Auditory Research Bulletin 2007 Biennial Edition

Auditory Research Bulletin 2007 Biennial Edition
Cover/qxp
1/22/08
10:28 AM
Page 1
Auditory Research Bulletin
DEC07 2-091520
© 2007 by Advanced Bionics Corporation or its affiliates. All rights reserved. Printed in U.S.A.
2007 BIENNIAL EDITION
www.BionicEar.com
Auditory
Research Bulletin
2007 BIENNIAL EDITION
auditory
Research Bulletin
bienniAl edition 2007
Published by Advanced Bionics® LLC, Valencia, California, USA. © 2008.
All rights reserved.
Production design and editing assistance for this publication were provided by
ReMarks Editing, Port Gamble, Washington. E: [email protected]
Advanced Bionics
Founded in 1993, Advanced Bionics is a global leader in implantable, hightechnology neurostimulation devices for the treatment of deafness. Advanced
Bionics is dedicated to improving the quality of life for hearing impaired
individuals through the application of advanced technology and the delivery
of high-quality products and services. As the only American cochlear implant
manufacturer, Advanced Bionics is committed to offering lifetime services and
support for our recipients.
Advanced Bionics®, Auria®, CII Bionic Ear®, CLARION®, Connect and Discover™,
Firefly®, Harmony™, HEARPRINT™, HiResolution®, HiRes®, HiRes Fidelity 120™*,
HiRes 90K™, HiFocus®, HiFocus Apex™, HiFocus Helix®, Platinum BTE®, Platinum
Series™, PowerCel™, SoundWave™, and T-Mic® are trademarks of Advanced
Bionics, LLC, Valencia, California, USA.
*In the United States, HiRes Fidelity 120 is an optional feature for adults only.
See package insert for details.
Contents
Contents
Preface. . . . .v
Forewords. . . .vi
Acknowledgments. . . .xi
List of Research Summaries. . . .xii
Basic Science, Electrode Development & Medical Applications. . . .17
HiResolution Sound Processing. . . .35
Psychophysics, Modeling & Novel Processing. . . .69
Processors & Accessories. . . .93
Bilateral Implantation. . .125
Music Studies. . .141
Objective Measures. . .155
Clinical Tools. . .179
Contributing Authors. . .196
Participating Research Centers. . .200
Research Staff Worldwide. . .206
Preface
More than 23,000 patients worldwide enjoy the listening benefits provided by cochlear
implant systems from Advanced Bionics. Over 17,000 patients are benefiting from
HiResolution® Sound—designed not only to enable speech understanding but also to
simulate the fullness and richness of sound experienced by normal hearing individuals in
real-world environments. With the new Harmony™ HiResolution Bionic Ear System,
we are continuing our efforts to bring unprecedented hearing performance to cochlear
implant recipients.
This biennial edition of the Auditory Research Bulletin highlights basic and clinical research
contributed by more than 250 dedicated professionals who are studying the hearing benefits of the Bionic Ear in today’s patients as well as exploring the potential for additional
benefits in the future. These efforts span the gamut from inner ear to central auditory
physiology, from psychoacoustics to speech perception, and from clinical programming to
everyday listening. We are proud to participate in these shared efforts that reach around
the globe.
We thank all of our partners in science for their tireless commitment to advancing the field
of cochlear implantation. Advanced Bionics looks forward to continuing its collaborative
research endeavors worldwide with investigators, researchers, and clinicians bringing the
gift of sound to individuals who are hearing impaired or deaf.
Jeffrey H. Greiner
President
Advanced Bionics
Forewords from the Americas
cochlear implants.
. .a technological miracle
In the field of bioengineering, cochlear implants represent a technology
that has evolved and produced clinical benefits far beyond our expectations.
I have had the privilege of witnessing and experiencing the evolution of
cochlear implant technology since William House developed the first single
channel implant in the early 1970s. Cochlear implants were introduced to
Latin America by the mid 1970s, and we began our program at Clínica Rivas
in Bogotá, Colombia, in the early 1990s. Today, over 40 cochlear implant
teams are well established in at least 10 Latin American countries including
Argentina, Brazil, Chile, Colombia, Mexico, and Venezuela.
Our team is pleased to participate in research with Advanced Bionics and
its Bionic Ear technology. We thank the many individuals worldwide who
have dedicated their lives to the development of a technology that can give to
deaf people, especially children, a new or second chance to enjoy the world of
sound and spoken language.
Jose Antonio Rivas, M.D.
Clínica Rivas, Bogotá, Colombia
vi Advanced Bionics® Auditory Research Bulletin 2007
. . .moving beyond language.
The collective impact that cochlear implants have had on the quality of
life of deaf individuals has been truly remarkable. Through a combination
of basic scientific and clinical advances, we can now offer patients a treatment for hearing loss that far exceeds therapeutic options for the loss of
any other sensory modality. From simple environmental sound awareness to
open-set speech perception in challenging listening conditions (for example,
cell phones), we have now arrived at the point where language perception
is reliably achieved in most recipients with postlinguistic deafness. In light
of these achievements, however, it is important to remember that the auditory world consists of more than spoken language. Of all auditory stimuli,
music represents in many ways the most challenging category of sound for
cochlear implant users. Music is an abstract stimulus capable of somehow
conveying emotion and inducing joy in the listener. The neural demands of
music perception—with multiple instrument lines, each with their own individual rhythmic contour and timbral features—are astonishingly high. As a
result, accurate music perception remains an elusive hope for most individuals
with cochlear implants.
The technological innovations of the Advanced Bionics HiResolution
processing strategy represent the most significant advance towards realizing the goal of sophisticated sound perception beyond language. Through a
process referred to as “current steering,” HiResolution Sound is now able to
provide directed electrical stimulation to neural regions between electrodes
rather than limiting stimulation at immediately adjacent sites. The implications of current steering for pitch perception are tremendous. Rather than a
limited number of pitch percepts linked to each individual electrode contact,
this technology allows us to utilize the concept of virtual channels to deliver
the rich spectrum of pitches that one may hear, whether listening to an iPod
or to a symphony orchestra. In providing a viable solution for the problem of
accurate pitch perception, current steering represents the most crucial first
step of progress beyond language. This bulletin provides an overview of the
research exploring these remarkable advances, which represent a true highlight in the evolutionary development of cochlear implant technology.
Charles J. Limb, M.D.
Johns Hopkins University, Baltimore, Maryland, USA
—continued—
vii
Forewords
Forewords from Asia Pacific & Europe
cochlear implants.
. .a global challenge
The majority of the world’s population is in the Asia-Pacific region. Although
there is a general lack of epidemiological data, we believe that the incidence
of hearing impairment is no less in Asia than in other parts of the world.
Clinicians know that the indicated treatment for an estimated 1.8 million
profoundly deaf individuals throughout the region is cochlear implant technology. However, the treatment demand can never be met given current
economics—except in certain relatively wealthier areas of Hong Kong, Korea,
Singapore, and Japan where the cost-effectiveness of cochlear implantation is
well appreciated.
Currently in China, merely two percent of the 30,000 profoundly deaf
newborns are implanted. Establishing the cost-effectiveness of cochlear
implantation is imperative if we are to solve the problem of widespread
under-treatment for these children. In addition, hearing screening, rehabilitation, and habilitation programs are required as well as information campaigns
that educate the general public about cochlear implant technology as a treatment option for deafness.
Advanced Bionics has been instrumental in continuing research for the
benefit of the deaf population in the Asia-Pacific region, as a key part of
meeting the global challenge. By sponsoring research on HiResolution sound
processing in tonal speakers, the company has joined with cochlear implant
professionals in the region to open new frontiers in our efforts to treat deafness. I am privileged to participate as a partner and facilitator in these research
efforts. I compliment Advanced Bionics on its mission to support health care
providers, and I wish this research bulletin success.
Michael C. F. Tong, M.D.
The Chinese University of Hong Kong, Hong Kong
viii Advanced Bionics® Auditory Research Bulletin 2007
. . .the challenge continues.
When I first became acquainted with the cochlear implant in the early 1970s,
it was not at all evident that one day we would not be surprised to see an
implant recipient answering the phone. At that time, physiologists warned
that such a crude system could not work and that deaf people should not
be subjected to such an experiment. However, history shows that perseverant
research and development can accomplish what once was considered impossible. The development of today’s cochlear implants would not have been
possible without exemplary cooperation between manufacturers and independent research groups.
In spite of the significant benefits experienced by today’s implant users,
some of the early challenges have not yet been resolved. Even though speech
perception by implant recipients is possible because of the impressive capacity
of the human brain to cope with degraded information, we may be facing
the limits of this capacity. Thus, the challenges for future implant development are many. Stimulation via a limited number of electrodes in an electrically conductive environment leads to much spread of excitation. Frequency
resolution in cochlear implants falls short of the frequency resolution in the
normal ear. Pitch perception is still difficult for implant recipients. It seems to
me that improving frequency resolution is a key issue in solving these problems. For example, the electrode-neuron interface must be improved. Moreover, more research is required to understand the contribution of temporal
fine structure to pitch perception, as well as to determine how to deliver that
temporal information to the auditory nerve. Although temporal fine structure
can be included relatively easily in coding strategies, clinical results have not
shown improved pitch perception thus far.
Although challenges continue to exist, much progress has been made in the
development of user-friendly hardware and software. Sound processors have
become smaller, electrodes hug the modiolus, fitting sessions are shorter, and
remote fitting has become an option. These advances are important contributions from the manufacturers to the field. Let us hope that close cooperation
between manufacturers, clinics, and research groups will continue to advance
the development of cochlear implants—always mindful that cochlear implant
recipients depend increasingly on these devices and, by extension, on us.
Guido Smoorenburg, Ph.D.
University Medical Center, Utrecht, the Netherlands
ix
Forewords
Acknowledgments
In the second half of 2006, Advanced Bionics introduced the Harmony™
HiResolution® Bionic Ear System, the newest of its cochlear implant systems designed
to provide usable hearing to individuals with severe-to-profound deafness. The Harmony
system represents a giant step in cochlear implant design by combining the latest in DSP
technology with advanced knowledge about electrical hearing in an ongoing endeavor to
deliver sound quality that far surpasses the capability of earlier devices from Advanced
Bionics and its competitors.
Cochlear implants have served as a technological compensation for deafness for over two
decades. Until the advent of Harmony, implant design and testing has focused on helping
individuals to understand conversational speech in quiet. As a result, most cochlear implant
users today have little difficulty hearing speech when the acoustic conditions are favorable.
However, hearing in everyday life involves much more. Implanted listeners should be able
to hear in noise, to experience subtle environmental sounds, to hear when speakers are
far away—and to enjoy music. In other words, implant users should be connected to the
entire world of sound. To this end, the new Harmony HiResolution Bionic Ear System
implements the family of HiResolution Sound (HiRes®) processing strategies, designed to
faithfully represent the three dimensions of sound—intensity (loudness), frequency (pitch),
and time. Through innovations like CD-quality 16-bit front-end analysis, dual-action
AGC, and current steering, HiRes sound processing preserves and delivers the acoustic
environment with high fidelity, thereby offering recipients an unprecedented opportunity
to maximize their hearing potential in the real world. Moreover, HiRes sound processing is
available to all Advanced Bionics implant users who were implanted after 2001 (including
all CII Bionic Ear® and HiRes 90K™ recipients).
The Auditory Research Bulletin Bienniel Edition 2007 is a compilation of reports on scientific and clinical studies conducted with the Harmony system and HiResolution Sound.
Included are contributions from surgeons and clinicians who describe the everyday listening
benefits experienced by Bionic Ear users and from basic researchers who are exploring
new ways to deliver electrical stimulation and to enhance electrical hearing. (An alphabetical index of contributing authors and a list of participating research centers appear at
the end of the bulletin.) Advanced Bionics thanks these investigators and clinicians for
their research efforts and their willingness to contribute to this publication. We also thank
the Research and Development and Clinical Research teams at Advanced Bionics for their
commitment to these collaborative efforts.
We hope you find this edition of the Auditory Research Bulletin a valuable reference that
expands your personal knowledge and supports your counseling of candidates and recipients. We look forward to continued collaboration as we explore further the capabilities of
the Harmony HiResolution Bionic Ear System.
Mary Joe Osberger, Ph.D.
Director, Global Clinical Research
Advanced Bionics
xi
List of Research Summaries
Basic Science, Electrode Development & Medical Applications
17
Slim Lateral Electrode: Histological Evaluation Using a Modified Polishing Technique................. 18
Micro-CT Anatomical Measurements in the Human Cochlea....................................................... 20
Contralateral Effects of Unilateral Cochlear Implantation on Auditory Nerve Synapses
in Congenitally Deaf White Cats............................................................................................. 22
Effects of Intracochlear High-Rate Electrical Stimulation on the Spontaneous Activity
of the Higher Auditory Pathway............................................................................................................25
An Animal Model for Electrical Stimulation of the Inferior Colliculus.......................................... 26
The Effect of Unilateral Multichannel Cochlear Implantation on Bilateral Tinnitus....................... 29
Cochlear Implantation in Unilateral Deafness Associated with Ipsilateral Tinnitus:
A Case Study........................................................................................................................... 30
Using Electrical Stimulation to Suppress Tinnitus......................................................................... 32
HiResolution Sound Processing
HiRes with Fidelity 120: Speech Recognition and Strategy Preference
Multicenter Study (1) in North America................................................................................... 36
HiRes with Fidelity 120: Everyday Listening Benefits
Multicenter Study (2) in North America................................................................................... 38
HiRes with Fidelity 120: Sound Quality and Music Listening
Multicenter Study (3) in North America................................................................................... 40
HiRes 120 Benefits in Adults: Multicenter Study in Europe.......................................................... 42
Sound Quality and Music Perception with HiRes 120: Multicenter Study in Europe ................... 44
HiRes with Fidelity 120 Benefit in Native Speakers of Korean...................................................... 48
Pediatric Results Following Conversion to HiRes 120 from Standard HiRes on the
Harmony Processor................................................................................................................. 53
Programming Children in HiResolution Sound............................................................................ 54
Speech Perception in Children Using Two Generations of Advanced Bionics
Cochlear Implants: Three-Year Results..................................................................................... 56
Auditory Skill Development in Newly Implanted Children: Effect of Sound Processing................ 58
xii Advanced Bionics® Auditory Research Bulletin 2007
35
HiResolution Sound Processing—continued
Auditory and Language Abilities in Children: Comparison of Two Different
Cochlear Implant Systems....................................................................................................... 60
HiRes 120 in Children Using the HiRes 90K Device: Preliminary Report..................................... 62
HiRes 120 and Lexical Tone Perception in Mandarin-Speaking Children..................................... 64
Psychophysics, Modeling & Novel Processing
69
Channel Interaction Patterns with Simultaneous Stimulation: Psychophysical
and Physiological Measures.................................................................................................... 70
Psychophysical Measurement of Spread of Excitation Using Single and Dual Maskers................. 72
Effects of High-Rate Pulse Trains on Modulation Detection in Cochlear Implant Users................. 74
Effects of Stimulation Rate on Fundamental Perceptual Abilities
Underlying Speech Perception................................................................................................ 76
Electrode Discrimination in Noise in Cochlear Implant Recipients.............................................. 78
Spectral Peak Resolution Assessment of Cochlear Function.......................................................... 80
The HRStream Research Interface................................................................................................ 82
A Novel Signal Processing Strategy for Current Steering in Cochlear Implants.............................. 84
Comparison of Sound Processing Strategies: FFT vs. Sinusoidal Approximation (SinEx)
Implemented with Current Steering......................................................................................... 86
Current Steering and Current Focusing in Cochlear Implants: Comparison of Monopolar,
Tripolar, and Virtual Channel Electrode Configurations............................................................ 88
Processors & Accessories
93
Evaluation of the Harmony Sound Processor: Multicenter Study in North America...................... 94
HiRes and HiRes 120 on the Harmony: Results in Experienced Users
of Current Steering Strategies................................................................................................... 96
Results Using the Harmony Sound Processor with HiRes and HiRes 120 Processing................... 98
Input Dynamic Range: Effect on Speech Perception in Noise....................................................... 100
—continued on next page—
List of Research Summaries xiii
List of Research Summaries—continued from previous page
Processors & Accessories—continued
Harmony Processor and T-Mic Benefits in Children: Speech Perception....................................... 102
Harmony Processor and T-Mic Benefits in Children: Questionnaire Results................................. 104
Benefits of Using Directional Microphones in a Diffuse Noise Environment................................ 106
Remote Microphone Technology Benefits in Adults..................................................................... 108
iConnect/FM Benefit in Adults..................................................................................................... 110
Evaluation of the iConnect FM Adapter in Adults......................................................................... 112
Evaluation of the iConnect FM Adapter in Children..................................................................... 114
iConnect/FM System Benefits in School-Aged Children............................................................... 116
iConnect Benefits in Children...................................................................................................... 118
Optimizing FM Systems for the Auria Speech Processor............................................................... 120
Dual-Loop AGC Improves Speech Perception in Noise for CII and HiRes 90K Implant Users....... 122
Bilateral Implantation
125
Simultaneous Bilateral Cochlear Implantation in Adults: Comparison of Bilateral and
Unilateral Benefit Multicenter Study (1) in North America....................................................... 126
Simultaneous Bilateral Cochlear Implantation in Adults: Everyday Listening Benefits
Multicenter Study (2) in North America................................................................................... 128
Bilateral Benefit in Twins Implanted Simultaneously with Two Devices
before Twelve Months of Age.................................................................................................. 130
Pediatric Bilateral Implantation: Two Case Studies....................................................................... 138
Music Studies
Music Benefits in Adults Using HiRes with Fidelity 120: Multicenter Study in North America...... 142
Music Perception in Adult HiRes 120 Users................................................................................. 144
Development of Musical Behavior in Children with Cochlear Implants....................................... 146
Video Assessment of the Development of Musical Behavior in Children with Cochlear Implants.....149
Effects of Pitch, Rhythm, and Melody Training on Speech and Music Development in Children
with Cochlear Implants........................................................................................................... 150
The CIRCLE Game: A Music Education Program for Pediatric Cochlear Implant Users................. 152
xiv Advanced Bionics® Auditory Research Bulletin 2007
141
Objective Measures
155
European Multicenter Objective Measures Study......................................................................... 156
Smart NRI Multicenter Study in Europe: Preliminary Results........................................................ 160
European Banded Neural Response Imaging (NRI) Study: One-Year Results................................. 162
Banded Neural Response Imaging (NRI) Measurements: Comparison to Single-Channel
NRI Responses and Psychophysical Measurements Over Time................................................ 164
Relationship between Electrical Stapedial Reflex Thresholds and Speech Perception
in Adult Users of the HiResolution Bionic Ear.......................................................................... 166
Evaluation of Neural Refractoriness as a Tool for Setting Implant Program Levels......................... 168
Comparison of Intraoperative Stapedius Reflex Thresholds and Postoperative M Levels................ 170
Relationships among tNRIs, ESRTs, and Psychophysical Measures of Loudness: Implications
for Programming the HiResolution Bionic Ear......................................................................... 172
Amplitude Growth and Spatial Spread of Neural Responses in HiRes 90K Users......................... 175
Channel Interaction in Cochlear Implant Users............................................................................ 176
Clinical Tools
179
Everyday Listening Questionnaire: A New Evaluation Tool for Adult Cochlear Implant Users....... 180
Cochlear Implant Outcomes and Quality of Life in Adults with Prelinguistic Deafness................ 182
More than the Words: Cochlear Implant Users’ Perception of the “Indexical” Properties
of Speech................................................................................................................................ 184
DVD Version of the UCL-CUNY Sentence Test............................................................................. 188
Hearing with a Cochlear Implant: The Everyday Listening Questionnaire I................................... 190
Impact of Cochlear Implant Technology on Everyday Listening: The Everyday Listening
Questionnaire II...................................................................................................................... 192
Using Tele-Medicine to Program Cochlear Implants Remotely..................................................... 194
List of Research Summaries xv
Basic Science, Electrode Development &
Medical Applications
Clinically, cochlear implantation provides unmistakable benefit for hearing
sound and understanding speech. New basic research is defining the anatomical and physiological processes underlying the perceptual benefits of electrical
stimulation as well as exploring alternative sites for stimulating the auditory
system—especially the inferior colliculus. Clinical research is expanding the
applications of cochlear implantation not only for restoring hearing sensation
but also for alleviating debilitating tinnitus.
This section summarizes various investigations into these clinical research
areas that are taking place around the globe.
17
Slim Lateral Electrode: Histological Evaluation Using a
Modified Polishing Technique
Thomas Lenarz, Prof., M.D., Ph.D.1
Timo Stöver, M.D., Ph.D.1
Peter Erfurt1
Yassaman Khajehnouri, Dipl.-Inform.2
1 Medizinische Hochschule Hannover, Hannover, Germany
2 Advanced Bionics, Europe
“The Slim Lateral prototype was easily
inserted via the round window—a key
design goal for an atraumatic electrode.”
Figure 1. Temporal bone implanted with the Slim Lateral electrode array. The electrode array is confined to scala tympani
and sits in a lateral position.
Figure 2. X-ray image showing the Slim Lateral array inserted to
approximately 180 degrees in a human cochlea. The electrode
sits in a relatively lateral position within the scala tympani.
18
The HiFocus Slim Lateral electrode array is a prototype design based on the HiFocus 1j array but with
approximately half of the cross sectional area. The
largest array dimension varies between 0.42 mm at
the tip to 0.6 mm at the base. An insertion depth
of between 16.5 and 19.5 mm is achieved when
inserting to either the first or second marker contacts
at the proximal end of the array. The Slim Lateral
electrode has 16 medial facing electrode contacts
spread evenly along a 13.4 mm active length. The
design was created to address cases where there is
a desire to preserve residual low-frequency acoustic
hearing. However, the length of the array is long
enough so that, if residual hearing were lost over
either the short or long term, electrical stimulation alone still could support high levels of speech
perception.
The original Slim Lateral electrode design was
revised to improve the stability of the array and to
minimize trauma to cochlear structures. The current
prototype has a more rounded tip than earlier
designs, the angle of the collar has been modified to assist sealing at the round window, and a
softer silicone is used to promote better bending—
particularly towards the electrode tip. To simplify
the surgical process, a release button has been added
to the insertion tool.
The aim of this study was to evaluate insertion
trauma for the most recent Slim Lateral design.
Insertions were made into five fresh frozen human
temporal bones. All insertions were made via the
round window. First, the promontory was drilled
away and an incision was made in the round window
membrane. The electrode then was introduced
through the incision. An insertion tool with a 5-mm
stylet inserted into the proximal end of the array
acted as a handle, giving the surgeon control of the
array during surgery. When the array was inserted
to the desired insertion depth, a button was pressed
Advanced Bionics® Auditory Research Bulletin 2007
Recent Public at ion
In Vivo Estimates of the Position of
Advanced Bionics Electrode Arrays in the
Human Cochlea
on the insertion tool to release the electrode array.
Insertions were uneventful in all five bones.
Next, the five bones were prepared for histological
analysis. Each fixed bone, with the electrode array
still in place, was progressively ground down to
reveal a flat surface equivalent to a cut through the
bone. The exposed surface was polished and then
photographed. By repeating this process many times,
a series of images were created showing the detail
within the temporal bone.
Margaret W. Skinner, Ph.D.1
Timothy A. Holden, B.S.E.1
Bruce R. Whiting, Ph.D.1
Arne H. Voie, Ph.D.3
Barry Brunsden1
J. Gail Neely, M.D.1
Eugene A. Saxon, M.Eng., M.M.E.2
Timothy E. Hullar, M.D.1
Charles C. Finley, Ph.D.3
1 Washington University, St. Louis, MO, USA
2 Spencer Technologies, Seattle, WA, USA
3 University of North Carolina, Chapel Hill, NC, USA
This study describes a new microscopy technique for
determining the position of electrode arrays within
the cochlea. Data were compared to spiral computed
tomography (CT) data from 15 patients implanted with
The images revealed that all five electrode arrays
were located entirely in the scala tympani with no
obvious basilar membrane damage. The electrode
array location was always lateral (Figure 1). No tip
was observed to be folded back onto itself. The mean
insertion depth was 188 degrees (range: 180-210).
Figure 2 shows an x-ray of a typical electrode array
position.
the Advanced Bionics HiFocus I, Ij, or Helix arrays.
These temporal bone findings are encouraging and
indicate that the current Slim Lateral design is an
important step towards creating an atraumatic
electrode array. The grinding histological method
produced excellent images, allowing the location of
the electrode array to be observed all the way through
the temporal bone. This technique also minimized
histological preparation trauma that might be
confused with any insertion-related trauma.
tissue boundaries in the patient scan data to choose the
Ongoing clinical studies with a variety of electrode
designs suggest that round-window insertion can
lead to high levels of hearing preservation. The Slim
Lateral prototype was easily inserted via the round
window—a key design goal for an atraumatic electrode. Nonetheless, if individual anatomy prevents
a round window approach, the Slim Lateral array
could be inserted via a cochleostomy of no more
than 1.0 mm in diameter.
Specifically, ANALYZE imaging software was used to
register 3D-image volumes from patients’ preoperative
and postoperative scans and from a single body donor
whose unimplanted ears were scanned clinically with
Micro-CT and with orthogonal-plane fluorescence optical
sectioning (OPFOS) microscopy. Using this registration,
the atlas of OPFOS images of soft-tissue within the
donor cochlea were compared with the bone and fluid/
midmodiolar axis position and judge electrode position
in scala tympani (ST) or scala vestibuli (SV) including
distance to medial and lateral scalar walls. The angular
rotation 0o start point is a line joining modiolar axis and
middle of cochlear canal entry from the vestibule. Group
mean array insertion depth was 477o (range: 286o – 655o).
Word recognition scores were negatively correlated (r =
-0.59; p = 0.028) with the number of electrode contacts in
SV. Further refinement of the technique may help develop
improved surgical insertion techniques and assist in the
clinical programming of cochlear implant systems based
upon individual patient characteristics.
Ann Otol Rhinol Laryngol (2007) 116(4) Suppl197:1-24.
19
Basic Science, Electrode Development & Medical Applications Micro-CT Anatomical Measurements
in the Human Cochlea
Andrzej Zarowski, M.D.1
Erwin Offeciers, Prof., M.D., Ph.D.1
Andrei Postnov, Ph.D.2
Filiep Vanpoucke, Dr.Ir.3
Stefaan Peeters, Prof. Dr.Ir.2
1 Medical Center Sint Augustinus, Wilrijk, Belgium
2 University of Antwerp, Antwerp, Belgium
3 Advanced Bionics European Research Center,
Antwerp, Belgium
Micro-CT is an innovative computerized tomography technique that is highly optimized for accuracy.
It allows for scanning of samples with dimensions in
the centimeter range including in vivo scanning of
small laboratory animals. Excellent resolution (down
to 5 μm) makes micro-CT a potential improvement
over standard histological techniques (Postnov et al,
2006).
As examples, Figure 1 shows the comparison
between a Micro-CT image and a corresponding
histological section. The main advantage of microCT is minimal sample manipulation that results in
artefact-free data sets.
Figure 1. Micro-CT (top) and histological (bottom) cross
sections of a human cochlea.
20
This study obtained anatomical data on human
cochleas and characterized the variability of the size
and the length of selected intracochlear structures,
including the cochlear scalae, the organ of Corti, and
Rosenthal’s canal. Eight human cadaveric temporal
bones (four right, four left) were scanned with the
Skyscan-1076 Micro-CT machine. A resolution of
18μm, a Ti-filter, and scanning times of up to four
hours were used in order to achieve good visualization of the intracochlear soft tissues. Acquired
shadow projections were reconstructed as virtual
slices. For further analysis, a custom Matlab software
tool was created. First, in order to have a uniform
description of the intracochlear anatomy, the slice
orientation was rotated to “Cochlear View”—
that is, reconstructions were made perpendicular to
the axis of the internal auditory canal. Next, twodimensional cross sections through the cochlear
scalae were determined according to the angular
position (every 30 grades). For accurate evaluation of
the size of a cross section, the software tool allowed
a perpendicular orientation of these two-dimen-
Advanced Bionics® Auditory Research Bulletin 2007
This study indicates that micro-CT is a very accurate
technique that provides an artefact-free quasi-histological quality evaluation of intracochlear anatomy.
Micro-CT has the potential to assist surgeons and
audiologists in determining the position of a cochlear
implant electrode array to optimize implant benefit.
Height (mm)
Surface (mm2)
2
Millimeters
The shape and sizes of the human cochleas showed
substantial variability. Average dimensions and standard deviations for the most important intracochlear
structures were measured. As an example, Figure 2
shows the anatomical dimensions and variability
for the scala tympani. Figure 3 shows the cumulative distance from the round window for the medial
and the lateral side of the scala tympani, the organ
of Corti, and Rosenthal’s canal. These figures can
be used to estimate whether an implanted electrode
array is located laterally or periomodiolarly based on
its insertion angle and insertion length.
Width (mm)
2.5
1.5
1
.5
0
0
100
200
300
400
500
600
700
Angle (Degrees) Relative to the Round Window
Figure 2. Scala tympani dimensions and variability.
Cumulative Distance from RW (mm)
sional sections relative to the lateral cochlear wall.
In each of the rotated two-dimensional sections, a
set of 10-20 characteristic points that describe the
intracochlear topology were marked—for example,
the “corner” points of the bony and the membranous
labyrinth, the membranes (basilar and Reissner’s),
the position of the organ of Corti, and the location of Rosenthal’s canal. The characteristic points
were used for calculation of the distances necessary
for evaluation of the anatomical variability of each
intracochlear structure.
“Micro-CT has the potential
to assist surgeons and audiologists
in determining the position
of a cochlear implant electrode array
to optimize implant benefit.”
Scala Tympani
Lateral Side
40
Scala Tympani
Modiolar Side
Organ of Corti
Rosenthal’s Canal
35
30
25
20
15
10
5
0
0
100
200
300
400
500
600
700
Angle (Degrees)
Reference
Postnov A, Zarowski A, De Clerck N, Vanpoucke F, Offeciers FE, Van Dyck D,
Peeters S. (2006) High resolution micro-CT scanning as an innovative tool for
evaluation of the surgical positioning of cochlear implant electrodes. Acta
Otolaryngol 126(5):467-474.
Figure 3. Cumulative distance from the round window for the
medial and the lateral side of the scala tympani, the organ of
Corti, and Rosenthal’s canal.
21
Basic Science, Electrode Development & Medical Applications Contralateral Effects of Unilateral Cochlear Implantation
on Auditory Nerve Synapses in Congenitally Deaf
White Cats
Jahn N. O’Neil, Ph.D.
Christa Baker, B.S.E.
David K. Ryugo, Ph.D.
Johns Hopkins University, Baltimore, MD, USA
Our long-term study goals are to understand the
neural mechanisms of hearing, to explore the consequences of deafness on brain development and
function, and to determine the restorative effects of
cochlear implantation on the central auditory pathways. The congenitally deaf white cat provides an
animal model that enables us to examine pathologic
changes such as abnormal circuitry, neuronal death,
and synaptic anomalies that have been widely associated with congenital deafness. This model mimics
the Scheibe deformity in humans featuring early
onset, profound hearing loss.
Our previous studies in young, congenitally deaf
cats have shown that unilateral cochlear stimulation
largely reverses the synaptic abnormalities between
auditory nerve fibers and spherical bushy cells
(SBCs) caused by congenital deafness. The significance of this observation is that we have identified
a specific pathology and can begin studying the
phenomenon with the goal of understanding the
neurobiology of synaptic plasticity and developing
intervention treatment strategies.
Unilateral hearing, however, remains a problem for
the listener in terms of speech recognition and spatial
hearing. With a growing rise of patients receiving
two cochlear implants, it is timely to begin analyzing
and comparing brain responses to unilateral versus
bilateral cochlear implants, addressing one central
question: Is there physical evidence of brain changes
that is consistent with favoring bilateral over unilateral implantation?
One of the largest synaptic endings in the brain
originates from myelinated auditory nerve fibers and
is known as the endbulb of Held. Endbulbs transmit
signals from the auditory nerve to the postsynaptic
SBCs in the anteroventral cochlear nucleus and are
implicated in the pathways that process the temporal
features of sound. Endbulbs are highly branched in
animals with normal hearing and are significantly
less branched in congenitally deaf animals. The brain
region where auditory nerve fibers make functional
contact with SBCs is known as the cochlear nucleus.
This region is the focus of our studies because it
receives incoming signals from the cochlea and gives
rise to the ascending auditory pathway. Pathologic
changes at this site would disrupt the processing of
auditory signals at higher brain centers.
In the present study, congenitally deaf kittens were
unilaterally implanted between three to six months
of age with a six-channel cochlear implant. These
animals received electrical stimulation for 35 hours
per week over a period from 9-24 weeks with a
custom-built CII sound processor. The electrode
selection and M-level settings of the sound processor
were determined for each animal using behavioral
cues and electrical compound action potentials
(ECAPs). After a brief training period, the animal
learned that a specific sound signaled a food reward,
indicating that the cochlear implant enabled the
processing of biologically relevant stimuli.
22
Advanced Bionics® Auditory Research Bulletin 2007
At the end of the stimulation period (two to three
months), tissue was collected for histologic analysis.
The pre- and postsynaptic features of the endbulb of
Held were examined using light and electron microscopy in normal hearing cats, congenitally deaf cats,
and congenitally deaf cats with a unilateral cochlear
implant. We used auditory nerve synapses contralateral to the cochlear implant as a control structure
for the implanted cats—a strategy providing us with
four separate cohorts of measurements.
As shown in Figure 1, endbulbs are represented in
electron micrographs as a large profile adjacent to
the SBC. They contain clear, round synaptic vesicles
and exhibit multiple asymmetric membrane thickenings with the postsynaptic cell body. Because the
endbulb is a large and elaborately branched structure, the sections are very thin and the magnification
relatively high. Therefore, only part of the endbulb is
seen in electron micrographs. In normal hearing cats,
the postsynaptic densities (PSDs) are marked by a
dome-shaped membrane thickening that houses the
transmitter receptors. These structures are assumed
to represent synaptic release sites (Figure 1A). In
congenitally deaf cats, the PSDs are longer and
flatter and there is a notable loss of synaptic vesicles
(Figure 1B). Synapses ipsilateral to the cochlear
implant are restored to their normal appearance
(Figure 1C). Surprisingly, the synapses contralateral
to the cochlear implant also were affected (Figure
1D). These synapses exhibited a mixture of features
that reflected synapses from both normal and deaf
cats. Some PSDs were dome-shaped and small, but
many were flat and larger. These results demonstrate that there are definable effects on synaptic
morphology for auditory nerve synapses contralateral to the cochlear implant.
Three-dimensional reconstructions of endbulb
synapses through serial sections emphasize PSD
“The results so far indicate that
cochlear implantation prevents or reverses
synaptic abnormalities that are induced
by congenital deafness.”
Figure 1. Electron micrographs of representative auditory
nerve endbulbs (EB) in the cochlear nucleus from (A) a normal
hearing cat, (B) a congenitally deaf cat, (C) the ipsilateral side
to a cochlear implant in a congenitally deaf cat, and (D) the
contralateral side to a cochlear implant in a congenitally deaf
cat. Postsynaptic densities (PSDs) appear as asymmetric membrane thickenings (*). Note that the size and shape of PSDs
in the cochlear nucleus ipsilateral to the cochlear implant
appear identical to those from a normal hearing cat (A). In
contrast, many of those from an untreated deaf cat exhibit flattened and hypertrophic synapses. The contralateral terminals
from implanted cats exhibit morphology whose characteristics
lie between those of normal hearing and deaf cats. That is,
auditory nerve synapses are clearly affected by the cochlear
implant even though it is indirect. (Note: SBC = spherical
bushy cell.)
—continued on next page—
23
Basic Science, Electrode Development & Medical Applications —continued from previous page—
“These data are highly relevant to issues
regarding which ear to implant
when one is better than the other,
as well as considerations of
the interval between implantation of a
second cochlear implant when
bilateral implants are considered.”
Figure 2. En face view of postsynaptic densities reconstructed
through serial sections and rotated. The congenitally deaf cats
exhibit hypertrophied synapses—in contrast to the normal
shapes and distributions in the stimulated deaf cats. Postsynaptic densities in the cochlear nucleus contralateral to implant
appear to be modestly affected by stimulation.
differences between the cohorts of cats. As illustrated in Figure 2, the en face view presents the SBC
membrane that lies under the auditory nerve ending.
The shaded and lined regions represent the reconstructed synapses. Statistical analysis revealed that
there were no significant differences between the size
of postsynaptic densities from the auditory nerve
fibers of normal-hearing and implanted cats (ipsilateral). However, the PSDs of congenitally deaf cats
were significantly larger (Ryugo et al, 2005), whereas
those on the contralateral side appear to exhibit a
partial “savings” that preserves synaptic morphology.
This work is still in progress as we seek to determine
to what extent unilateral cochlear implant stimulation maintains auditory nerve synapses from the
contralateral deaf ear.
The results so far indicate that cochlear implantation prevents or reverses synaptic abnormalities
that are induced by congenital deafness. Data gathered to date are limited to unilateral implantation
in the deaf white cat model. It is noteworthy that
unilateral cochlear implantation not only has direct
effects on the auditory system, but also has indirect
effects to help preserve the structure of synapses
in the contralateral ear. Perhaps it is not surprising
that activity in one auditory nerve has an impact
throughout the central auditory system given the
presence of commissural and descending pathways.
In light of the relationship between structure and
function, it follows that the indirect stimulation by
the unilateral cochlear implant will exert functional
change as well. These data are highly relevant to
issues regarding which ear to implant when one is
better than the other, as well as considerations of the
interval between implantation of a second cochlear
implant when bilateral implants are considered.
Future studies are proposed that will compare the
effects of unilateral versus bilateral cochlear implantation on auditory structure and function.
Reference
Ryugo DK, Kretzmer EA, Niparko JK. (2005) Restoration of auditory nerve synapses
in cats by cochlear implants. Science 310 (5753):1490-1492.
24
Advanced Bionics® Auditory Research Bulletin 2007
Effects of Intracochlear High-Rate Electrical Stimulation
on the Spontaneous Activity of the Higher Auditory Pathway
Dietmar Basta, Ph.D.1, 2
Romy Goetze, M.Sc.2
Moritz Groeschel, M.Sc.2
Oliver Janke, M.D.2
Patrick Boyle, M.Sc.3
Arne Ernst, M.D.1
1 Unfallkrankenhaus Charité Medical School, Berlin, Germany
2 Humboldt University, Berlin, Germany
3 Advanced Bionics, Europe
The advantage of high-rate stimulation up to 5000
pulses per second per channel (pps/ch) for speech
perception in cochlear implant recipients has been
well demonstrated. The neurophysiological consequences of stimulation rates above 800 pps (i.e., the
expected maximum firing rate of a single auditory
nerve fiber) also have been observed (Oertel et al,
2002). For example, higher rates can decrease the
amplitudes of electrical evoked compound action
potentials and lengthen their latencies. Some studies
indicate that increasing stimulation rates abolishes
EABR responses for several hours and prolonged
stimulation at higher rates increases EABR thresholds (Tykocinski et al, 1995; Xu et al, 1997). In
contrast, other studies demonstrate that EABR
thresholds and survival rate of spiral ganglion cells
do not depend on stimulation rate (Cords et al,
1997; Reuter et al, 1997).
Because the data concerning the effects of stimulation rate are inconsistent, the upper rate limit for
safe and physiologically acceptable electrical stimulation is not clear. Another important point is that
earlier studies primarily used summating potentials,
which are not sufficiently sensitive to monitor minor
cellular or synaptic changes—indicating the initiation of long-term physiological changes following
high-rate electrical stimulation. Thus, it is important to investigate the electrophysiological consequences of chronic high-rate electrical stimulation
on the auditory pathway at a cellular level—that is,
the effect of stimulation on spontaneous neuronal
activity. Moreover, because a change in spontaneous
activity in the neuronal network of higher auditory
centers could contribute to a change in the auditory
perception of cochlear implant users, this information is clinically relevant.
Therefore, this study investigated the effect of chronic
intracochlear high-rate stimulation on the spontaneous activity in living brain slices of the medial
geniculate body (MGB) and primary auditory cortex
(AC). Sixteen guinea pigs were implanted unilaterally with a HiRes 90K implant and HiFocus electrode array. After recovery from surgery, the implants
were activated in eight animals with a stimulation
rate of 5000 pps/ch and a pulse duration of 11 µs
per phase using the HiRes strategy. The other eight
animals were not activated and served as controls.
The speech processor parameters were set based on
NRI data and behavioral reactions to sound. After
12 weeks of daily use (16 hours per day) in a standardized species-relevant acoustical environment,
single-unit recordings were made in living brain
slices of the MGB and AC.
Based on recordings from 334 neurons, no statistically significant differences were found in spontaneous activity between the implanted and
nonimplanted sides of the MGB or AC in the
activated animals. There also were no differences in
spontaneous activity between the study and control
animals on the implanted sides. This finding indicates that there were no negative effects of high-rate
electrical stimulation found in the higher auditory
system. Subsequent to these data, the effects of
chronic intracochlear high-rate electrical stimulation
on the lower auditory pathway now is under investigation. The results of these studies together will
improve understanding of the safety and efficacy of
stimulation parameters (rate, duration) for cochlear
implant users.
References
Cords SM, Reuter G, Issing P, Lenarz T. (1997) Development of the evoked auditory
brain stem response amplitude peak IV during chronic electrical, intracochlear,
multichannel high-rate stimulation. Am J Otol 18(6 Suppl.):S22-S23.
Oertel D, Fay RR, Popper AN. eds. (2002) Integrative Functions in the Mammalian
Auditory Pathway, Springer Handbook of Auditory Research, Vol. 15, New York:
Springer.
Reuter G, Cords SM, Issing P, Lenarz T. (1997) Acute and chronic effects of electrical
intracochlear multichannel high rate stimulation on the auditory brainstem response
in neonatally deafened kittens. Adv Otorhinolaryngol 52:46-51.
Tykocinski M, Shepherd RK, Clark GM. (1995) Reduction in excitability of the
auditory nerve following electrical stimulation at high stimulus rates. Hear Res
88(1-2):124-142.
Xu J, Shepherd RK, Millard RE, Clark GM. (1997) Chronic electrical stimulation
of the auditory nerve at high stimulus rates: a physiological and histopathological
study. Hear Res 105(1-2):1-29.
25
Basic Science, Electrode Development & Medical Applications An Animal Model for Electrical Stimulation of the
Inferior Colliculus
Douglas C. Fitzpatrick, Ph.D.
University of North Carolina, Chapel Hill, NC, USA
Some deaf individuals cannot be helped by a cochlear
implant because they do not have an auditory nerve.
The most common cause of bilateral loss of the auditory nerve is neurofibromatosis type 2 (NF2). Some
NF2 patients currently are treated with a brainstem
implant in the cochlear nucleus. However, benefits
tend to be limited to detection of environmental
An alternative site for stimulation is the inferior
colliculus (IC), which plays a central role in the
auditory pathways. In contrast to the cochlear
nucleus—which has three primary divisions, each
with a separate tonotopy and distinct cell types—
the central nucleus of the IC has a single tonotopic
representation of frequency and a homogeneous
cellular structure consisting of a patterned array of
only two cell types.
0.2
0.1
Response (mV)
0.0
-0.1
0.6
0.4
0.2
1
2
0.0
-0.2
3
0
0.5
1
1.5
2
2.5
Time (ms)
Figure 1. Intra-IC evoked potentials. (A) Response at one suprathreshold intensity. The stimulus was a biphasic current pulse
with alternating phase to reduce the artifact. The fast recovery
of recording in the presence of the large artifact is possible
because of a high-bandwidth custom amplifier. Blue represents
the unfiltered response. Red represents the response after filtering (100-20,000 Hz) to reduce artifact. (B) Illustrates a level
series showing threshold and >20-dB dynamic range. Arrow 1:
At 75 µA, an effect of the electrical artifact prior to the peak of
the neural response is evident. Arrow 2: Threshold was 15 µA.
Arrow 3: The response was still increasing at 150 µA for at least
a 20-dB dynamic range.
26
sounds and improved ability to read lips (Colletti
& Shannon, 2005). In contrast, patients who have
lost auditory nerve function as a result of trauma can
achieve a high degree of speech recognition through
cochlear nucleus stimulation alone. These results
suggest that either the tumor or its removal causes
damage to the cochlear nucleus.
We have been developing an animal model to study
physiological and behavioral responses to electrical
stimulation in the IC. This rabbit model offers several
advantages including no requirement for anesthesia
during physiological recordings, chronic implantation allowing long term studies, and the ability to
conduct physiological and behavioral experiments in
the same animal. The electrode is a “Michigan-type”
design (NeuroNexusTech, Inc.) with 16 iridium
contacts on a silicon carrier separated by 200 µm over
a 3 mm distance. Each 1250 µm2 contact is activated
by cyclic voltammetry to achieve a final impedance
of approximately 100 kOhms. Most importantly,
because the IC is deep in the brain, there is a 7 mmlong ribbon cable providing separation between the
electrode and an Omnetics connector mounted on
the skull.
The physiological measurements consist of intraIC evoked potentials. An example is shown in
Figure 1A. The stimulus was a 30 µs/phase biphasic
pulse with an amplitude of 45 µA in a monopolar
Advanced Bionics® Auditory Research Bulletin 2007
A level series of the post-filtered responses is shown
in Figure 1B. In the positive phase of the response,
the peak increased systematically up to a current
level of 75 µA (arrow 1). At this stimulation level, the
artifact caused an early initial rise before the neural
response reached its maximum. For higher stimulation levels, the artifact dominated the positive phase,
so the negative phase is the more reliable measure
of the neural response. The threshold current for the
negative phase to reach more than three standard
deviations from the noise floor was 15 µA (arrow 2).
The response increased systematically as the current
level increased up to the maximum of 150 µA
(arrow 3) for a dynamic range of at least 20 dB. The
time to the peak response shortened, as expected for
a neural potential. In this animal, these responses
have been stable for more than eight months.
The behavioral sensitivity to electrode stimulation
was measured in the same animal. The stimulus
was a 50-Hz train of 60 µs/phase biphasic pulses
E8
E10
E12
E14
E16
Stimultating Electrode (E11)
Level = 30
Level = 60
Level = 90
Level = 120
«
The pattern of responses across the array reflects the
current flow through the IC (Figure 2). Responses
that have some “width” to them are neural responses,
while those that are extremely narrow, such as in
the upper left hand corner, reflect only artifact. The
largest responses were obtained from the closest
contacts. There was asymmetry, with more current
flowing toward contacts deep to the stimulating
electrode (right of the dotted line).
E2 electrode
E4 11 E6
Stim
level=15 uA
Response (mV)
configuration. The initial phase of each pulse was
alternated to reduce stimulus artifact. The recording
electrode was an adjacent contact. A fast recovery
from saturation (due to the electrical artifact) was
obtained by using a high-bandwidth, custom amplifier and is apparent in the unfiltered response (blue).
With post-filtering (100-20,000 Hz), the magnitude
of the artifact was greatly reduced (red).
“. . .we have developed a rabbit model
to study the effects of electrical stimulation
of the IC physiologically and behaviorally.
We have observed a close match between
physiological and behavioral thresholds.”
0.30
Level = 150
0.15
0.00
-0.15
1
2»
Time (ms)
Figure 2. Evoked responses obtained from different contacts
while stimulating Electrode 11. The largest responses were
obtained from the closest contacts. There was asymmetry indicating nonuniform current flow.
—continued on next page—
27
Basic Science, Electrode Development & Medical Applications —continued from previous page—
Aug–Sep ‘06
Dec ‘06–Jan ’07
All
90
Percent Correct
70
50
30
10
0
0
10
20
30
40
50
60
70
80
Current Level (uA)
Figure 3. Psychometric functions for current level for the same
animal’s test results shown in Figures 1 and 2. The behavioral
paradigm was conditioned avoidance (Heffner & Heffner,
1995). The threshold (50% correct in the go/no go task) was
approximately 20 µA. The shape of the function was stable
over a period of several months. Error bars in the combined
curve represent standard errors.
delivered to contact 11. Two psychometric functions measured several months apart were similar
(Figure 3, dotted lines). The threshold (50% correct
in a go/no go task) in the average curve (solid line
and filled circles) was just under 20 µA. Because the
pulse width was different from the physiological
data shown in Figures 1 and 2, the physiological
threshold on contact 10 was measured using a 30 µs/
phase pulse train and was found to be 7.5 µA. Thus,
there was good correspondence between the behavioral and physiological thresholds when the same
stimuli were used.
In summary, we have developed a rabbit model to
study the effects of electrical stimulation of the IC
physiologically and behaviorally. We have observed
a close match between physiological and behavioral
thresholds. Future studies will use more complex
stimuli such as vowel sounds and voice onset time
to address the spectral and temporal resolution of IC
responses to speech signals.
References
Colletti V, Shannon RV. (2005) Open set speech perception with auditory brainstem
implant? Laryngoscope 115:1974-1978.
Heffner HE, Heffner RS. (1995) Conditioned avoidance. In: Klump GM, Dooling
RJ, Fay RR, Stebbins WG, eds. Methods in Comparative Psychoacoustics. Basel:
Birkhauser Verlag.
28
Advanced Bionics® Auditory Research Bulletin 2007
The Effect of Unilateral Multichannel Cochlear
Implantation on Bilateral Tinnitus
Nicola Quaranta, M.D.1
Susana Fernandez-Vega, M.D.1
Chiara D’Elia, M.D.2
Roberto Filipo, Prof., M.D.2
Antonio Quaranta, M.D.1
1 University of Bari, Bari, Italy
2 University of Rome La Sapienza, Rome, Italy
Tinnitus is a common condition that reportedly
affects between 66% and 86% of cochlear implant
candidates before implantation (Quaranta et al,
2004). Subjects affected by bilateral profound sensorineural hearing loss and bilateral tinnitus who have
received one implant represent a study group that
demonstrates the effect of unilateral electrical stimulation on bilaterally perceived tinnitus.
Ninety-eight adult patients have been implanted
at the University of Bari and the University of
Rome La Sapienza. In order to quantify the effect
of cochlear implantation on ipsi- and contralateral
tinnitus, 41 patients who reported bilateral tinnitus
before implantation were evaluated. Of these, 23 had
been implanted with a Nucleus device, 15 with an
Advanced Bionics device, and three with a Med-El
device. Demographics, as well as the cause of deafness, the type of cochlear implant used, and the
speech coding strategy used, were obtained from
clinical reports. All patients were asked to complete
a questionnaire that evaluated the presence, location,
and intensity of tinnitus before and after cochlear
implantation. They also completed the Tinnitus
Handicap Inventory (THI), a 25-item self-report
questionnaire (Newman et al, 1996).
Before cochlear implantation, tinnitus intensity
was graded as mild in 7 cases, moderate in 19 cases,
severe in 12 cases, and troublesome in 3 cases. After
surgery, 7 patients (17%) reported the perception of
“new” tinnitus. With the cochlear implant turned
off, tinnitus was abolished in 23 patients (56.1%)
in the implanted ear and in 22 patients (53.6%) in
the contralateral ear. When the implant was turned
on, tinnitus was abolished in the ipsilateral ear in 27
patients (65.8%) and in the contralateral ear in 27
patients (65.8%). Fourteen patients (34.1%) reported
the absence of tinnitus with the implant either on
or off, while 15 (36.6%) reported the persistence of
tinnitus with the implant either on or off. In 9 cases
(21.9%), tinnitus was present with the implant off
and disappeared with implant on, while in 3 cases
(7.3%), tinnitus was absent with the implant off
and reappeared on both sides when the implant was
turned on.
Statistical analysis did not show significant differences among the three implants. With the device
on, patients using an ACE strategy reported that
tinnitus was abolished in a greater proportion than
patient using a SAS strategy (p = 0.002). Further
analysis showed a trend of a greater suppression
of tinnitus in HiRes users compared to SAS users
(p = 0.07) and ACE users compared to SPEAK
users (p = 0.07).
Questionnaire results showed a significant reduction of total THI score and of the scores for each
subscale (p < 0.001). On average, cochlear implantation induced a significant improvement of
tinnitus-related distress, suppression of tinnitus in
a significant number of patients, and reduction of
tinnitus intensity in the remaining patients.
In summary, these clinical results indicate that
today’s multichannel cochlear implants provide
effective tinnitus suppression. Unilateral electrical
stimulation has beneficial effects on both ipsilateral
and contralateral tinnitus, even when the implant
is switched off. Higher rate stimulation strategies
appear to be more effective in reducing tinnitus than
analogue or lower pulse rate strategies.
References
Newman CW, Jacobson GP, Spitzer JB. (1996) The development of the Tinnitus
Handicap Inventory. Arch Otolaryngol Head Neck Surg 122: 143-8.
Quaranta N, Wagstaff S, Baguley DM. (2004) Tinnitus and cochlear implantation.
Int J Audiol 43:245-251.
29
Basic Science, Electrode Development & Medical Applications Cochlear Implantation in Unilateral Deafness
Associated with Ipsilateral Tinnitus: A Case Study
Andreas Büchner, Ph.D.
Anke Lesinski-Schiedat, M.D.
Rolf-Dieter Battmer, Ph.D.
Yassaman Khajehnouri, M.Sc.*
Thomas Lenarz, Prof., M.D., Ph.D.
Medizinische Hochschule Hannover, Hannover, Germany
* also with Advanced Bionics, Europe
Most treatments for tinnitus, such as retraining or
masking, require acoustic stimulation. In subjects
who are deaf in the ear suffering from tinnitus, these
acoustic approaches are not applicable. On the other
hand, tinnitus suppression utilizing electrical stimulation of the auditory nerve has been reported to be
successful by various research groups (for example,
Battmer et al, 1989). For subjects with near normal
hearing on the non-implanted side, however, the
sound quality produced by a cochlear implant system
may prove unacceptable. To achieve the best conditions for using a cochlear implant in the management of tinnitus, a HiRes 90K device may prove
ideal because (1) the new current steering technique
can provide increased spectral resolution (Koch et
al, 2007), (2) high stimulation and update rates offer
good temporal resolution, and (3) high-rate pulse
trains may suppress tinnitus, even if the recipient
rejects hearing through the implant (Rubinstein et
al, 2003).
CI alone (direct audio input)
80
Percent Correct
60
40
20
0
Initial
1 mo
2 mos
3 mos
4 mos
5 mos
6 mos
Test Interval
Figure 1. Scores over time for the HSM sentence test in quiet
with the CI alone (signal coupled to the processor via direct
audio input). It should be noted that at the one- and two-month
intervals, the subject had not been listening daily in the CIalone condition. Thereafter, he began intensive aural rehabilitation by listening to audio-recorded books.
30
A study is underway at Medizinische Hochschule
Hannover to evaluate cochlear implantation as a
treatment for severe-to-profound unilateral deafness associated with ipsilateral tinnitus. Currently
one subject has been implanted and has reached six
months of experience. The subject is 48 years of age
with a threshold of less than 20 dB HL on the left
side and a sudden hearing loss in 2001 followed by
tinnitus on the right side. He scored 100% on the
Freiburger monosyllabic word test in his hearing ear.
Despite these good results, he decided to participate in the study, hoping for a potential cure for his
tinnitus. The subject was implanted unilaterally with
a HiRes 90K device in September 2006.
For the two days directly after surgery, the subject
continued to suffer from severe tinnitus. During
the recovery period, the tinnitus returned to a
similar level as prior to surgery. At first fitting, the
subject understood speech in quiet immediately but
described the sound quality as hollow. He reported
that his tinnitus was softer and less distracting
immediately following initial stimulation. However,
after returning home, he stated that the tinnitis level
had increased but was not distracting. Since surgery,
the subject has experienced debilitating tinnitus in
only two instances—both times in psychologically
stressful situations. He reports that the electrical
stimulation has only a short-term effect on the
tinnitus, which returns to previous levels in a couple
of minutes after the device is switched off.
Speech perception via the cochlear implant (CI)
alone was measured with the HSM sentence test
in quiet. To ensure that the normal hearing ear did
not receive additional input, the test was performed
via direct input to the speech processor. During the
one-week rehabilitation phase after initial stimulation, the CI side received specific training. The
results obtained at the end of this week were better
than during the following two months when the
subject was not attending to the CI alone. Therefore, he was encouraged to train the implanted ear
regularly by listening to audio books via direct input
to the speech processor. As a result of that training,
the sentence test results improved to about the same
level as during the rehabilitation phase (Figure 1).
Advanced Bionics® Auditory Research Bulletin 2007
Even though no general conclusions can be drawn
from this one case, the results are encouraging overall.
The subject accepts the cochlear implant and uses it
during the working day. For this patient, even though
speech perception using the cochlear implant alone
is significantly poorer than the speech perception of
typical users, the tinnitus suppression is beneficial.
It is likely that the normal-hearing ear is dominant
for speech understanding in everyday situations.
Further intensive training through the cochlear
implant alone may improve the implanted ear’s
speech perception ability.
NH only
NH+CI
NH only
NH+CI
NH only
NH+CI
0
-2
dB SNR
-4
-6
-8
-10
-12
1
0
dB SNR
-1
-2
-3
-4
-5
-6
0
-1
-2
dB SNR
Because the subject had normal hearing in the nonimplanted ear, the study also investigated whether
the cochlear implant could restore spatial hearing
to some degree. The Oldenburger Sentence Test was
administered in noise with separated signal and noise
sources. In all conditions, the signal was presented
from the front while the noise was either presented
from the CI side (90°), from the front (0°), or from
the NH side (-90°). When noise was presented from
the CI side, the implant mainly contributed information about the noise. This test condition usually
reveals any binaural squelch effects. However, this
subject showed no improvement when the CI input
was added to the NH ear (Figure 2A). When noise
was presented from the left, his scores improved—
the improvement attributable to the headshadow
effect (Figure 2B). At some appointments, the
subject also showed a weak trend towards better
understanding when the noise and signal both originated from the front (Figure 2C). Most likely, in
this condition, the NH ear was so dominant that the
additional input provided by the implant was not
significant. The CI did not improve sound localization for this subject. However, his localization ability
using the NH ear alone was already similar to most
bilateral cochlear implant users.
-3
-4
-5
-6
Initial
3 mos
4 mos
6 mos
Test Interval
References
Battmer RD, Heermann R, Laszig R. (1989) Suppression of tinnitus by electric
stimulation in cochlear implant patients [German]. HNO 37(4):148-152.
Koch DB, Downing M, Osberger MJ, Litvak L (2007) Using current steering to
increase spectral resolution in CII and HiRes 90K users. Ear Hear 28(Suppl 2):38S41S.
Rubinstein JT, Tyler RS, Johnson A, Brown CJ. (2003) Electrical suppression of tinnitus
with high-rate pulse trains. Otol Neurotol 24(3):478-485.
Figure 2. Results for the Oldenburger Sentence Test with
separated signal and noise sources. (A) signal presented from
the front, noise from the cochlear implant (CI) side; (B) signal
presented from the front, noise from the normal-hearing (NH)
side; (C) signal and noise presented from the front. (Note that a
lower SNR indicates better performance.)
31
Basic Science, Electrode Development & Medical Applications Using Electrical Stimulation to Suppress Tinnitus
Nearly 36 million Americans suffer from tinnitus—
a ringing or other type of noise that seems to
originate in the ear or head. Tinnitus occurs in the
absence of an external stimulus and often is associated with hearing loss. Numerous treatments have
been explored to alleviate this handicap, which can
significantly affect quality of life. However, tinnitus
often is resistant to treatments such as acoustic
masking, biofeedback, drugs, or surgery.
Fan-Gang Zeng, Ph.D.1
Qing Tang, M.S. 1
Jeff Carroll, M.S.1
Andrew Dimitrijevic, Ph.D.1
Arnold Starr, M.D.1
Leonid Litvak, Ph.D.2
Jannine Larky, M.A.3
Nikolas H. Blevins, M.D.3
1 University of California, Irvine, CA, USA
A few studies have suggested that electrical stimulation may reduce tinnitus in cochlear implant users
(Brackmann, 1981; Rubinstein et al, 2003). In this
case study, a 46-year-old man was implanted with
a HiRes 90K device to control debilitating tinnitus
in his right ear. He had essentially normal hearing
in his left ear so that he could match both tinnitus
and electrical stimulation in the right ear to acoustic
stimulation in the left ear. The subject was able to
match his tinnitus to an acoustic stimulus of 40008000 Hz at 70-90 dB SPL. The effect of electrical
stimulation on his tinnitus was evaluated systematically as a function of pulse rate from 25 to 5000 Hz,
pulse duration from 10 to 500 µsec per phase, electrode position from apex to base, and stimulation
configuration from monopolar to bipolar mode.
2 Advanced Bionics, Valencia, CA, USA
3 Stanford University, Palo Alto, CA, USA
Electric: 100 Hz (1,0)
Tinnitus
10
8
Rebound
6
4
Delay
Suppression
Loudness Estimates
2
0
0
200
400
Electric, 5000 Hz (1,0)
600
800
Tinnitus
10
8
No suppression
6
4
2
Total adaptation
0
0
100
200
300
400
Time (Seconds)
Figure 1. Loudness estimates of tinnitus and electrical stimulation as a function of time—with the turquoise bar above the
x-axis representing the duration of the masker. (A) Low-rate
simulation (100 pps) suppresses tinnitus and does not adapt; (B)
High-rate stimulation (5,000 pps) adapts and does not suppress
tinnitus.
32
Notably, monopolar low-rate pulsatile stimulation
(40-100 pulses per second) with short pulse duration
delivered to the most apical electrode was the most
effective suppressor of the tinnitus (Figure 1). This
result contrasted with previous studies suggesting
that high-rate stimulation could effectively reduce
tinnitus in implant users (Rubinstein et al, 2003).
The suppression could be observed in event-related
evoked potentials, which reflected a waveform difference related to the presence and absence of tinnitus.
Based upon these data, an innovative acoustic waveform employing a Gaussian-enveloped sinusoid and
optimized programming of electric parameters was
created to allow the subject to use his behind-the-
Advanced Bionics® Auditory Research Bulletin 2007
ear processor to suppress tinnitus effectively at home.
This case study underscores the need to customize
electric stimulation for tinnitus suppression and
suggests that complementary electrical stimulation,
rather than masking, is the brain mechanism underlying the unexpected results seen in this subject.
Acknowledgment
This research was supported by the National Institute for Deafness and Other
Communicative Disorders (RO1 DC002267).
References
Brackmann DE. (1981) Reduction of tinnitus in cochlear implant patients. J Larynol
Otol Suppl 4:163-165.
Rubinstein JT, Tyler RS, Johnson A, Brown CJ. (2003) Electrical suppression of
tinnitus with high-rate pulse trains. Otol Neurotol 24(3):478-485.
Zeng FG, Tang Q, Carroll J, Dimitrijevic A, Starr A, Litvak L, Larkey J, Blevins N.
Optimizing electric stimulation to suppress tinnitus. Paper presented at the Association for Research in Otolaryngology, Denver, CO, USA. 10-15 February, 2007.
New Study Init iat i ve
Hearing Performance, Emotional Impact, and Quality of Life Following Reimplantation
Thomas Lenarz, Prof., M.D., Ph.D.1
Yassaman Khajehnouri, Dipl. Inform.2
1 Medizinische Hochschule Hannover, Hannover, Germany
2 Advanced Bionics, Europe
The lifetime and reliability of cochlear implants have
benefit over time. These measures will be applied as soon
improved
in
as a device is suspected to be failing and will continue
the number of device failures over time. However,
significantly,
resulting
in
a
decrease
after reimplantation. All subjects will be evaluated 1, 3, 6,
reimplantation still may be required occasionally because
and 12 months after initial stimulation of the new device.
of device failure, infection, or other medical indications.
Children will be retested at 18 and 24 months after the
When reimplantation is necessary, there may be a
second device is activated. In cases where the contralateral
considerable impact on performance, emotional balance,
ear receives the second implant, comparisons of pitch
and quality of life for implant users. The aim of this study
perception and timbre will be made between ears. Also
is to take a proactive look at reimplantation and to monitor
in those cases, reasons for implanting the contralateral ear
closely both the positive and negative effects it has on the
rather than the same ear will be documented.
individual, the family, and the cochlear implant center.
A prime motivation is to determine the likelihood of
This study will be conducted at Medizinische Hochschule
equivalent benefit from the new implant compared to the
Hannover (MHH) and other European implant centers
benefit from the original device.
yet to be determined. Initial results from 113 reimplanted
patients at MHH indicate that postoperative performance
Five groups of subjects will be followed in the study:
was either equal to or better than the best scores measured
postlinguistically
prelinguistically
with the original implant before failure. The most frequent
deafened adults, teenagers, children 5–12 years of age, and
indication for cochlear reimplantation was device failure.
children younger than five years of age. Clinical test results
Only 14 of 113 reimplanted patients in MHH have had
and a questionnaire survey will be the main outcome
a replacement device because of a medical indication.
measures used. For data analyses, the five groups will
Contralateral implantation was performed in only two of
be treated as one factor with different levels to allow for
the 113 patients.
deafened
adults,
correlation analyses. Therefore, to the extent possible, the
same questionnaire will be used for all subjects regardless
It is anticipated that the information from this study will be
of group.
helpful to clinicians who counsel implant users requiring
reimplantation. The data also should be useful in addressing
Clinical testing or observation as well as the questionnaire
questions about device failure and reimplantation posed by
will monitor changes in speech perception and implant
prospective candidates and their families.
33
Basic Science, Electrode Development & Medical Applications HiResolution Sound Processing
In 2006, Advanced Bionics introduced HiRes Fidelity 120™ (HiRes 120)*, a
programming option to standard HiResolution® Sound (HiRes®) processing.
The goal of HiRes 120 is to build on the strengths of standard HiRes by
improving representation of the stimulus spectrum in the electrical stimulation
pattern. Using active current steering, additional spectral bands are created by
precisely varying the proportion of current delivered simultaneously to adjacent electrodes pairs. When all 16 electrodes are enabled, 120 total spectral
bands are created. Current steering is made possible by 16 independent and
programmable current sources in the CII Bionic Ear® and HiRes 90K™
devices.
Reports from around the world have demonstrated that standard HiRes
surpassed previous-generation technologies in providing improved language
skills, speech perception, and speech intelligibility to implant recipients of all
ages. Moreover, studies have shown that HiRes 120 offers additional benefits for speech understanding in noise and sound quality for some recipients.
Notably, HiRes 120 users report improved music listening and satisfaction—
benefits previously thought to be an unrealistic expectations for many cochlear
implant users. In addition, HiRes 120 offers advantages for speakers of tonal
languages where spectral contrasts convey lexical meaning.
The studies in this section summarize various investigations into the music
listening and sound quality benefits enjoyed by HiRes and HiRes 120 users.
*In the United States, an optional feature for adults only. See package insert for details.
35
HiRes with Fidelity 120:
Speech Recognition and Strategy Preference
Multicenter Study (1) in North America
“The strong preference for HiRes 120
suggests that standard speech test batteries
are not able to reflect some of the
important qualitative benefits
reported by recipients.”
HiRes (baseline)
HiRes 120 @ 3 mos
100
Mean Percent Correct
85.7
HiRes @ 3 mos
91.7 89.4
80
60
60.2
64.1
61.2
60.5
67.9
62.2
40
20
0
CNC Words
HINT in Quiet
HINT +8 dB SNR
Figure 1. Mean scores for three speech tests at baseline with
HiRes, at three months with HiRes 120, and after being refit
with HiRes. Asterisks indicate a significant difference from
baseline (p < .05).
HiRes (baseline)
HiRes 120 @ 3 mos
100
Percent Correct
80
60
40
20
0
Individual Scores
Figure 2. Individual HINT-in-fixed-noise scores at baseline with
HiRes and at three months with HiRes 120 for 25 subjects with
baseline scores < 70%. Scores are rank ordered from lowest
to highest standard HiRes scores. The average improvement
between standard HiRes and HiRes 120 was 13.7%, and the
improvement was significant (p < .05).
36 The flexibility of the CII Bionic Ear and HiRes 90K
electronics platform allows for the continued evolution of HiResolution Sound introduced in 2003.
The HiRes with Fidelity 120 sound processing option
(HiRes 120), released commercially in December
2006, is designed to build on the strengths of standard HiRes by creating electrical stimulation patterns
that more faithfully represent the frequency spectrum of the acoustic input signal. Through improved
spectral analyses and current steering, HiRes 120 is
intended to allow implant users to take advantage of
their residual place-pitch perception capabilities by
delivering spectral information in higher resolution
compared to conventional processing. The increased
spectral resolution, in combination with the fine
temporal resolution already implemented in HiRes,
may lead to better speech perception in noise and
improved music appreciation.
A recent clinical study documented the listening
benefits of HiRes 120 sound processing in 55
adult CII and HiRes 90K users. Performance with
standard HiRes was assessed at the baseline visit
and compared with HiRes 120 performance after
three months of listening experience. Subsequently,
subjects were refit and retested with standard HiRes.
Outcome measures included speech perception in
quiet and noise, music and sound quality ratings,
self-reported benefits, and a preference questionnaire.
The speech perception tests included CNC words,
HINT sentences in quiet, and HINT sentences
in noise (using a fixed signal-to-noise ratio and
an adaptive paradigm). For the 55 subjects, mean
speech perception scores were significantly higher
for HiRes 120 compared to baseline with HiRes
and to scores after subjects were refit with HiRes
(Figure 1). Subjects who showed the lowest scores
at baseline (HINT-in-noise score < 70%) appeared
Advanced Bionics® Auditory Research Bulletin 2007
Participating Centers and Clinics in North America
Carle Clinic Association, Urbana, Illinois
House Ear Clinic, Los Angeles, California
Houston Ear Research Foundation, Houston, Texas
Indiana University, Indianapolis, Indiana
Johns Hopkins University, Baltimore, Maryland
Mayo Clinic, Rochester, Minnesota
Midwest Ear Institute, Kansas City, Missouri
New York University, New York, New York
Northwestern University, Chicago, Illinois
Ottawa Hospital (Civic Campus), Ottawa, Ontario
University of Texas (Southwestern Medical Center), Dallas, Texas
University of Massachusetts, Worcester, Massachusetts
Washington University, St. Louis, Missouri
At the three-month test interval, subjects indicated
their preferred sound processing option (HiRes 120
or standard HiRes) and their strength of preference
on a scale from 1 (weak preference) to 10 (strong
preference). Of the 55 subjects, 47 (85.5%) preferred
HiRes 120. The strength of preference was 7.9
(range: 1-10). The strength of preference was rated as
8 or higher by 28 of the 47 subjects and 18 subjects
rated it as 10.
These data indicate that HiRes 120 is a usable
sound processing option that can provide improved
speech recognition for some Bionic Ear recipients
compared to standard HiRes. Moreover, HiRes 120
was preferred by the majority of study participants.
The strong preference for HiRes 120 suggests that
standard speech test batteries are not able to reflect
some of the important qualitative benefits reported
by recipients.
Better with Standard HiRes
Equivalent
Better with HiRes 120
4
3
2
1
0
dB
to exhibit the greatest gains when using HiRes 120,
particularly for the HINT sentences in quiet and
in fixed noise (Figure 2). The adaptive HINT paradigm proved to be sensitive to differences in sound
processing for high-performing subjects. Sentence
reception thresholds (sSRTs) represent the SNR in
dB at which subjects hear 50% of HINT sentences
in 52 dB SPL speech-spectrum noise. sSRTs were
lower (better) for HiRes 120 than for HiRes for 18
of 30 subjects who scored better than 70% on the
HINT-in-fixed-noise test at baseline (Figure 3). For
5 of 30 subjects, sSRTs were similar between HiRes
and HiRes 120 (within 1 dB). Seven of 30 subjects
showed better results with standard HiRes on this
particular measure.
-1
-2
-3
-4
-5
-6
Individual Difference Scores
Figure 3. HiRes 120 sSRT minus standard HiRes sSRT for 30
subjects who scored 70% or greater on the HINT-in-fixed noise
test at baseline. Scores are rank ordered from smallest to greatest difference in sSRT. A negative value indicates greater benefit
with HiRes 120.
HiResolution Sound Processing 37
HiRes with Fidelity 120: Everyday Listening Benefits
Multicenter Study (2) in North America
HiRes (baseline)
HiRes 120 @ 3 mos
Percent of Respondents
100
80
60
40
20
In
Q
In
Sm uiet
all
Gr
In
ou
La
p
rg
e
Gr
ou
p
At
Ho
m
e
At
W
or
k
In
aV
eh
icl
e
In
Le
ctu
re
At
aP
In
ar
ty
aR
es
ta
ur
an
Lis
t
te
TV nin
/R g t
ad o
In io
aM
ov
i
Te U e
lep sin
Us ho g a
ne
in
ga
Ce
Ph llul
on ar
e
0
Figure 1. Percent of subjects responding “clear” or “extremely
clear” for understanding speech in 13 listening situations.
Asterisk indicates situations in which there was > 15% shift in
the percentage when listening with HiRes 120 compared to
standard HiRes.
HiRes (baseline)
Percent of Respondents
100
HiRes 120 @ 3 mos
80
60
40
20
A recent clinical study documented the listening
benefits of HiRes with Fidelity 120 (HiRes 120)
sound processing in 55 adult CII Bionic Ear and
HiRes 90K users. Performance with standard HiRes
was assessed at the baseline visit and compared
with HiRes 120 performance after three months
of listening experience. Subsequently, subjects were
refit and retested with standard HiRes.
As part of the study, subjects rated the sound clarity
and ease of listening in everyday situations using a
5-point Likert scale, with 5 being best (clearest or
easiest). Ratings were made at baseline with standard HiRes and after three months of HiRes 120
use. After three months of HiRes 120 use, a greater
proportion of the 55 subjects rated their experience
as “clear” or “very clear” for all 13 listening situations.
There was a 10% or greater shift in the proportions
for seven situations, and over a 15% shift in the
proportions for three situations after three months
of HiRes 120 experience (Figure 1).
Similarly, a greater proportion of subjects rated their
experience as “easy” or “very easy” for all situations
but one (at home) after using HiRes 120. There
was a 10% or greater shift in the proportions for
seven situations and greater than a 15% shift in the
proportions for four situations after three months of
HiRes 120 experience (Figure 2).
Pa
rty
Re
sta
ur
an
t
Lis
te
TV nin
/R g t
ad o
io
In
aM
ov
ie
At
a
In
a
Ve
hi
cle
In
Le
ctu
re
At
W
or
k
In
a
In
Qu
iet
In
Sm
all
Gr
ou
In
p
La
rg
e
Gr
ou
p
At
Ho
m
e
0
Figure 2. Percent of subjects responding “easy” or “very easy”
when listening in 11 everyday situations. Asterisk indicates situations in which there was a >15% shift in the percentage when
listening with HiRes 120 compared to standard HiRes.
38 Advanced Bionics® Auditory Research Bulletin 2007
Forty-seven of 55 subjects (85.5%) reported a preference for the HiRes 120 over standard HiRes. Those
47 subjects also indicated those aspects of sound that
were better with HiRes 120 using the scale: strongly
disagree, disagree, neutral, agree, or strongly agree.
As seen in Table 1, over 80% of the subjects agreed
or strongly agreed that HiRes 120 resulted in better
overall sound quality and more natural sounding
speech, and that environmental sounds were easier
to distinguish and sounded more distinct compared
with standard HiRes. Over half of them reported
that it was easier to understand speech in noise with
HiRes 120.
“The reported benefits extend beyond
speech perception and may encompass
qualitative everyday listening benefits—
outcomes not assessed typically
in studies of implant efficacy.”
These subjective test results indicate that HiRes
120 is a viable sound processing option that may
offer additional hearing advantages for some CII
and HiRes 90K recipients in a variety of listening
environments. The reported benfits extend beyond
speech perception and may encompass qualitative
everyday listening benefits—outcomes not assessed
typically in studies of implant efficacy.
Table 1. Distribution of preference ratings on questionnaire items regarding speech
and environmental sounds for subjects who preferred HiRes 120.*
Statement of Perceived Benefit
Strongly
Disagree
Disagree
Neutral
Agree
Strongly
Agree
Agree or
Strongly Agree
2.1%
0%
12.8%
46.8%
38.3%
85.1%
0%
2.1%
17.0%
48.9%
31.9%
80.9%
2.1%
10.6%
29.8%
27.7%
29.8%
57.5%
Environmental sounds are easier to
distinguish.
0%
0%
19.1%
42.6%
38.3%
80.9%
Environmental sounds are more distinct.
0%
2.1%
12.8%
51.1%
34.0%
85.1%
Quality is better.
Speech is more natural.
Speech is easier to understand in noise.
* n = 47
HiResolution Sound Processing 39
HiRes with Fidelity 120:
Sound Quality and Music Listening
Multicenter Study (3) in North America
“The reported benefits include
improved everyday sound quality
and music appreciation—outcomes once
thought to be beyond expectations for
cochlear implant recipients.”
HiRes (baseline)
1.0
Mean Rating
0.8
HiRes 120 @ 3 mos
*
*
*
*
0.6
0.4
0.2
0.0
Men’s
Voices
Women’s
Voices
Children’s
Voices
Environmental
Sounds
Figure 1. Mean ratings of sound clarity with standard HiRes at
baseline and after using HiRes 120 for three months (on a scale
of 0 to 1, with 1 being best) (n = 55). Ratings were significantly
higher for all sound categories with HiRes 120 (* p < .005).
10
Mean Rating
8
HiRes (baseline)
*
6
HiRes 120 @ 3 mos
HiRes @ 3 mos
*
4
2
0
Pleasantness
Distinctness
Figure 2. Mean ratings of music pleasantness and distinctness
(scale: 1 to 10) at baseline with HiRes, after three months with
HiRes 120, and at three months with HiRes (n = 55). Ratings
were significantly higher for HiRes 120 compared to baseline
with HiRes (* p < .005).
40 A recent clinical study documented the listening
benefits of HiRes with Fidelity 120 (HiRes 120)
sound processing in 55 adult CII Bionic Ear and
HiRes 90K users. Performance with standard HiRes
was assessed at the baseline visit and compared
with HiRes 120 performance after three months
of listening experience. Subsequently, subjects were
refit and retested with standard HiRes.
As part of the study, subjects listened to and rated
the clarity of men’s, women’s, and children’s voices
as well as environmental sounds (on a scale of
0 = unclear to 1 = clear) with standard HiRes and
after using HiRes 120 for three months (Iowa
Sound Quality Test, Tyler et al, 2004). The ratings
for clarity of voices and environmental sounds were
significantly higher for HiRes 120 compared to
baseline with HiRes (p < .005) (Figure 1).
Subjects also listened to instrumental music passages
and rated their pleasantness and distinctness on a
scale from 0 (extremely unpleasant/indistinct) to
10 (extremely pleasant/distinct). Ratings of music
pleasantness and distinctness were significantly
higher for HiRes 120 compared to baseline with
HiRes (Figure 2).
Using a questionnaire format, subjects rated various
aspects of listening to music in their everyday lives.
Mean ratings for frequency of listening to music
(1= never to 5 = very often) and satisfaction with
listening to music (1 = very dissatisfied to 5 = very
satisfied) were higher with HiRes 120 than with
standard HiRes (Figure 3). After three months using
HiRes 120, there was an 11% shift in the proportion
of adults reporting that they listen to music “often”
or “very often” compared to their baseline ratings
with standard HiRes. There was a 24% shift in the
proportion of subjects stating that they are “satisfied” or “very satisfied” when listening to music using
HiRes 120 compared to standard HiRes.
Forty seven of 55 subjects (85.5%) reported a preference for HiRes 120 over standard HiRes. Those
47 subjects also indicated those aspects of music
listening that were better with HiRes 120 using
Advanced Bionics® Auditory Research Bulletin 2007
the scale: strongly disagree, disagree, neutral, agree,
or strongly agree. Table 1 shows that 55-72% of the
subjects who preferred HiRes 120 reported improvement in listening to various aspects of music (agree
or strongly agree), with the exception of recognizing
lyrics.
Reference
HiRes 120 @ 3 mos
4
Mean Rating
These qualitative clinical results indicate that HiRes
120 is a viable sound processing option that may
offer additional hearing advantages for some CII and
HiRes 90K recipients in a variety of listening environments. The reported benefits include improved
everyday sound quality and music appreciation—
outcomes once thought to be beyond expectations
for cochlear implant recipients.
HiRes (baseline)
5
3
2
1
0
Frequency
Satisfaction
Figure 3. Mean ratings for frequency of listening to music
(1 = never to 5 = very often) and satisfaction with listening to
music (1 = very dissatisfied to 5 = very satisfied) at baseline
with HiRes and after three months with HiRes 120 (n = 55).
Ratings were higher with HiRes 120 than with standard HiRes.
Tyler R, Witt S, Dunn CC. (2004) Trade-offs between better hearing and better
cosmetics. Am J Audiol 13:193-199.
Table 1. Distribution of preference ratings on questionnaire items
regarding music for subjects who preferred HiRes 120.*
Statement of Perceived Benefit
Strongly
Disagree
Disagree
Neutral
Agree
Strongly
Agree
Agree or
Strongly Agree
Music sounds better.
0%
4.3%
23.4%
38.3%
34.0%
72.3%
Music sounds more natural.
0%
0%
31.9%
31.9%
34.0%
66.0%
Music sounds richer.
0%
2.1%
29.8%
34.0%
34.0%
68.1%
Melody is more enjoyable.
0%
4.3%
27.7%
40.4%
25.5%
66.0%
Rhythm is more noticeable.
2.1%
2.1%
31.9%
34.0%
27.7%
61.7%
Singers’ voices are distinguishable from
instruments.
0%
2.1%
27.7%
44.7%
23.4%
68.1%
Singers’ voices sound more distinct.
0%
4.3%
38.3%
31.9%
23.4%
55.3%
Lyrics are recognizable.
0%
10.6%
38.3%
27.7%
21.3%
48.9%
Individual instruments are distinguishable.
0%
8.5%
31.9%
46.8%
12.8%
59.6%
Individual instruments sound more distinct.
0%
6.4%
23.4%
53.2%
17.0%
70.2%
* n = 47
HiResolution Sound Processing 41
HiRes 120 Benefits in Adults:
Multicenter Study in Europe
“These early results suggest that . . .
CII and HiRes 90K implant recipients
experience real benefits in everyday life
when using HiRes 120
on the Harmony processor.”
Percent Difference Score
40
30
20
10
0
-10
-20
-30
Individual Scores
Figure 1. Individual speech-in-noise results for 31 subjects
expressed as a difference between the three-month score with
HiRes 120 and the baseline score with standard HiRes. A positive score indicates improved performance with HiRes 120.
A negative score indicates better performance with HiRes.
42 A European multicenter study currently is underway
to evaluate the benefit of HiRes with Fidelity 120
(HiRes 120) sound processing. To date, over 90
subjects have been enrolled at 14 sites. There are two
study populations. One group uses HiRes 120 on
their body-worn PSP processors. The other group
includes individuals who used previous generation
processors (various models) before being fit with
HiRes 120 on the new Harmony™ processor. In
addition to conventional tests of speech perception,
questionnaires are used to evaluate speech quality,
music appreciation, and processor satisfaction.
Results to date show that HiRes 120 provides
improved benefit for understanding speech in tests
that simulate everyday listening in noise. Figure 1
shows speech-in-noise data for 31 adult implant
users with at least nine months of implant experience prior to enrolling in the study. Because the
data originate from adults in several countries, the
speech scores include results from the German
HSM Sentence test (+10 dB SNR), the Italian PPV
sentence test (+5 dB SNR), the Swedish sentence
test (+5 dB SNR), or the Dutch CVC test (+10
dB SNR). Figure 1 displays the percent difference between scores obtained after three months
of HiRes 120 use (15 subjects using Harmony,
16 subjects using a PSP) and at baseline with standard HiRes sound processing implemented on the
subject’s initial processor. Positive difference scores
indicate an improvement with HiRes 120 and negative difference scores indicate better performance
with standard HiRes. A highly significant number
of subjects (23 of 31, 74%) showed improved performance with HiRes 120 (sign test, p < 0.01).
Advanced Bionics® Auditory Research Bulletin 2007
Participating Centers and Clinics in Europe
Birmingham Adult Cochlear Implant Programme, Birmingham, United Kingdom
Emmeline Cochlear Implant Programme, Cambridge, United Kingdom
Hôpital Avicenne, Bobigny, France
Hôpital Beaujon, Clichy, France
Hôpital Edouard Herriot, Lyon, France
Hôpital Saint Antoine, Paris, France
Karolinska University Hospital Huddinge, Stockholm, Sweden
Klinikum d J W Goethe Universität, Frankfurt, Germany
Klinikum d. Albert-Ludwigs-Universität, Freiburg, Germany
Leiden University Medical Centre, Leiden, The Netherlands
Medizinische Hochschule Hannover, Hannover, Germany
Royal National Throat, Nose, & Ear Hospital, London, United Kingdom
Universitätsklinikum Schleswig Holstein, Kiel, Germany
University of Rome La Sapienza, Rome, Italy
HiRes 120 was designed to improve spectral resolution and pitch perception. When implemented
on the Harmony, with its high-fidelity front-end
signal processing, listeners are expected to experience improved hearing in noise and sound quality.
These early results suggest that these design goals
have been met and that CII and HiRes 90K implant
recipients experience real benefits in everyday life
when using HiRes 120 on the Harmony processor.
Standard HiRes
10
HiRes 120
9
8
Mean Rating
Subjects also completed a series of Speech Quality
Questionnaires (Figure 2). Results from 28 subjects
show significant improvements with HiRes 120
for sound clarity, pleasantness, and overall speech
quality (p < 0.05). All areas were rated at least 7 out
of 10 with HiRes 120. In addition, subjective ratings
for listening in background noise with HiRes 120
showed a highly significant improvement (p < 0.01).
These positive speech perception and subjective
evaluation results are reflected in the strategy preference in that 80% of subjects prefer HiRes 120 to
standard HiRes.
7
*
*
*
6
**
5
4
3
2
1
n=13
0
Clarity
Pleasantness
Quality
Reduced Noise
Interference
Figure 2. Mean subjective ratings of sound clarity, pleasantness, quality, and reduced noise interference for HiRes 120 and
standard HiRes (n = 28). The differences in clarity, pleasantness,
and quality are significant (* = p < .05). The difference for listening in noise is highly significant (** = p < .01).
HiResolution Sound Processing 43
Sound Quality and Music Perception with HiRes 120:
Multicenter Study in Europe
Deborah Vickers, Ph.D.
University College London, London, United Kingdom
“. . .preliminary results show a trend
toward better everyday sound quality
and music perception with HiRes 120
compared to standard HiRes.”
Standard HiRes
10
HiRes 120
9
Mean Rating
8
7
*
6
5
4
3
2
1
In
y
Le
ss
Ti
nn
y
om
Bo
ss
al
Qu
nt
sa
ea
Pl
Le
ity
ss
ne
ity
ar
Cl
Re
du
ce
te No d
rfe ise
re
nc
e
0
Figure 1. Mean ratings for six dimensions of everyday listening
with HiRes at baseline and after three months of HiRes 120
experience (n = 19). Note: * = p < .05.
Standard HiRes
10
HiRes 120
9
Mean Rating
8
7
6
5
4
3
2
1
s
es
ra
ln
tu
Na
ss
Ti
nn
y
y
Le
Qu
Bo
om
ss
Le
e
Pl
ss
ne
nt
a
as
al
ity
0
Figure 2. Mean ratings of five aspects of music with HiRes at
baseline and after three months of HiRes 120 experience
(n = 19).
44 HiRes with Fidelity 120 (HiRes 120) sound
processing uses current steering technology to
enhance the spectral representation of sound. The
improved representation of frequency information is expected to improve sound quality and the
perception of music. As part of a multicenter study
of HiRes 120 in Europe, these abilities were assessed
using the Quality Assessment Questionnaire (QAQ)
developed by Vickers and colleagues (Advanced
Bionics, 2005). Using the QAQ, subjects rate their
everyday listening experience as well as music
samples along various dimensions (e.g., quality,
clarity) on a scale from 1 (poorest) to 10 (best). The
QAQ provides a quick and efficient method that is
sufficiently sensitive for highlighting differences in
perception resulting from different sound processing
strategies. The questionnaire was developed and
validated during a clinical study of standard HiRes
sound processing in which it was shown that HiRes
improved music quality and timbre perception
compared to conventional sound processing.
The QAQ is being used in two study protocols
aimed at evaluating HiRes 120. The first protocol
evaluates the difference between standard HiRes and
HiRes 120 in listeners with longer than six months
of experience with HiRes. Subjects are switched to
HiRes 120 and subsequently evaluated after one,
three, and six months of HiRes 120 experience. The
second study uses a crossover protocol where newly
implanted subjects are programmed initially with
either HiRes or HiRes 120 and then switched after
three months to the alternate strategy. They are evaluated at three and six months after their implants
are activated.
Preliminary data are available for 19 subjects
enrolled in the first protocol who have reached
the three-month test interval. Subjects rated
various sound dimensions of everyday listening
Advanced Bionics® Auditory Research Bulletin 2007
Participating Centers and Clinics in Europe
Birmingham Adult Cochlear Implant Programme, Birmingham, United Kingdom
Karolinska University Hospital Huddinge, Stockholm, Sweden
Klinikum d. Albert-Ludwigs-Universität, Freiburg, Germany
Klinikum der J. W. Goethe Universität, Frankfurt, Germany
Medizinische Hochschule Hannover, Hannover, Germany
University College London, London, United Kingdom
University of Rome La Sapienza, Rome, Italy
8
*
7
6
5
4
3
2
1
In
Pl
y
ss
Ti
nn
y
om
Bo
Le
ss
Le
al
Qu
nt
sa
ea
Re
du
ce
te No d
rfe ise
re
nc
e
ss
ne
ity
ar
Cl
ity
0
Figure 3. Mean ratings for six dimensions of everyday listening
after three months with HiRes and three months with HiRes
120 (n = 10). Note: * = p < .05.
Standard HiRes
10
HiRes 120
9
8
*
7
6
5
4
3
2
1
s
Na
tu
ra
ln
es
ny
in
ss
T
Le
oo
m
ss
B
Le
P
ss
ne
nt
sa
a
le
al
ity
ity
ar
Cl
y
0
Qu
In summary, preliminary results show a trend toward
better everyday sound quality and music perception with HiRes 120 compared to standard HiRes.
This ongoing study will be exploring how these
percepts change as listeners gain more experience
with the increased spectral information provided by
HiRes 120.
HiRes 120
9
Mean Rating
At this point, there are only 10 subjects enrolled in
the second protocol who have reached the six-month
test interval (three months with their first strategy
and then three months experience with their second
strategy). Six were programmed first with HiRes 120
and four were programmed first in standard HiRes,
which may bias the HiRes results because of potential learning effects. In everyday sound experience,
HiRes 120 was rated significantly higher for clarity
of speech (p = 0.039) (Figure 3). On the music questionnaire, HiRes 120 was rated significantly better
for tinniness (p = 0.025), thereby suggesting that
high frequency sound quality had been improved
with HiRes 120 (Figure 4). For the music ratings,
there also were trends for higher ratings of naturalness, quality, and boominess with HiRes 120, but
additional data are required to determine if these
differences are significant.
Standard HiRes
10
Mean Rating
(Figure 1), as well as attributes of dynamic music
samples (Figure 2) with standard HiRes and after
three months of HiRes 120 use. In general, HiRes
and HiRes 120 were rated similarly for most questionnaire items. However, HiRes 120 showed
significant improvements in the reduction of noise
interference compared to HiRes for these experienced HiRes users (p = 0.017) after three months.
Figure 4. Mean ratings of six aspects of music after three
months with HiRes and three months with HiRes 120 (n = 10).
Note: * = p < .05.
Reference
Advanced Bionics (2005) European HiResolution multicenter study: sound quality
results. In Auditory Research Bulletin 2005, Valencia, CA: Advanced Bionics
Corporation, 92-93.
HiResolution Sound Processing 45
New Study Init iat i ve
Speech Identification and Speech Masking Release with HiRes 120 Sound Processing:
Multicenter Clinical Study in Europe
Christian Lorenzi, Prof., Ph.D.1; Olivier Sterkers, Prof., M.D.2; Bruno Frachet, Prof., M.D.3;
Bernard Meyer, Prof., M.D.4; Eric Truy, Prof., M.D.5, 6; Julie Bestel, Ph.D.7
1 Ecole Normale Supérieure, Paris, France (Research Group [GRAEC] CNRS 2967); 2 Hôpital Beaujon, Clichy, France; 3 Hôpital Avicenne,
Bobigny, France; 4 Hôpital St Antoine, Paris, France; 5 Hôpital Edouard Herriot, Lyon, France; 6 Université Claude Bernard, Lyon, France;
7 Advanced Bionics, Europe
Many studies have demonstrated that for normal-hearing
processing, then theoretically listeners should experience
listeners speech identification performance and speech
(1) improved consonant identification in quiet, specifically
reception
improved
for place of articulation (a spectrally-based phonetic feature),
substantially when maskers have slow spectral and temporal
(2) improved vowel identification in quiet (because of
fluctuations compared to when the maskers do not fluctuate.
better formant frequency representation), and (3) improved
Several mechanisms may be important in this masking
masking release for at least the slowest spectral modulations
release (MR) effect. For example, dip listening (an ability to
in a noise masker.
thresholds
(SRTs)
in
noise
are
take advantage of relatively short or narrow temporal and
spectral minima in the fluctuating background to detect
The goal of this new project is to test these three predictions
speech cues) and auditory grouping mechanisms (which use
in newly implanted adults who alternately use standard
disparities in spectral and fine-structure information between
HiRes and HiRes 120 sound processing on the Harmony
speech and background) seem to play a major role. (For a
processor. The protocol is a crossover design where half of
review, see Füllgrabe et al, 2006.) However, this masking
the listeners will be fit with standard HiRes for three months
release effect is severely reduced in listeners with cochlear
and then HiRes 120 for three months (Group A). The other
damage and completely absent in deaf listeners who use
half will use HiRes 120 for the first three months and then
cochlear implants (for example, Bacon et al, 1998; Eisenberg
standard HiRes for the second three months (Group B).
et al, 1995; Festen & Plomp, 1990; Gilbert & Lorenzi, 2006;
The first six months after initial stimulation serve as an
Lorenzi et al, 2006; Nelson et al, 2003; Peters et al, 1998;
“acclimatization” period. At the end of six months, subjects
Summers & Molis, 2004).
will experience a four-month post-acclimatization period
(when speech intelligibility in quiet is stabilized). During this
Current sound processors perform a limited spectral
post-acclimatization period, Group A will be fit with HiRes
decomposition of incoming sounds. Across the various types
for two months and then with HiRes 120 for the final two
of processors, at best 16 to 24 adjacent, non-overlapping
months. Group B will be fit with the two sound processing
frequency bands spanning the 50-8000 Hz range critical
strategies in the opposite order.
for speech perception are used to encode and represent the
spectral structure of speech stimuli and background maskers.
Consonant identification and phonetic feature reception will
Consequently, sound processors provide poor frequency
be measured in quiet and in the presence of a stationary
resolution compared to normal auditory processes and offer
unmodulated or a spectrally modulated noise masker. In the
extremely limited access to the fine spectral information in
masked conditions, a gated noise masker will be added to
both the speech and masker signals. Processing limitations
each utterance (and refreshed in each trial of a given session).
smooth (or fill in) the spectral gaps within the background
The noise masker is a speech-shaped noise encoded using 32
maskers, thereby preventing implant users from listening in
non-overlapping auditory (gammatone) filters (one equivalent
the spectral valleys.
rectangular bandwidth wide) logarithmically spaced and
centered between 80 and 8,020 Hz. This speech-shaped
The new HiRes with Fidelity 120 (HiRes 120) coding
noise will be either unmodulated or spectrally modulated to
scheme developed by Advanced Bionics aims to increase
produce periodic “spectral gaps” in the noise masker.
the transmission of fine spectral information using a
strategy based upon current steering. If HiRes 120 improves
In each experimental condition, the noise masker will be
spectral resolution compared to standard HiRes sound
added to each speech utterance at a fixed signal-to-noise
—continued on next page—
46 Advanced Bionics® Auditory Research Bulletin 2007
New Study Init iat i ve
Pitch Perception and Speech Recognition with HiRes and HiRes 120
in Adult Cochlear Implant Users
Gail Donaldson, Ph.D
Patricia Dawson, B.A.
Lamar Borden, B.A.
University of South Florida, Tampa, FL, USA
In this planned study, experienced users will be fit with
their current strategy (either HiRes or HiRes 120) on a
Speech recognition depends on the listener’s ability to
Harmony processor at the baseline visit. Subjects will be
discriminate the spectral information embedded in the
tested with this strategy after two weeks of Harmony use and
speech signal. The purpose of this study is to determine
subsequently fit with the opposite strategy (either HiRes or
whether cochlear implant users demonstrate improved
HiRes 120). Two months later, subjects will be assessed with
perception of spectral speech cues when they are using
the second strategy and then refit with their original strategy.
HiRes with Fidelity 120 (HiRes 120), a speech processing
Final testing will occur two weeks thereafter. Tests will
strategy designed to provide enhanced spectral resolution.
include a psychophysical test of spectral resolution (place-
The hypothesis is that implant users with good place-
pitch perception), vowel and consonant recognition, and
pitch discrimination will achieve increased perception of
sentences in noise (BKB-SIN). A comparison of two-week and
spectral cues with the HiRes 120 strategy as compared to
final results with the original strategy will reveal any learning
HiResolution (HiRes) sound processing—an equivalent
effects that occur. Comparisons of HiRes and HiRes 120
strategy that provides standard spectral resolution. In
speech results with spectral resolution data will determine
contrast, implant users with poor place-pitch capability
whether listeners with good place-pitch discrimination are
may have equivalent perception of spectral cues with the
able to take advantage of the enhanced spectral resolution
enhanced and standard strategies.
provided by HiRes 120.
—continued from previous page—
ratio (SNR). The SNR will be determined individually for
References
each subject in a preliminary experiment so as to yield a
Bacon SP, Opie JM, Montoya DY. (1998) The effects of hearing loss and noise
masking on the masking release for speech in temporally complex background.
J Speech Lang Hear Res 41:549-563.
mean consonant identification performance between 30
and 40% across four repeated sessions in which the speechshaped noise is unmodulated. Speech identification will be
assessed using a psychoacoustic research system (GRAECIntelliTest) placed in a sound-treated chamber. All stimuli will
be delivered in the free field via two Bose loudspeakers.
Masking release will be defined as the change in consonant
identification performance in modulated versus unmodulated
noise. Masking release also should be reflected in differences
between the reception of voicing, manner, and place cues in
modulated versus unmodulated noise. The study is scheduled
to begin in early autumn 2007.
Eisenberg LS, Dirks DD, Bell TS. (1995) Speech recognition in amplitudemodulated noise of listeners with normal and listeners with impaired hearing.
J Speech Lang Hear Res 38:222-233.
Festen J, Plomp R. (1990) Effects of fluctuating noise and interfering speech on
the speech reception threshold for impaired and normal hearing. J Acoust Soc
Am 88:1725-1736.
Füllgrabe C, Berthommier F, Lorenzi C. (2006) Masking release for consonant
features in temporally fluctuating background noise. Hear Res 211:74-84.
Gilbert G, Lorenzi C. (2006) The ability of listeners to use recovered envelope
cues from speech fine structure. J Acoust Soc Am 119:2438-2444.
Lorenzi C, Husson M, Ardoint M, Debruille X. (2006) Speech masking release in
listeners with flat hearing loss: Effects of masker fluctuation rate on identification
scores and phonetic feature reception. Int J Audiol 45:487-495.
Nelson PB, Jin S-H, Carney AE, Nelson DA. (2003) Understanding speech in
modulated interference: cochlear implant users and normal-hearing listeners.
J Acoust Soc Am 113:961-968.
Peters RW, Moore BCJ, Baer T. (1998) Speech reception thresholds in noise
with and without spectral and temporal dips for hearing-impaired and normally
hearing people. J Acoust Soc Am 103:577-587.
Summers V, Molis MR. (2004) Speech recognition in fluctuating and continuous
maskers: effects of hearing loss and presentation level. J Speech Lang Hear Res
47:245-256.
HiResolution Sound Processing 47
HiRes with Fidelity 120 Benefit in
Native Speakers of Korean
Hong-Joon Park, M.D.1
Seung-Chul Lee, M.D.1
Young-Myoung Chun, M.D.1
Jee-Yeon Lee, M.S.2
1Soree Ear Clinic, Seoul, Korea
2Hallyn University, Chuncheon, Korea
“. . .these results show that
Korean speakers exhibit improved
speech understanding in quiet and
in noise with HiRes 120.”
100
HiRes (baseline)
*
80
Percent Correct
HiRes 120 (1 mo)
*
*
HiRes 120 (3 mos)
*
HiRes (3.5 mos)
*
*
*
60
40
20
0
Monosyllabic
Words
Bisyllabic
Words
HINT in
Quiet
HINT in
Noise
Figure 1. Mean scores for four speech perception tests with
HiRes (baseline), after using HiRes 120 for one and three
months, and after being refit with HiRes for two weeks. Asterisk
indicates significant difference between HiRes 120 and baseline HiRes (p < 0.05).
48 The purpose of this study was to examine performance with HiRes with Fidelity 120 (HiRes 120)
in native speakers of Korean. Eleven adults with
postlinguistic onset of severe-to-profound hearing
loss participated in the study. Mean age at implant
was 41 years (range: 17-70 years). All subjects used
standard HiRes sound processing from the time of
device activation. Implant experience ranged from
approximately 6 to 18 months.
A within-subjects design was used to compare
performance with standard HiRes and HiRes 120.
Baseline performance was assessed with standard
HiRes. Subjects then were fitted with HiRes 120 and
returned for reevaluation after one and three months
of use. After three months of HiRes 120 experience, subjects were refitted with standard HiRes and
returned two weeks later for reassessment. Speech
perception was evaluated with monosyllabic words,
bisyllabic words, and Korean HINT sentences in
quiet and in noise (+8 dB signal-to-noise ratio).
Speech stimuli were presented at 60 dB SPL.
Figure 1 summarizes the scores for the four speech
perception tests. The same pattern of results was
evident for all tests in that performance was higher
with HiRes 120 at one and three months compared
with standard HiRes at baseline and at the final test
session. The difference between baseline HiRes and
three-month HiRes 120 performance was significant for all four speech tests (p < 0.05).
Advanced Bionics® Auditory Research Bulletin 2007
Percent Correct
60
40
20
0
S9
S10
S4
S11
S2
S5
S8
S3
S6
S1
S7
Figure 2. Individual baseline HiRes and three-month HiRes
120 scores for monosyllabic words. Score pairs are rank
ordered by baseline HiRes results.
HiRes (baseline)
100
HiRes 120 (3 mos)
80
60
40
20
0
S2
CNT
In summary, these results show that Korean speakers
exhibit improved speech understanding in quiet
and in noise with HiRes 120. All subjects preferred
the new HiRes 120 sound processing option. These
results, as well as the subjective comments, are
consistent with data reported for English speakers
(Advanced Bionics, 2007).
HiRes 120 (3 mos)
80
CNT
All subjects preferred HiRes 120 to standard HiRes.
The mean strength of preference was 6.1 (range:
3-10) on a scale from 1 (weak preference) to 10
(strong preference). Subjects also indicated various
aspects of sound that were better with HiRes 120
using a five-point Likert scale (strongly disagree,
disagree, neutral, agree, strongly agree). Over 80%
of subjects agreed or strongly agreed that music
sounded better with HiRes 120, and 55-64% agreed
or strongly agreed that the overall quality of sound
was better, that instruments were more distinguishable, and that singers’ voices were more distinct with
HiRes 120.
HiRes (baseline)
100
Percent Correct
Figures 2 and 3 show individual scores for the monosyllabic word test and the HINT-in-noise test. For
monosyllabic words, eight subjects obtained higher
scores with HiRes 120, one subject demonstrated
equivalent performance, and two subjects obtained
higher scores with standard HiRes (Figure 2). For
the HINT in noise, ten subjects obtained higher
scores with HiRes 120 and one subject demonstrated
equivalent performance with HiRes and HiRes 120
(Figure 3).
S9
S7
S4
S5
S3
S1
S10
S8
S11
S6
Figure 3. Individual baseline HiRes and three-month HiRes
120 scores for HINT sentences in noise. Score pairs are rank
ordered by baseline HiRes results.
Reference
Advanced Bionics. (2007) HiRes with Fidelity 120 Clinical Results. Valencia, CA:
Advanced Bionics Corporation.
HiResolution Sound Processing 49
New
New Study
Study Init
Init iat
iat ii ves
ves
Using Current Steering to Increase Spectral
Resolution in Cantonese-Speaking
Adult Cochlear Implant Users
Betty P. K. Luk, M.A.
Terence K. C. Wong, M.Sc.
Michael C. F. Tong, Prof., M.D.
The Chinese University of Hong Kong,
Hong Kong Special Administrative District,
People’s Republic of China
This ongoing study is aimed at determining whether
Tone Perception in Cantonese-Speaking
Subjects Using Harmony and HiRes 120
Kathy Y. S. Lee, Ph.D.
Betty P. K. Luk, M.A.
Terence K. C. Wong, M.Sc.
Michael C. F. Tong, Prof., M.D.
Charles Andrew van Hasselt, M.D.
The Chinese University of Hong Kong,
Hong Kong Special Administrative District,
People’s Republic of China
postlinguistically deafened cochlear implant users can hear
additional spectral channels—or pitches—using current
Studies have shown that cochlear implant users require a
steering. The number of spectral bands is defined as the
greater number of spectral channels to understand speech
number of distinct pitches that can be heard as current is
in noise and for lexical tone perception. Results from North
delivered to targeted locations along the cochlea. Through
America and Europe indicate that adults and children derive
simultaneous delivery of current to pairs of adjacent
substantial benefit from HiRes with Fidelity 120 (HiRes 120)
electrodes, stimulation can be “steered” to sites between the
sound processing, which provides more spectral information
contacts by varying the proportion of current delivered to each
than previous sound processing strategies. However, there
electrode of the pair. Ten Cantonese-speaking adult HiRes
has been limited investigation of standard HiRes and HiRes
90K implant users will be evaluated with their standard HiRes
120 processing in subjects who are native speakers of a tonal
programs using the Cantonese Word Recognition Test and the
language. The goal of this in-progress study is to examine
Cantonese Hearing-In-Noise Test. Then, spectral resolution
tonal language perception in adults with cochlear implants
(number of spectral bands) will be determined using a
who are native speakers of Cantonese.
psychophysical procedure implemented on a research
software platform. Specifically, subjects are required to
Ten adult HiRes 90K device users are participating in the
loudness balance stimulation on selected pairs of electrodes
study. Baseline data are obtained with standard HiRes and
along the implanted array and to indicate the smallest pitch
compared with HiRes 120 after three months of use. Subjects
difference that can be heard when stimulation is delivered
are evaluated on a tone identification test that includes all
simultaneously to each electrode pair. The number of spectral
15 tone contrasts in Cantonese. Everyday listening benefits
bands will be compared to speech perception ability, program
are assessed with a questionnaire. Early results indicate that
parameters, and demographic information to determine how
some subjects show improved tone perception with HiRes
spectral resolution contributes to implant benefit and to
120 compared with standard HiRes.
provide information for enhancing sound processing.
Reference
Reference
Luk BPK, Wong TKC, Tong MCF. Increasing spectral bands using current steering
in cochlear implant users. Paper presented at the 6th Asia Pacific Symposium
on Cochlear Implants and Related Sciences, Sydney, Australia. 30 October-4
November, 2007.
Lee KYS, Luk PBK, Wong TKC, Tong MCF, van Hasselt CA. Tone perception
results with Harmony and HiRes 120 in Cantonese-speaking subjects. Poster
presented at the 6th Asia Pacific Symposium on Cochlear Implants and Related
Sciences, Sydney, Australia. 30 October-4 November, 2007.
50 Advanced Bionics® Auditory Research Bulletin 2007
New Study Init iat i ves
Benefits of HiRes 120 in Mandarin-Speaking
Pediatric Cochlear Implant Users
Benefits of HiRes 120 in Cantonese-Speaking
Adult Cochlear Implant Users
Yu-Tuan Chang, M.D.
Hui-Mei Yang, M.S.
Jiunn-Liang Wu, M.D.
National Cheng Kung University Hospital, Tainan, Taiwan
Betty P. K. Luk, M.A.
Terence K. C. Wong, M.Sc.
Kathy Y. S. Lee, Ph.D.
Michael C. F. Tong, Prof., M.D.
In HiRes with Fidelity 120 (HiRes 120), the input signal is
The Chinese University of Hong Kong, Hong Kong Special Administrative District, People’s Republic of China
analyzed in greater spectral detail than with standard HiRes
sound processing to deliver a maximum of 120 spectral bands
With earlier technologies, the number of channels of
to the cochlea. The increased spectral resolution offered
stimulation provided by a cochlear implant has been
by HiRes 120 is expected to provide more benefit than
limited by the number of contacts in the electrode array.
standard HiRes, especially for speakers of tonal languages.
Presently, through simultaneous delivery of current to pairs
The objective of this study is to compare performance with
of adjacent electrodes, stimulation can be “steered” to sites
HiRes 120 to performance with standard HiRes in Mandarin-
between electrode contacts by varying the proportion of
speaking children. Nine children, ages 3-12 years, who
current delivered to each electrode of the pair. Clinical data
have been implanted with CII or HiRes 90K implants are
indicate that some cochlear implant users can differentiate
enrolled in the study. First, subjects are evaluated with HiRes.
multiple and unique pitches when current is “steered”
Subsequently, they are fit with HiRes 120 and evaluated
between adjacent electrode contacts. HiRes with Fidelity
immediately, then after one and three months of HiRes 120
120 (HiRes 120) uses 15 analysis channels and can deliver
use. Performance measures include a consonant recognition
stimulation to as many as 120 sites along the cochlea. The
test, a tone discrimination test, the Peabody Picture
objective of this study is to compare performance with
Vocabulary Test, the Lexical Neighborhood Test, and the
HiRes 120 implemented on the Harmony sound processor to
Speech Perception in Noise (SPIN) test. A strategy preference
performance with standard HiRes. Ten Cantonese-speaking
questionnaire will be completed by each child (or parents/
subjects with postlinguistic onset of severe-to-profound
caretakers) at the end of the study.
hearing loss who have been implanted with a CII or HiRes
90K cochlear implant are participating in this study. Initially,
Reference
Chang Y-T, Yang J-M, Wu J-L. Speech perception benefit of HiRes with Fidelity
120 in Mandarin-speaking children with cochlear implants. Paper presented
at the 6th Asia Pacific Symposium on Cochlear Implants and Related Sciences,
Sydney, Australia. 30 October-4 November, 2007.
subjects are evaluated with HiRes on the Auria processor.
Next, they are fit with a HiRes program on the Harmony
and evaluated after one week of use. Subsequently, each
subject is fit with HiRes 120 for three months and evaluated
thereafter. Subjects are tested using the Cantonese Word
Recognition Test and the Cantonese Hearing-In-Noise Test.
Finally, a strategy-preference questionnaire is completed by
each subject at the end of the study. Preliminary data suggest
variability in preference regarding the two sound processing
strategies under study.
Reference
Luk BPK, Wong TKC, Lee KYS, Tong MCF. Clinical trial of HiRes 120 on
Cantonese-speaking cochlear implant users. Presented at the 6th Asia
Pacific Symposium on Cochlear Implants and Related Sciences, Sydney,
Australia. 30 October-4 November, 2007.
HiResolution Sound Processing 51
New Study Init iat i ves
Evaluation of HiRes with Fidelity 120 in Adults with No Prior Cochlear Implant Experience:
Participating Centers and Clinics in North America
Arkansas Children’s Hospital, Little Rock, Arkansas
Northwestern University, Chicago, Illinois
Oregon Health Sciences University, Portland, Oregon
Ottawa Hospital-Civic Campus, Ottawa, Ontario
Stanford University, Palo Alto, California
Sunnybrook & Women’s College Health Sciences Centre, Toronto, Ontario
To date, results with HiRes with Fidelity 120 (HiRes 120)
Unlike fixed-level tests, the BKB-SIN is designed to avoid
processing have been obtained in previously implanted adults
floor and ceiling effects and is quick to administer. The BKB-
who had an average of two years of device use at the time
SIN data will indicate how reception thresholds for sentences
they were converted from previous strategies. Most of these
change over time, thereby indicating how the ability to hear
individuals experienced enhanced speech perception, sound
speech in noise evolves in newly implanted individuals.
quality, and music appreciation with the software upgrade
Speech perception data will be obtained at a baseline visit
(Advanced Bionics, 2007; Firszt et al, submitted). Thus, there
(prior to implantation) and then tracked one, three, and six
has been no investigation of the benefits experienced over
months following device activation.
time for newly implanted adults who use HiRes 120 from
the time of device activation. The purpose of this study is to
References
evaluate speech recognition in quiet and noise during the first
Advanced Bionics. (2007) HiRes with Fidelity 120 Clinical Results. Valencia, CA:
Advanced Bionics.
six months of device use in adults who receive HiRes 120 as
their initial sound processing strategy. Speech perception will
be assessed with CNC words, HINT sentences in quiet and
in fixed noise, and the BKB-SIN (Etymotic Research, 2007).
Etymotic Research, Inc. BKB-SIN™ Speech-in-Noise Test. http://www.etymotic.
com (accessed 3 October, 2007).
Firszt JB, Holden LA, Reeder R, Skinner MW. Spectral channels and speech
recognition in cochlear implant recipients who use HiRes 120 sound processing.
Otol Neurotol, submitted.
Speech Recognition and Everyday Listening Benefits in Adults with Cochlear Implants
Elizabeth L. Roth, Au.D.
Kaiser Permanente, Sacramento, CA, USA
To date, HiRes with Fidelity 120 (HiRes 120) results have
to engage in social situations requiring hearing abilities for
been obtained in previously implanted adults who had an
effective communication). Because there is limited data on
average of two years of device use at the time they were
how self-reported listening and social benefits evolve in
programmed with HiRes 120 sound processing. Most of these
newly implanted adults, this study will evaluate listening
individuals experienced enhanced speech perception, sound
benefits in adults who are programmed with HiRes 120
quality, and music appreciation with the software upgrade
sound processing at initial device activation. In addition, the
(Advanced Bionics, 2007; Firszt et al, 2007). The most
study will examine changes in speech recognition in quiet
striking improvements were revealed through self-assessment
and in noise over time with HiRes 120. Performance on
questionnaires that evaluated benefits in everyday listening
speech perception tests and questionnaires will be tracked
situations. These novel tools provided a means for assessing
after 1, 3, 6, and 12 months following device activation.
listening benefits that were not captured in traditional tests of
speech recognition—that is, in challenging conditions such
as multiple conversations and talkers, listening at a distance,
listening in noise, and listening to sounds at low intensities.
The questionnaires also revealed some of the social benefits
that accompany improved listening abilities with a cochlear
implant (for example, greater self-confidence and willingness
52 References
Advanced Bionics. (2007) HiRes with Fidelity 120 Clinical Results. Valencia, CA:
Advanced Bionics.
Firszt JB, Koch DB, Downing M, Litvak L. (2007) Current steering creates
additional pitch percepts in adult cochlear implant recipients. Otol and Neurotol
(2007) 28(5):629-636.
Advanced Bionics® Auditory Research Bulletin 2007
Pediatric Results Following Conversion to HiRes 120
from Standard HiRes on the Harmony Processor
Thierry Van Den Abbeele, Prof., M.D.1
Nathalie Noël-Petroff, M.D.1
Laure Arnold, M.Sc.2
Natacha Crozat-Teissier, M.D.1
Patrick Boyle, M.Sc.2
“All five children preferred HiRes 120
immediately after programming and after
one month of HiRes 120 experience.”
1 Hôpital Robert Debré, Paris, France
2 Advanced Bionics, Europe
A recent study carried out at the Hôpital Robert
Debré in Paris showed that a prototype version of the
Harmony processor could be fit easily on pediatric
HiRes 90K implant users without reprogramming.
Children accepted the new processor readily and
showed stable speech performance after one month
of use (Crozat-Teissier et al, 2006). A continuation
of that study examined the effects of programming
the children with HiRes with Fidelity 120 (HiRes
120) sound processing. The five subjects, ages 6-16
years, first were tested with standard HiRes on the
commercial Harmony processor. Subsequently, they
were programmed with HiRes 120 and evaluated
immediately after switchover. Children went home
with both programs on their processors and were
instructed to use both HiRes and HiRes 120 for at
least one month and to keep a battery log. Parents
completed a benefits questionnaire after at least one
month of HiRes and HiRes 120 use.
The conversion to HiRes 120 was uneventful and
the most comfortable levels (M levels) with HiRes
120 were similar to M levels with HiRes. All five
children preferred HiRes 120 immediately after
programming and after one month of HiRes 120
experience. Their comments included “I hear better”
and it sounds “more beautiful.” Parents reported
that their children required less repetition for
speech understanding, that they required a lower
TV volume, and that they could hear whispering
with HiRes 120. More environmental sounds were
perceived (such as birds, wind, skis on snow). One
child now plays the bassoon. Parents also reported
that the lighter weight of the Harmony was preferable to the previous behind-the-ear processor
and that the battery lasted a full day (alleviating
the need to carry extra batteries to school). Suggestions for improvement included a smaller case and a
more robust connection between the processor and
headpiece.
This ongoing study will continue to follow the progress of these children along with 11 other children
who have been programmed with HiRes 120 on the
body-worn Platinum Sound Processor (PSP).
References
Crozat-Teissier N, Noël-Petroff N, Arnold L, Boyle P, Van Den Abbeele T. The Auria+
processor: first results with children. Presented at the 8th Pediatric Cochlear Implant
Conference Venice, Italy. 26 March, 2006.
Van Den Abbeele T, Noël-Petroff N, Arnold L, Crozat-Teissier N, Boyle P. Pediatric
results following switch-over from HiRes to HiRes 120. Presented at the 11th
International Conference on Cochlear Implants in Children, Charlotte, North
Carolina, USA. 11-14 April, 2007.
HiResolution Sound Processing 53
Programming Children in HiResolution Sound
Kim Veekmans, M.Sc.
Mark Daft, B.Sc.
Kelvin Hawker, Ph.D.
Susan Johnson, M.Sc.
Tracey Twomey, M.Sc.
Nottingham Cochlear Implant Programme,
Nottingham, United Kingdom
At the Nottingham Cochlear Implant Programme,
39 children have received HiRes 90K devices, one
of whom was implanted bilaterally. This study was
undertaken to determine average program parameters in this group of pediatric recipients, to investigate how soon a program is optimal after initial
programming, and to evaluate our programming
techniques based upon those findings. The average
age at initial programming was 4.8 years (range: 1.215.2 years). Etiologies included genetic (3), Usher
syndrome (1), Waardenburg syndrome (1), meningitis (4), cytomegalovirus (4), premature birth (3),
and unknown (23). Eleven children were diagnosed
with one or more additional difficulties such as
cognitive or physical impairments and autistic spectrum disorder. Two children experienced a period of
non-use because of severe illness and one teenager is
an inconsistent user.
500
Clinical Units
450
400
350
300
250
200
150
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16
Electrode
Figure 1. Box and whisker plots of six-month M levels for each
of the 16 electrodes (n = 26). Median, 25th percentile, 75th
percentile, and range are shown for each electrode.
54 Programming was carried out using the SoundWave software at the following intervals: 1-2 days,
4 weeks, 8 weeks, 3 months, 6 months, and 12
months. Behavioral data as well as electrophysiologic measures were used to set program parameters. HiRes-S programs were created routinely
using Speech Bursts for all children. For fine tuning
and in more complex cases, tone bursts were used
to create programs. A HiRes-P program was evaluated in one child, but the HiRes-S program proved
more beneficial. The standard input dynamic range
(60 dB) and the extended frequency range (333–
6665 Hz) were used for all children. When possible
and appropriate, sound detection thresholds for 0.5,
1, 2, and 4 kHz were measured (in dB HL) in order
to verify the program settings.
Results showed that the greatest increases in
M levels occurred at the day-2, 4-week, and
3-month test intervals. Most M levels were between
200 and 300 Clinical Units (15–23 nC/phase).
After six months, the programs stabilized and only
minor increases in M levels were required. After
this time, the programming focused on fine tuning.
Figure 1 shows the M levels for 26 children who
have reached the six-month test interval. There was
considerable variability among subjects for M levels
as well as variability in stimulation rate (median:
2047 pps/channel) and pulse width (median: 15.3
µs/phase). For 6 of the 26 children, the pulse width
was set manually because there was insufficient
loudness growth. The M-level data indicate that
upper programming levels are similar for the electrodes within each speech-burst band, which reflects
the technique used for routine programming. There
also is a trend for more variability and a larger spread
towards higher levels of stimulation for the basal
Advanced Bionics® Auditory Research Bulletin 2007
“All subjects except one were able
to detect sound at 35 dB HL for the
four audiometric frequencies by the
time of their six-month test interval.“
Neural Response Imaging (NRI) was used routinely
with all children intraoperatively. NRI measurements were repeated postoperatively in 59% of the
children with at least three months of device use.
NRI responses when not obtainable, tended to be
for the basal electrodes. Electrical stapedial reflex
recordings were attempted in 64% of subjects who
were over three years of age at initial programming
(n = 20). Reflex thresholds were measurable in 67%
of those children, and the data were used along with
behavioural data to set M levels.
Sound detection levels were measured for 21 of the
26 children with six-month data (Figure 2). Thresholds were not obtainable in five children either
because of test fatigue or because of concomitant
other difficulties. All subjects except one were able to
detect sound at 35 dB HL for the four audiometric
frequencies by the time of their six-month test
interval. This ability coincides with the six-month
stabilization of their program parameters. However,
some children took a longer time developing the
ability to detect sound at 4 kHz. This delay may
be explained by the children’s unfamiliarity with
hearing high-frequency sounds, by the fact that
higher stimulation levels were required for this part
of the array, or both. These factors appear to coincide only for 4 kHz and not for lower frequencies. In
view of this finding, it is possible that the M levels
were increased unnecessarily for the basal part of
the array. This programming issue currently is under
investigation.
500 Hz
1 kHz
2 kHz
4 kHz
100
Cumulative Percent
channels. In our experience, this trend may reflect
two underlying causes. First, some children require
more stimulation in the high frequencies and therefore uncomfortable levels appear to be higher at the
basal end of the array. Second, some children initially
detect sound better at the low and mid frequencies
than at the high frequencies, which has led us to
increase the stimulation level for the more basal end
of the array. When comparing the HiRes 90K data
to average map data and charge requirements for the
Cochlear device (N24) from a previous study (Veekmans et al, 2006, n = 139), the M/C levels are set
higher for the HiRes 90K device than for the N24R
device. However, the HiRes 90K levels fall within
the interquartile range of the Cochlear data and the
map/program profiles are similar.
80
60
40
20
0
Day 1
Day 2
4 wks
8 wks
3 mos
6 mos
Postoperative Test Interval
Figure 2. Cumulative percentage of children able to detect
sound at 30 dB HL at initial stimulation through six months of
implant use (n = 21). Sound detection was measured for four
audiometric frequencies.
Reference
Veekmans K, Johnson S, Robbins A. (2006) Charge requirements for children, does
this influence our programming technique? Presented at the 8th Pediatric Cochlear
Implant Conference, Venice, Italy. 24-27 March, 2006.
HiResolution Sound Processing 55
Speech Perception in Children Using Two Generations of
Advanced Bionics Cochlear Implants: Three-Year Results
Ersilia Bosco, Cl.Psychol.
Patrizia Mancini, M.D.
Luciana D’Agosta, Sp.Th.
Gabriella Traisci, Sp.Th.
Chiara D’Elia, M.D.
Roberto Filipo, M.D.
University of Rome La Sapienza, Rome, Italy
“. . . the children using
HiResolution processing who were
implanted before two years of age
and tested at 24 months
tend to attain follow-up scores similar
to those attained at 36 months by children
implanted at a later age. . .”
The development of speech perception in children
provided with a cochlear implant depends on both
audiological and nonaudiological variables (Bosco
et al, 2005; Geers et al, 2003; O’Donoghue et al,
1998; Waltzman et al, 1997). Technological evolution of cochlear implants has led to modern sound
processing strategies, such as HiResolution, that
are intended to provide more fine structure and
place-pitch information than earlier sound coding
methods. The original Clarion cochlear implant
system provided eight channels of spectral information and relatively low rates of stimulation (typically
800 pulses per second per channel) and implemented both the Continuous Interleaved Sampling
(CIS) and the Simultaneous Analogue Stimulation
(SAS) strategies. The contemporary CII and HiRes
90K devices support the family of HiRes strategies.
With standard HiRes, up to 16 channels of spectral
56 information is provided and temporal information
is delivered by as many as 5100 pulses per channel.
In addition, the quality of the electronics used in
today’s processors has improved, allowing increases
in the acoustic (input) dynamic range. In this longterm investigation, we have been focusing on the
effect of different sound coding strategies on the
speech perception skills of children implanted with
different generations of Advanced Bionics devices.
The study population consists of 49 children divided
into two experimental groups. The Standard Resolution Mode (SRM) group is comprised of 16 Clarion
1.2 (CIS or SAS) users. The HiResolution Mode
(HRM) group consists of 33 CII or HiRes 90K
users. All children were prelinguistically deaf and
differed in age at implantation as well as etiology of
deafness. Tests are selected for each child according
to Erber’s hierarchical model (1982), making use of
a speech perception test battery adapted to the age
of each child (Bosco et al, 2005). Tests are administered at the initial fitting and at 3, 6, 9, 12, 24, and 36
months thereafter. Results are pooled and analyzed
across the different tests according to Erber’s fourcategory hierarchy of auditory skills: detection, identification, recognition, and comprehension.
The data from the 12-, 24-, and 36-month followup intervals have been analyzed for three subgroups
of children: (1) children implanted before two years
of age (HRM: n = 5, mean age = 1.8 years, SRM:
n = 4, mean age 2 years), (2) children implanted
between two and five years of age (HRM: n = 17,
mean age 3.1 years, SRM: n = 6, mean age 2.6 years),
and (3) children implanted over five years of age
(HRM: n = 11, mean age 9 years, SRM: n = 6, mean
age 8.9 years).
Advanced Bionics® Auditory Research Bulletin 2007
References
Bosco E, D’Agosta L, Mancini P, Traisci G, D’Elia C, Filipo R. (2005) Speech
perception results in children implanted with Clarion devices: Hi-resolution and
Standard resolution modes. Acta Otolaryngol 125(2):148-158.
Erber NP. (1982) Auditory Training. Washington, DC: Alexander Graham Bell
Association for the Deaf.
Geers A, Brenner C, Davidson L. (2003) Factors associated with development of
speech perception skills in children implanted by age five. Ear Hear 24:24S-35S.
O’Donoghue GM, Nikolopoulos TP, Archbold SM, Tait M. (1998) Speech perception
in children after cochlear implantation. Am J Otol 19:762-767.
SRM
100
Percent Correct
80
Implant at
< 2 years of age
HRM
Implant between
2–5 years of age
Implant at
> 5 years of age
60
40
20
0
12
24
36
12
24
36
12
24
36
Implant Use (months)
Figure 1. Pooled speech recognition results for three groups of
children, according to age of implant, evaluated after 12, 24,
and 36 months of implant use with previous generation (SRM)
versus HiResolution (HRM) sound processing.
SRM
100
80
Percent Correct
Three-year results show that the HRM children
develop better speech perception skills at 12, 24,
and 36 months postimplant compared to the SRM
children (Figures 1 and 2). Moreover, the children
using HiResolution processing who were implanted
before two years of age and tested at 24 months tend
to attain follow-up scores similar to those attained
at 36 months by children implanted at a later
age (2-5 years and > 5 yrs). At the 36-month test
interval, there is no significant difference in performance between children implanted between 2-4
years and after five years of age. These results might
be influenced by the small number of subjects
belonging to the SRM group, and further data will
be needed from other cochlear implant centers in
order to confirm these findings. Formal testing and
clinical observation show that HRM users appear
to make a more natural use of acoustic information available from their environment. Clinical
experience has been corroborated by anecdotal
reports by experienced caregivers such as parents,
teachers, and speech therapists. The HRM children
appear to benefit from incidental learning—
particularly idiomatic and dialectal expressions,
swear words, diminutives, nursery rhymes, songs,
and rhyming verses.
Implant at
< 2 years of age
HRM
Implant between
2–5 years of age
Implant at
> 5 years of age
60
40
20
0
12
24
36
12
24
36
12
24
36
Implant Use (months)
Figure 2. Pooled speech comprehension results for three groups
of children, according to age of implant, evaluated after 12, 24,
and 36 months of implant use with previous generation (SRM)
versus HiResolution (HRM) sound processing.
Waltzman SB, Cohen NL, Gomolin RH, Green JE, Shapiro WH, Hoffman RA, Roland
JT. (1997) Open set speech perception in congenitally deaf children using cochlear
implants. Am J Otol 18:342-349.
HiResolution Sound Processing 57
Auditory Skill Development in Newly Implanted
Children: Effect of Sound Processing
Jose Antonio Rivas, M.D.1
Adriana Rivas, M.D.1
Jorge E. Almario, M.D.2
Jose Alberto Prieto, M.D.2
Maria Piedad Nuñez2
1 Clínica Rivas, Bogotá
2 Clínica San Rafael, Bogotá
“. . . the children programmed with
HiRes develop auditory skills earlier
than children programmed with CIS,
particularly if they were implanted
before three years of age.”
HiRes (n=6)
CIS (n=3)
An ongoing longitudinal clinical study is evaluating
auditory development in children implanted with
the HiRes 90K implant between 2004 and 2006.
Children were divided into groups based upon their
age at time of implantation and were assigned to be
programmed with either Continuous Interleaved
Sampling (CIS) or HiResolution (HiRes) sound
processing. To date, 26 children have been followed
after implantation. Subjects were evaluated with the
Protocolo de Habilidades Auditivas de la Universidad de Navarra in which a hierarchical battery of
tests evaluates auditory detection, discrimination,
identification, recognition, and comprehension using
materials appropriate for the age of the child. Children were evaluated before implantation and at 3, 6,
9, 12, 18, and 24 months after initial stimulation.
100
For sound processing assignment and data analyses,
children were divided into four groups—those
implanted between 12 and 36 months of age, those
implanted at 4 years of age, those implanted between
5 and 8 years of age, and children implanted at
9 years of age and older. Auditory skill development
was compared over time and by sound-processing
type.
Percent Detection
80
60
40
20
0
3 mos
6 mos
9 mos
12 mos
18 mos
24 mos
Duration of Implant Use
Figure 1. Mean auditory detection scores for children implanted
between one and three years of age during the first 24 months
of implant use (total n = 9, HiRes = 6, CIS = 3).
58 Figure 1 shows mean detection scores over time
for the youngest group of children. All children
were able to detect sound accurately by 24 months
after implantation, but the children using HiRes
attained that skill at a faster rate than the children
using CIS.
Advanced Bionics® Auditory Research Bulletin 2007
Figure 3 shows individual comprehension abilities
for seven children implanted at five years of age or
older (three children in that age group had 0% scores
up to 24 months and are not shown). All seven used
HiRes sound processing. Although there is a wide
range of ability, all children show improvement over
time on this highest level of auditory skill development.
These results indicate that all children receive
benefit from their devices during the first 24 months
of use. Moreover, the children programmed with
HiRes develop auditory skills earlier than children
programmed with CIS, particularly if they were
implanted before three years of age.
HiRes (n=6)
CIS (n=3)
100
Discrimination
Identification
Percent Correct
80
60
40
20
0
3 mos
6 mos
9 mos
12 mos
3 mos
6 mos
9 mos
12 mos
Duration of Implant Use
Figure 2. Mean discrimination and identification scores for children implanted between one and three years of age during the
first 12 months of implant use (total n = 9, HiRes = 6, CIS = 3).
S1
S2
S3
S4
S5
S6
S7
100
80
Percent Correct
Figure 2 shows mean discrimination and identification scores for the youngest group in the first 12
months after initial stimulation. Again, all subjects
improve over time, but the children using HiRes
develop these skills faster and to a greater degree
than the children using CIS.
60
40
20
0
3 mos
6 mos
9 mos
12 mos
18 mos
24 mos
Duration of Implant Use
Figure 3. Individual auditory comprehension scores for seven
children implanted at five years of age or older. All children use
HiRes sound processing.
HiResolution Sound Processing 59
Auditory and Language Abilities in Children:
Comparison of Two Different Cochlear Implant Systems
Maria Cecilia Bevilacqua, Ph.D.
Natália Barreto Frederigue, Ph.D.
Leandra Tabanez Nascimento, Ph.D.
Márcia Yuri Tsmura Kimura
Adriane Lima Mortari Moret, Ph.D.
Orozimbo Costa, M.D.
University of Sao Paulo, Bauru, Brazil
“The results of this study suggest
differences in acquisition of
auditory and oral language skills for
children implanted with either
HiRes 90K or Nucleus 24 devices,
at least in those skill areas assessed. . .”
Children receive cochlear implants at increasingly
younger ages because of technological advances that
have led to more reliable diagnosis of hearing loss
(Kirk, 2000; Kishon-Rabin et al, 2005). However,
the evaluation of auditory and oral language abilities remains a challenge because traditional tests of
speech understanding are not appropriate for these
young children. Alternatively, measures that use
parental report or other types of questionnaires and
classification scales designed to capture behavior in
everyday situations are useful tools with these young
children. In this study, auditory and oral language
abilities were tracked with two tools that use a
parent-interview technique to assess the development of auditory skills (Infant-Toddler Meaningful
Auditory Integration Scale or IT-MAIS) and oral
language (Meaningful Use of Speech Scale or
MUSS). Both tools were translated and adapted to
Portuguese for this study.
Thirty-four children participated in the study. Eighteen used the HiRes 90K implant (Group 1) and
16 children used the Nucleus 24 device (Group 2).
Children in each group were matched based on age
at implantation (range = 1 year, 2 months to 3 years,
5 months). Mean age at implant was 2 years,
2 months for Group 1 and 2 years for Group 2.
Children in both groups had similar socioeconomic
status and were followed by the same team of professionals. The IT-MAIS and MUSS were administered prior to implantation, and after 3, 6, and 12
months of device use.
60 Advanced Bionics® Auditory Research Bulletin 2007
The results of this study suggest differences in acquisition of auditory and oral language skills for children implanted with either HiRes 90K or Nucleus
24 devices, at least in those skill areas assessed by the
IT-MAIS and MUSS. These differences are likely
related to the sound processing implemented with
each device. Further investigation is needed to validate these results.
Nucleus 24
100
80
70
60
50
40
30
20
10
0
n=13 n=15
n=13 n=15
preimplant
3 months
n=6
6 months
n=1
n=1
12 months
Figure 1. Mean scores and standard deviations for the IT-MAIS
over time for children implanted with HiRes 90K or Nucleus 24
devices.
Nucleus 24
HiRes 90K
90
References
80
Percent Correct
Kishon-Rabin L, Taitelbaum-Swead R, Ezrati-Vinacour R, Hildesheimer H. (2005)
Prelexical vocalization in normal hearing and hearing-impaired infants before and
after cochlear implantation and its relation to early auditory skills. Ear Hear
26(Suppl 4):17S-39S.
n=6
Assessment Period
100
Kirk KI. (2000) Challenges in cochlear implant outcomes. In: Niparko J, ed.
Cochlear Implants: Principles and Practices. New York: Lippincott Williams &
Wilkins, 225-259.
HiRes 90K
90
Percent Correct
There was significant improvement between the
preimplant and three-month scores on the IT-MAIS
(Figure 1) and MUSS (Figure 2) for both groups of
children. Notably, there was a trend for better performance by the children in Group 1 (HiRes 90K users)
than Group 2 (Nucleus 24 users) on both measures,
with significant differences between groups on the
MUSS at three months (p = 0.029).
70
60
50
40
30
* p = 0.029
20
10
n=13
n=15
n=13 n=15
n=6
n=6
n=1
n=1
0
preimplant
3 months
6 months
12 months
Assessment Period
Figure 2. Mean scores and standard deviations for the MUSS
over time for children implanted with HiRes 90K or Nucleus 24
devices.
HiResolution Sound Processing 61
HiRes 120 in Children Using the HiRes 90K Device:
Preliminary Report
Patrizia Mancini, M.D.
Ersilia Bosco, Cl.Psychol.
Luciana D’Agosta, Sp.Th.
Gabriella Traisci, Sp.Th.
Chiara D’Elia, M.D.
Filipo Roberto, M.D.
University of Rome La Sapienza, Rome, Italy
Excellent results with standard HiRes processing
have been demonstrated for both adults (Koch et
al, 2004) and children (Bosco et al, 2004). However,
standard HiRes delivers only 16 spectral bands,
thereby limiting the available spectral information.
Psychophysical work has shown that CII and HiRes
90K implant users can hear additional pitch percepts
when adjacent electrode contacts are stimulated
simultaneously using current steering (Koch et al,
2007). This finding led to development of HiRes 120
in which eight spectral bands are created for each of
the 15 electrode pairs, thereby increasing the spectral
resolution. This ongoing work is evaluating standard
HiRes and HiRes 120 in children implanted with a
HiRes 90K device.
Table 1. Demographics for the three groups of children.
Group
Number
Mean
Age
(Months)
Mean
Device Use
(Months)
Initial HiRes
6
59.3
31.7
HiRes-to-HiRes
120 Conversion
6
53.7
25.6
Initial HiRes 120
5
30.2
2.5
62 The results from three groups of children have been
analyzed to date, one fitted from the beginning in
HiRes, one converted from HiRes to HiRes 120,
and one fitted from the beginning with HiRes 120.
The ages and duration of implant use are shown
in Table 1. All children were fitted on the basis of
Neural Response Imaging thresholds, the responses
confirmed by careful behavioural assessment, and
verified using free field audiometry. Clinical SoundWave software was used for all fittings and all children used the PSP bodyworn processor. For the
converted group, no changes were made to either
hardware or psychophysical levels when changing
from standard HiRes to HiRes 120.
A speech perception test battery (Bosco et al, 2004)
was presented by experienced clinicians using live
voice or prerecorded materials. Analyses focused on
determining performance using Erber’s hierarchical
levels of detection, identification, recognition, and
comprehension (Erber, 1982). Additional measures
of speech perception included the ASSE (Govaerts
et al, 2006) in which pairs of carefully processed
consonants are presented in a sequence containing
an odd-one-out to verify both discrimination and
identification of phonemes. Results are reported as
percent improvement for discrimination between
two different phonemes (u-i, u-a, i-a, m-z, s-f, z-s,
f-z) and percent overall improvement in identification. To complement these measures two structured
parental interviews—the MAIS (Robbins et al,
1991) and the PRISE (Kishon-Rabin et al, 2005)
—were administered.
Figure 1 shows the baseline and three-month mean
scores for the three groups on Erber’s four categories
of speech perception. Although the demographics for
the HiRes and the HiRes-to-HiRes 120 conversion
groups are similar, the mean scores for the children
in the conversion group are higher for identification,
recognition, and comprehension at the three-month
test interval. Note that the children programmed
solely with HiRes 120 show lower scores overall.
Advanced Bionics® Auditory Research Bulletin 2007
3 months
100
80
60
40
20
Identification
Recognition
20
Hi
Re
s1
s
n
Re
sio
Hi
er
nv
Co
20
Re
s1
s
n
Re
sio
Hi
Hi
er
nv
Co
20
Re
s1
s
n
Re
sio
Co
Discrimination
Hi
Hi
nv
er
20
s1
Re
sio
Hi
er
nv
Hi
Re
s
n
0
Co
Parental responses in structured interviews and
questionnaires revealed that the children converted
to HiRes 120 showed increased attention, improved
ability to discriminate loudness and pitch, and
better comprehension of speech than the HiRes
children—particularly for noisy listening situations.
Interestingly, the data also suggested greater attention to rhythm and melody for this group. Moreover, speech production quality and intelligibility
were reported to be better for the children receiving
HiRes 120 as their initial strategy. These findings
were complemented by anecdotal reports of greater
incidental learning of nursery rhymes, short songs,
and poems in the initial HiRes 120 group. Further
work is required to monitor performance over time,
particularly for the younger children programmed
initially with HiRes 120.
baseline
Percent Correct
However, these children are much younger than the
children in the other two groups. Figure 2 shows the
mean ASSE scores for the HiRes and HiRes-toHiRes 120 conversion groups. The HiRes children
improved slightly whereas the conversion group
showed significant gains at the three-month test
interval.
“. . . speech production quality and
intelligibility were reported to be better
for the children receiving
HiRes 120 as their initial strategy.”
Comprehension
Figure 1. Mean performance at baseline and three months
for the HiRes group, HiRes/HiRes 120 conversion group, and
HiRes 120 group across the categories of discrimination, identification, recognition, and comprehension.
100
baseline
3 months
References
Bosco E, D’Agosta L, Mancini P, Traisci G, D’Elia C, Filipo R. (2004) Speech
perception results in children implanted with Clarion devices : Hi-Resolution and
Standard resolution modes. Acta Otolaryngol 125:148-158.
Erber N.P. ( 1982) Auditory Training. Washington, DC: Alexander Graham Bell
Association.
Govaerts PJ, Daemers K, Yperman M, De Beukelaer C, De Ceulaer G. (2006) A new
test to assess detection, discrimination, and identification in hearing impairment.
Coch Impl Int 2: 97-106.
Kishon-Rabin L, Taitelbaum-Swead R, Ezrati-Vinacour R, Hildesheimer M. (2005)
Prelexical vocalization in normal hearing and hearing-impaired infants before and
after cochlear implantation and its relation to early auditory skills. Ear Hear 26:17S29S.
Koch DB, Downing M, Osberger MJ, Litvak L. (2007) Using current steering to
increase spectral resolution in CII and HiRes 90K users. Ear Hear 28(2):38S-41S.
Percent Correct
80
60
40
20
0
HiRes Group
HiRes-to-HiRes 120
Conversion Group
Figure 2. ASSE identification scores at baseline and at the threemonth test interval for the HiRes group and the HiRes/HiRes
120 conversion group. A greater degree of improvement is seen
for the HiRes/HiRes 120 conversion group.
Koch DB, Osberger MJ, Segel P, Kessler D. (2004) HiResolution and conventional
sound processing in the HiResolution Bionic Ear: using appropriate outcome
measures to access speech recognition abilities. Otol Neurotol 9:214-223.
Robbins AM, Renshaw JJ, Berry SW. (1991) Evaluating meaningful auditory
integration in profoundly hearing-impaired children. Am J Otol 12(Suppl):144-150.
HiResolution Sound Processing 63
HiRes 120 and Lexical Tone Perception in
Mandarin-Speaking Children
Li Xu, M.D., Ph.D.1
Bo Liu, M.D.2
Xueqing Chen, M.Med.2
Yan Zheng, B.Med.2
Ying Kong, B.Med.2
Haihong Liu, M.Med.2
Ning Zhou, M.A.1
Demin Han, M.D., Ph.D.2
1 Ohio University, Athens, OH, USA
2 Beijing Tongren Hospital/Beijing Institute of Otorhinolaryngology, Beijing, People’s Republic of China
“. . .the results of the present study. . .
indicate that the temporal and
spectral information delivered
by HiResolution sound processing
may underlie an improved ability
to hear tonal contrasts in some of the
Mandarin-speaking children.”
A fundamental characteristic of tone languages (such
as Mandarin Chinese, Thai, and Vietnamese) is that
tone patterns in a syllable convey lexical meaning.
Lexical tone perception varies dramatically among
cochlear implant users who are native tone language
speakers, and most adult users have great difficulty
distinguishing the tonal contrasts important for
understanding speech (e.g., Huang et al, 1995; Sun
et al, 1998; Wei et al, 2004). Pediatric implant users
also have great difficulties in extracting the pitch
information needed to identify the lexical tones
(Ciocca et al, 2002; Lee et al, 2002; Wei et al, 2000;
Wong and Wong, 2004).
Studies suggest that a greater number of spectral
channels might provide the cues that are essential
for tone perception (Shannon, 2005). In fact, Xu
and others already have shown that spectral cues are
important for lexical tone perception and that fine
structure cues are essential for lexical tone perception
in normal-hearing Mandarin-speaking adults (Xu
et al, 2002; Xu and Pfingst, 2003; Kong and Zeng,
2006). Since HiRes with Fidelity 120 (HiRes 120)
sound processing captures and sends more spectral
information to the cochlea, this study evaluated the
potential benefit of HiRes 120 for improving tonelanguage perception in prelinguistically deaf children who are native speakers of Mandarin Chinese.
The hypothesis was that HiRes 120 represents the
pitch information and fine structure necessary
for distinguishing the tonal contrasts required for
understanding words in Mandarin.
Twenty children ages 4 to 17 years participated in the
study. Performance was compared between standard
HiRes and HiRes 120 sound processing for words
that differed in tonal contrast in a two-alternative
picture-pointing format. All children used HiRes
and HiRes 120 on the body-worn Platinum Sound
Processor (PSP). First, they were tested with standard HiRes. Subsequently, they were programmed
with HiRes 120 and reevaluated after three months
of HiRes 120 use. In addition, at the end of the
experiment, the parents of 18 children completed a
64 Advanced Bionics® Auditory Research Bulletin 2007
Figure 2 summarizes the preference results. The
parents of 15 of the 18 children (83%) preferred
HiRes 120 over standard HiRes, whereas parents of
three children indicated no preference. Strength of
preference ratings averaged 7.3, with the parents of
six children rating their preference as a 10 (strongly
preferring HiRes 120).
HiRes
100
HiRes 120 @ 3 months
90
80
Percent Correct
The results showed a range of performance on the
tonal contrast test for both HiRes and HiRes 120.
Although the mean scores were similar (HiRes
mean = 74.3%; HiRes 120 mean = 75.2%), individual
scores showed that some subjects derived more
benefit from HiRes 120, whereas others performed
better with standard HiRes. Figure 1 shows the
results for each subject at baseline with HiRes and
after using HiRes 120 for three months. Subjects are
rank-ordered from lowest to highest performer with
standard HiRes. Five children demonstrated higher
scores with HiRes 120—with improvements ranging
from 9 to 22 percent. In contrast, three subjects
performed better with standard HiRes. Twelve children showed equivalent performance for the two
sound processing strategies. Notably, six of those 12
subjects scored 90% or higher with standard HiRes,
thereby precluding the use of the tonal contrast test
for assessing any improvement with HiRes 120.
Although 5 of the 20 subjects scored around chance
(50%) with their best strategy, 15 children achieved
relatively high levels of performance with HiRes or
HiRes 120. This finding suggests that some children may rely more on temporal fine structure cues,
whereas others use predominantly spectral cues for
perception of tonal contrasts.
“Parental observations indicate strongly
that their children’s
hearing, speaking, and music listening
has improved with HiRes 120.”
70
60
50
40
30
20
10
0
Individual Scores
Figure 1. Individual baseline HiRes and three-month HiRes 120
scores on a Mandarin tonal contrast test. Score pairs are rank
ordered by baseline HiRes results (chance = 50% correct).
6
5
Number of Subjects
questionnaire that evaluated the preference of speech
processing strategy and their children’s hearing and
production abilities when using HiRes 120. Preference for HiRes 120 (versus standard HiRes) was
rated on a scale from 0 ( no preference) to 10 (strong
preference). Evaluation of children’s perception and
production abilities were rated on a five-point scale
(1 = strongly disagree, 5 = strongly agree).
4
3
2
1
0
0
1
2
3
4
5
6
7
8
9
10
Rating
Figure 2. Distribution of strength of preference ratings. Ratings
range from 0 (no preference) to 10 (strong preference HiRes
120). Parents of 15 children preferred HiRes 120 whereas
parents of three children expressed no preference.
—continued on next page—
HiResolution Sound Processing 65
—continued from previous page—
Table 1. Percent of parents who
agreed or strongly agreed with statements
regarding HiRes 120 benefits.†
Statement of Perceived Benefit
with HiRes 120
Percent Responding
“Agree” or
“Strongly Agree”
My child understands speech better—
especially in the presence of background
noise.
73%
My child’s speech is easier to understand.
87%
My child’s speech sounds more natural.
53%
My child alerts to more sounds in the
environment.
67%
My child shows more interest in music.
87%
My child sings more.
80%
Responses are summarized for parents who preferred HiRes 120
(n = 15).
†
Table 1 summarizes the questionnaire ratings for
the parents who preferred HiRes 120 (n = 15). These
results indicate that the majority of parents “agree” or
“strongly agree” that HiRes 120 provided improved
speech understanding, speech production, and music
benefits.
In summary, the results of the present study are
encouraging and indicate that the temporal and
spectral information delivered by HiResolution
sound processing may underlie an improved ability
to hear tonal contrasts in some of the Mandarinspeaking children. Parental observations indicate
strongly that their children’s hearing, speaking,
and music listening has improved with HiRes 120.
All 20 children continue to use HiRes 120 and
will be reevaluated on the tonal contrast test after
six months of HiRes 120 experience. In addition,
analyses of the children’s tone production at baseline
with HiRes and after six months with HiRes 120
will be compared in a future report.
References
Ciocca V, Francis, AL, Aisha R, Wong L. (2002) The perception of Cantonese lexical
tones by early-deafened cochlear implantees. J Acoust Soc Am 111:2250-2256.
Huang TS, Wang NM, Liu SY. (1995) Tone perception of Mandarin-speaking
postlingually deaf implantees using the nucleus 22-channel cochlear mini system.
Ann Otol Rhinol Laryngol Suppl 166:294–298.
Kong Y-Y, Zeng F-G. (2006) Temporal and spectral cues in Mandarin tone
recognition. J Acoust Soc Am 120:2830-2840.
Lee KYS, Van Hasselt CA, Chiu SN, Cheung DMC. (2002) Cantonese tone perception
ability of cochlear implant children in comparison with normal hearing children. Int
J Pediatr Otorhinolaryngol 63:137-147.
Shannon RV. (2005) Speech and music have different requirements for spectral
resolution. Int Rev Neurobiol 70:121-134.
Sun JC, Skinner MW, Liu SY, Wang FNM, Huang TS, Lin T. (1998) Optimization
of speech processor fitting strategies for Chinese-speaking cochlear implantees.
Laryngoscope 108:560-568.
Wei C, Cao K, Zeng F-G. (2004) Mandarin tone recognition in cochlear-implant
subjects. Hear Res 197:87-95.
Wei WI, Wong R, Hui Y, Au DKK, Wong BYK, Ho WK, Tsang A, Kung P, Chung E.
(2000) Chinese tonal language rehabilitation following cochlear implantation in
children. Acta Otolaryngol 120:218-221.
Wong AO, Wong LL. (2004) Tone perception of Cantonese-speaking prelingually
hearing-impaired children with cochlear implants. Otolaryngol Head Neck Surg
130:751-758.
Xu L, Pfingst BE. (2003) Relative importance of the temporal envelope and fine
structure in tone perception. J Acoust Soc Am 114:3024-3027.
Xu L, Tsai Y, Pfingst BE. (2002) Features of stimulation affecting tonal-speech
perception: Implications for cochlear prostheses. J Acoust Soc Am 112:247-258.
66 Advanced Bionics® Auditory Research Bulletin 2007
Psychophysics, Modeling &
Novel Processing
Today’s HiResolution® sound processing options require only a fraction of
the capability of the CII Bionic Ear® and HiRes 90K™ electronic platforms.
These devices are enabling research on the expanded benefits of electrical
hearing through mechanism such as current focusing, neural conditioning
(stochasticity), and physiologically or psychophysically based stimulation. In
the event research findings culminate in commercial implementations, CII and
HiRes 90K users will have access to these advances through software applications without having to undergo additional surgery.
The studies in this section summarize investigations into the efficacy and feasibility of several novel processing approaches that can be implemented with
these electronic platforms from Advanced Bionics.
69
Channel Interaction Patterns with Simultaneous
Stimulation: Psychophysical and Physiological Measures
Michelle L. Hughes, Ph.D.
Lisa J. Stille, M.A.
Boys Town National Research Hospital, Omaha, NE, USA
One of the advantages of speech processing strategies that implement simultaneous (SAS) or partially
simultaneous (HiRes-P, MPS, PPS) stimulation
is that faster overall stimulation rates can be used.
However, when two electrodes are stimulated
simultaneously, the respective electrical fields can
interact if the electrodes are not far enough apart.
If the current fields are in phase, summation can
occur—potentially resulting in overstimulation and/
or a change in pitch percept. If the electrical fields
are inverted in phase relative to each other, then
cancellation can occur—resulting in reduced audibility. Previous studies have assessed psychophysical
spatial interaction patterns for simultaneous stimulation (Favre and Pelizzone, 1993; Shannon, 1983).
However, corresponding physiological data have
not been published. Therefore, it was of interest to
Table 1. Probe and interaction electrode pairs used
for ECAP and psychophysical measures.
Probe
Electrode
Interaction
Electrodes
5
1, 2, 4, 6, 8, 10, 16
9
1, 4, 6, 8, 10, 12, 14, 16
12
1, 7, 9, 11, 13, 15, 16
determine whether physiological spatial interaction
patterns obtained with simultaneous stimulation
were predictive of psychophysical patterns obtained
with the same stimulus configuration.
The purpose of this study was to compare physiological and psychophysical spatial interaction patterns
obtained with simultaneous stimulation of two
electrodes. Physiological interaction patterns were
determined by measuring the electrically evoked
compound action potential (ECAP) using alternating polarity as the method of artifact reduction.
Psychophysical interaction patterns were measured
using a three-interval, two-alternative forced choice
task in which the subject indicated the interval that
contained the sound. The stimulus for both measures
consisted of a 50-microsecond per phase biphasic
current pulse presented at a rate of approximately
30 pulses per second. All measures were made using
the Bionic Ear Data Collection System (BEDCS).
Physiological and psychophysical thresholds were
obtained in quiet for three probe electrodes (5, 9,
12) in each subject. Thresholds for each probe electrode then were obtained in the presence of a fixedlevel, in-phase, subthreshold interaction stimulus
presented on a different electrode. The interaction
stimulus was identical to the probe stimulus, with
the exception of level. All measures were made using
monopolar coupling. The electrode pairs used are
listed in Table 1.
To date, data have been collected for five subjects—
three implanted with a CII Bionic Ear implant and
two with a HiRes 90K device. Figure 1 shows ECAP
and psychophysical thresholds for probe electrode
9 plotted as a function of interaction electrode for
subject C10. The top panel shows both physiological
and psychophysical thresholds in quiet and for the
interaction conditions. Psychophysical and ECAP
thresholds in quiet are indicated by dotted horizontal lines and small symbols (circle and square,
70 Advanced Bionics® Auditory Research Bulletin 2007
“These preliminary results suggest that
the ECAP may provide
a time-efficient alternative
to psychophysical measures
of interaction patterns
for simultaneous stimulation
in cochlear implants.”
respectively). Thresholds in the presence of the
interaction stimulus (+85 μA) are indicated by filled
circles (psychophysics) and open squares (ECAP).
For each measure, threshold shift was calculated as
the difference between the quiet threshold and the
interaction threshold.
3IFC + Interaction
ECAP + Interaction
3IFC Quiet Threshold
ECAP Quiet Threshold
300
Threshold (uA)
250
200
150
100
50
0
3IFC
This work was funded by NIDCD grants R03 DC007017 and P30 DC04662.
The authors thank Donna Neff, Hongyang Tan, and Walt Jesteadt from Boys
Town National Research Hospital and Leo Litvak from Advanced Bionics for their
assistance.
References
Favre E, Pelizzone M. (1993) Channel interactions in patients using the Ineraid
multichannel cochlear implant. Hear Res 66:150-156.
Shannon, RV. (1983) Multichannel electrical stimulation of the auditory nerve in
man. II: Channel interaction. Hear Res 12:1-16.
ECAP
100
80
60
40
20
0
C10, P9
2
Acknowledgements
Interaction Stimulus Level
120
Threshold Shift (uA)
Threshold shifts for each measure are shown in the
bottom panel of Figure 1, with interaction stimulus level indicated by the horizontal line. Ideally,
threshold shifts should not be greater than the
level of the interaction stimulus. These data show
maximum threshold shift for interaction electrodes
adjacent to the probe electrode, with less interaction
for greater separation between probe and interaction
electrodes. More interaction (greater threshold shift)
is seen for apical interaction electrode locations in
this subject. ECAP and psychophysical thresholdshift patterns in Figure 1 were highly correlated
(r = 0.92, p = 0.001). The group data thus far show
a similar trend. There were no significant differences
in threshold shift between ECAP and psychophysical measures for the 15 probe electrodes tested
across the five subjects (p = 0.42). These preliminary
results suggest that the ECAP may provide a timeefficient alternative to psychophysical measures of
interaction patterns for simultaneous stimulation in
cochlear implants.
4
6
8
10
12
14
16
Electrode, Interaction Stimulus
Figure 1. Physiological (open squares) and psychophysical
(filled circles) spatial interaction patterns for probe electrode
9 in Subject C10. Top panel (A): Thresholds in quiet and with
stimulation of the interaction electrode as a function of interaction electrode. Bottom panel (B): Threshold shift (interaction
threshold minus threshold in quiet) as a function of interaction
electrode.
Psychophysics, Modeling & Novel Processing 71
Psychophysical Measurement of Spread of Excitation
Using Single and Dual Maskers
J. Gertjan Dingemanse, M.Sc.
Johan H. M. Frijns, M.D., Ph.D.
Jeroen J. Briaire, M.Sc.
Leiden University Medical Center, Leiden, The Netherlands
When stimulating the cochlea using multiple electrode cochlear implants, the intention is to provide a
number of channels of spectral information, ideally
one for each individual electrode contact. However,
research indicates that current delivered to one electrode contact can interfere with current delivered
to a neighboring contact. This spread of excitation
reduces the number of available channels to typically
no more than seven (Fishman et al, 1997; Friesen et
al, 2001). A recent study undertaken at our center
has shown that the optimum number of electrode
contacts should be selected on an individual basis for
each patient (Frijns et al, 2003). Therefore, having
knowledge of electrode contact interaction may be
clinically useful.
τ = 52.3
C = 26.0
K = 126.7
R2 = 0.968
S1
140
120
100
80
60
Threshold Shift (uA)
40
20
0
τ = 60.6
C = 20.5
K = 108.2
R2 = 0.961
S2
140
120
100
80
60
40
20
0
10
100
1000
Time Delay (ms)
Figure 1. Recovery functions for two subjects. Data were fit easily using an exponential model.
72 This study took a psychophysical approach
to investigate the spread of excitation using a
forward-masking paradigm. In contrast to previous
single-masker experiments, two maskers were used,
one on either side of the probe contact. The goal was
to reduce off-site listening—the electrical analog of
off-frequency listening in acoustic hearing (Moore,
2003). Using two maskers reduces the contribution of fibers located on the unmasked side of the
probe in a single-masker experiment. The technique
is similar to using notched noise in an acoustic
masking experiment.
Six postlinguistically deafened CII implant
recipients participated in the study. All had been
implanted with the HiFocus electrode and had used
their devices for 3-19 months. Stimuli were delivered
via the Clarion Research Interface and SPAIDE
research software. Electrode 8, located in the middle
of the array, was used for the probe stimulus. Masker
stimuli were delivered to electrodes symmetrically
on either side of Electrode 8. The masker was a 300
ms duration biphasic pulse train followed by a 20
ms duration probe. Pulse widths of 32 μs per phase
(μs/ph) were used for both masker and probe.
Monopolar electrode coupling was employed with
the dual maskers presented sequentially to avoid
field summation. A three-interval forced-choice
paradigm was used for measurement of probe-alone
and masked thresholds.
In advance of the experiments, mid-range stimulation levels were determined for each subject. These
levels then were used for masker stimulation as well
as the maximum comfortable stimulation current for
both masker and probe. To verify the experimental
setup, an initial experiment measured recovery from
masking. Masked thresholds were measured for a
series of masker-to-probe delays (20 ms and 400 ms).
Electrode 8 was used for both probe and masker. A
nonlinear least squares method was used to estimate
a parameter for an exponential masking recovery
function (Nelson & Donaldson, 2002). Recovery
function data for all subjects were well represented
by the exponential model. Figure 1 shows data from
subjects S1 and S2. Recovery was almost complete
by between 100 and 300 ms and the time constant
for recovery ranged between 21.6 and 114.9 ms.
Advanced Bionics® Auditory Research Bulletin 2007
The masking recovery data are in line with previous
work, validating this experimental approach. The
mean time constant was 71 ms, indicating that central
processes dictate recovery. The masking pattern
widths were 4.4 mm for the single masker and, on
average, one contact more for dual masking. The
additional width may have resulted from a reduction
in off-site listening, as shown in Figure 2. Because
masking patterns were generally symmetrical, dual
maskers may offer a substantial time advantage
when making width measurements because it is
not necessary to measure both sides of the masking
pattern. The differences between single and dual
masker thresholds for zero-spaced masking (double
pulse on Electrode 8 for a dual masker) indicate that
some charge summation exists. However, given that
the pulse rates represent the subjects’ everyday sound
coding strategy, it appears worthwhile to maintain
these stimulation rates to investigate operation as
close to clinical use as possible. In summary, a dual
masker technique can provide practical insight into
electrode contact interactions and can separate local
from off-site listening effects.
References
Fishman KE, Shannon RV, Slattery WH. (1997) Speech recognition as a function of
the number of electrodes used in the SPEAK cochlear implant speech processor.
J Speech Lang Hear Res 40:1201–1215.
“. . .a dual masker technique can provide
practical insight into
electrode contact interactions and
can separate local from off-site
listening effects. . .”
Dual
Single Apical
Single Basal
180
S1
x
160
140
120
100
80
60
40
Threshold Shift (uA)
The spread of excitation measures were obtained for
a variety of probe-to-masker spacings. The amount
of masking was always greater for dual masking than
for a single masker. The width of the masking pattern
was 1.1 to 5.3 contacts for a single masker and 1.1
to 6.1 for a dual masker. (Contact spacing is 1.1 mm
for the HiFocus electrode array.) Excluding Subject
4, whose data was atypical, masking patterns were
significantly greater for dual than for single maskers
(n = 5, p = 0.001; t = 8.1) and the widths of singleand dual-masking patterns were highly correlated
(R2 = 0.90, with p < 0.01).
20
0
180
160
140
x
x
2
3
S2
120
100
80
60
40
20
0
0
1
4
5
Masker/Probe Distance (Number of Contact Spacings)
Figure 2. Spread of excitation for single and dual maskers for
two subjects. The change in probe current is plotted against
the physical spacing between masker and probe. The asterisk
indicates a significant difference between the single and dual
masker conditions.
Friesen LM, Shannon RV, Baskent D, Wang X. (2001) Speech recognition in noise as
a function of the number of spectral channels: Comparison of acoustical hearing and
cochlear implants J Acoust Soc Am 110:1150–1163.
Frijns JHM, Klop WM, Bonnet RM, Briaire J J. (2003) Optimizing the number of
electrodes with high-rate stimulation of the Clarion CII cochlear implant. Acta
Otolaryngol 123:138–142.
Moore BCJ. (2003). An Introduction to the Psychology of Hearing, 5th Ed. San Diego:
Academic Press.
Nelson DA, Donaldson G S. (2002) Psychophysical recovery from pulse-train
forward masking in electric hearing. J Acoust Soc Am 112:2932–2947.
Psychophysics, Modeling & Novel Processing 73
Effects of High-Rate Pulse Trains on Modulation
Detection in Cochlear Implant Users
Christina Runge-Samuelson, Ph.D.
Jamie Jensen, Au.D.
P. Ashley Wackym, M.D.
Medical College of Wisconsin, Milwaukee, WI, USA
“These results indicate that
the addition of constant-amplitude
high-rate pulse trains
can improve modulation detection.”
Maximum Threshold Change (dB)
Electrical stimulation of the auditory nerve produces
highly synchronized neural responses, which results
in a narrow dynamic range of hearing for cochlear
implant users. This high neural synchrony can be
attributed to the absence of functional cochlear hair
cells and synapses, which in a normal ear are responsible for spontaneous, stochastic neural activity.
Rubinstein et al (1999) have proposed that highrate electrical pulse trains may desynchronize auditory nerve fiber responses to electric stimuli, thereby
reintroducing stochastic neural activity. This effect
is thought to be a result of maintaining fibers in
various states of relative refractoriness (Litvak et al,
2001; Miller et al, 2001; Rubinstein et al, 1999). In
turn, the reintroduction of stochastic neural activity
may increase neural sensitivity, thereby widening the
E2
16
E6
E12
14
12
10
8
6
4
2
0
1
2
3
4
5
6
7
8
9
10
Subject
Figure 1. Maximum modulation detection threshold improvements (in dB) for three electrodes in 10 subjects.
74 electrical dynamic range. A wider dynamic range
would ameliorate some of the detrimental aspects of
hearing with a cochlear implant.
Amplitude modulation detection reflects the sensitivity to changes in amplitude of a particular stimulus.
One potential advantage of introducing high-rate
pulse trains is improved modulation detection.
Improved modulation detection has been observed
when other stimuli designed to induce stochastic
neural activity (such as noise) have been introduced
(Chatterjee and Oba, 2005; Chatterjee and Robert,
2001). This study hypothesized that constantamplitude high-rate pulse trains would improve
modulation detection for cochlear implant users.
Subjects were 10 adult CII or HiRes 90K users.
The modulated stimuli were 1 kHz pulse trains
(75 us/phase; 400 ms bursts) delivered in monopolar
mode that were amplitude modulated by a 100 Hz
sinusoid. The modulated stimuli were presented at
each subject’s most comfortable level. Modulation
detection thresholds (MDTs) for the modulated
pulse trains were measured for test electrodes 2, 6,
and 12. Constant-amplitude 5-kHz pulse trains
(11 us/phase) were delivered to the two electrodes on
either side of the test electrode. For each test electrode, several randomized levels of high-rate pulses
were evaluated for their effects on modulation detection. A three-alternative forced-choice task (with
one modulated burst and two unmodulated bursts)
was used and the modulation depth was varied to
determine MDT at a level of 79% correct on the
psychometric function. The MDTs were calculated
in dB of modulation depth, with a smaller dB value
indicating better modulation detection (i.e., lower
MDT).
Data were analyzed to determine the maximum
MDT change elicited by the addition of high-rate
pulses for each electrode. Figure 1 shows the subjects’
maximum MDT changes in dB for electrodes 2,
Advanced Bionics® Auditory Research Bulletin 2007
Recent Public at ion
Threshold and Channel Interaction in
Cochlear Implant Users: Evaluation of the
Tripolar Electrode Configuration
Julie Arenberg Bierer, Ph.D.
University of Washington, Seattle, WA, USA
6, and 12. All but one subject showed some MDT
improvement on at least two of the three electrodes
tested. Across all subjects and electrodes, maximum
MDT improvement ranged from 0.4 to 15 dB, with
an overall average improvement of 3.1 dB.
The configuration of the contacts in a cochlear implant
These results indicate that the addition of constantamplitude high-rate pulse trains can improve
modulation detection. This stimulation approach
may be beneficial for speech processing strategies because the increased ability to encode smaller
changes in amplitude has the potential to improve
speech perception (particularly in noise) and music
appreciation.
bipolar and monopolar stimulation in nine subjects who
electrode array determines the spatial distribution of
cochlear activation. The limitations imposed by the
spatial and temporal interactions among contacts, in turn,
affect the efficacy of the device. This study compared a
spatially restricted tripolar electrode configuration to
had been implanted with an Advanced Bionics HiFocus I
electrode array. Thresholds for single-channel stimulation
were higher and more variable for tripolar stimulation
compared to bipolar or monopolar mode. For individual
subjects, channel-to-channel threshold variability in the
tripolar configuration was presumed to reflect the status
of the interface between the electrode array and surviving
neurons. Greater variability was negatively correlated
with speech perception. Thus tripolar stimulation might be
Acknowledgement
used as a tool for determining the location of suboptimal
This research was supported by the National Institute on Deafness and Other
Communication Disorders (NIDCD R03DC006361).
stimulation—whether as a result of poor neural survival,
References
Chatterjee M, Oba SI. (2005) Noise enhances modulation detection by cochlear
implant listeners at moderate carrier levels. J Acoust Soc Am 118(2):993-1002.
Chatterjee M, Robert ME. (2001) Noise enhances modulation sensitivity in cochlear
implant listeners: stochastic resonance in a prosthetic sensory system? J Assoc Res
Otolaryngol 2(2):159-171.
Litvak L, Delgutte B, Eddington DK. (2001) Auditory nerve fiber responses to
electrical stimulation: Modulated and unmodulated pulse trains. J Acoust Soc Am
110(1):368-379.
Miller CA, Abbas PJ, Robinson BK. (2001) Response properties of the refractory
auditory nerve fiber. J Assoc Res Otolaryngol 2:216-232.
poor positioning of the electrode contact, or other factors.
Programming parameters then could be adjusted (such as
stimulation mode or inactivation of contacts) to optimize
benefit for individual listeners. Tripolar stimulation can be
implemented in all CII and HiRes 90K implant recipients
because each electrode contact is powered by a separate
programmable current source.
J Acoust Soc Am (2007) 121(3):1642-1653.
Rubinstein JT, Wilson BS, Finley CC, Abbas PJ. (1999) Pseudospontaneous activity:
stochastic independence of auditory nerve fibers with electrical stimulation. Hear
Res 127:108-118.
Psychophysics, Modeling & Novel Processing 75
Effects of Stimulation Rate on Fundamental Perceptual
Abilities Underlying Speech Perception
Tim Green, Ph.D.
Andrew Faulkner, Ph.D.
Stuart Rosen, Prof., Ph.D.
University College London, London, United Kingdom
“The pattern of results across pulse rate
was markedly different in these
“fixed-profile” conditions,
and spectral shape discrimination
was no longer substantially poorer
with 2,894 pulses per second.”
AM Threshold (dB re: Dynamic Range)
C1
C2
C3
C5
C6
-60
-50
-40
-30
-20
-10
482
723
1447
2894
5787
Pulse Rate (pps)
Figure 1. Thresholds for the detection of 10 Hz sinusoidal
amplitude modulation on a single electrode, expressed in dB
relative to the electrode’s dynamic range. The carrier level was
75% of the dynamic range.
Higher stimulation rates have been advocated on
the grounds that they might allow an enhanced
representation of temporal envelope information—
and lead to patterns of neural responses more closely
resembling those found in normal hearing. However,
studies of speech perception as a function of pulse
rate have not provided consistent evidence of these
expected benefits (e.g., Friesen et al, 2005). We have
taken an analytic approach intended to clarify the
effects of stimulation rate on fundamental auditory percepts related to speech and have examined
sinusoidal amplitude modulation (AM) detection,
modulation rate discrimination, and spectral shape
discrimination. Tests were conducted in users of
the CII implant using the CRI-2 research interface
and SPAIDE software. Performance was typically
assessed at stimulation rates of 482, 723, 1447, 2894,
5787 pulses per second (pps).
Higher pulse rates generally led to increases in
dynamic range and, at least in some cases, to more
gradual loudness growth, which might be beneficial in the process of compressing the acoustic
input onto the electrical dynamic range. However,
no consistent advantage for higher pulse rates was
apparent beyond effects that could be attributed to
increased dynamic range. AM detection thresholds
with stimulation of a single electrode in the middle
of the array were obtained from five implant users
at two carrier levels and two modulation rates:
10 Hz (a rate representative of the amplitude fluctuations associated with the dynamic spectral variation
that characterises speech) and 100 Hz (representative of modulation rates associated with voice
pitch). In most cases there was little effect of pulse
rate, though for one subject, thresholds increased
somewhat with stimulation rate. As an example,
Figure 1 shows detection thresholds for 10 Hz
modulation imposed on a carrier level of 75% of the
electrode’s dynamic range.
Similarly, there was little general effect of stimulation rate on the ability to detect changes in the
rate of modulation from a standard modulation
frequency of either 10 or 100 Hz. One subject
showed a beneficial effect of increasing pulse rate
on the discrimination of changes in modulation
rate from the 10 Hz, but not the 100 Hz, standard.
76 Advanced Bionics® Auditory Research Bulletin 2007
Speech perception performance with high stimulation rates may be limited by an increase in channel
interactions, as suggested by physiological data
obtained from implanted guinea pigs (Middlebrooks, 2004). We investigated the effect of stimulation rate on channel interactions using a test of
spectral shape discrimination. This ability is both
directly relevant to speech and dependent upon
the degree of channel interaction, since increased
channel interaction will necessarily lead to smearing
of spectral detail. Stimuli were derived from periodic complex tones containing three spectral peaks
(formants). The task was to discriminate a difference in the frequency of the second formant (F2)
from a standard value of 1500 Hz. Eight-channel,
CIS-like processing applied to the acoustic stimuli
preserved spectral shape information in the form
of across-channel amplitude differences but eliminated within-channel variations of pulse amplitude
over time. Therefore, performance was unaffected by
any influence of pulse rate on the representation of
temporal envelope information.
In an initial experiment, performance was generally
substantially poorer with stimulation at 2,894 pps
than at the other rates tested, across which performance was broadly similar (Figure 2, left-hand
panel). However, it was hypothesised that this may
have reflected variation of the Most Comfortable
Loudness (M) levels across pulse rate and electrodes.
Together with Threshold (T) levels, these values had
been determined using noise-burst stimuli, similar
to present clinical practice. Not only were M levels
with 2,894 pps lower overall than those with 5,787
pps, this difference was particularly pronounced for
the electrodes covering the F2 frequency region.
The experiment was repeated using T and M levels
which were constrained to have the same profile
across electrodes for each rate (though overall levels
differed). The pattern of results across pulse rate was
markedly different in these “fixed-profile” conditions,
and spectral shape discrimination was no longer
substantially poorer with 2,894 pulses per second
(Figure 2, right-hand panel). For two of the three
subjects tested, performance worsened significantly as pulse rate increased, suggesting increased
channel interactions. For the other subject there was
a nonsignificant tendency for better performance as
pulse rate increased.
Overall, these data provide no substantive evidence
for beneficial effects of high pulse rates and suggest
that high pulse rates may have detrimental effects
by increasing channel interactions. It would appear
possible that clinical reports of benefits from higher
rate speech processors may reflect effects of other
concomitant changes to the processing, rather than
stimulation rate per se.
C1
Just Discriminable Value of F2 (Hz)
Taken as a whole, the above results do not support
the hypothesis that higher stimulation rates result
in enhanced perception of temporal envelope information, and they are broadly consistent with other
recent studies that have examined the effects of pulse
rate on AM detection (Galvin & Fu, 2005; Pfingst,
Xu & Thompson, 2007).
2000
C2
C3
Original T and M
“Fixed Profile”
1900
1800
1700
1600
723
1447
2894
5787
1447
2894
5787
Pulse Rate (pps)
Figure 2. Spectral discrimination performance—represented
by the mean value of F2 that could just be discriminated from
the standard value of 1500 Hz as a function of pulse rate. Error
bars show standard deviations. The left-hand panel shows data
obtained with the originally used T and M levels, while the
right-hand panel shows performance with T and M levels
constrained to have the same profile across electrodes for
each rate.
References
Friesen LM, Shannon RV, Cruz RJ. (2005) Effects of stimulation rate on speech
recognition with cochlear implants. Audiol Neurotol 10(3):169-184.
Galvin JJ, Fu QJ. (2005) Effects of stimulation rate, mode and level on modulation
detection by cochlear implant users. J Assoc Res Otolaryngol 6:269-279.
Middlebrooks J. (2004) Effects of cochlear-implant pulse rate and inter-channel
timing on channel interactions and thresholds. J Acoust Soc Am 116:452-468.
Pfingst BE, Xu L, Thompson CS. (2007) Effects of carrier pulse rate and stimulation
site on modulation detection by subjects with cochlear implants. J Acoust Soc Am
121:2236-2246.
Psychophysics, Modeling & Novel Processing 77
Electrode Discrimination in Noise
in Cochlear Implant Recipients
Marc D. Eisen, M.D., Ph.D.1
Jonathan Kopelovich, M.D.2
Matthew Kohler, B.S.3
Leonid Litvak, Ph.D.4
Kevin H. Franck, Ph.D.3,5
Understanding how electrical stimuli are perceived
is an important step towards maximizing the benefit
obtained from a cochlear implant. One basic principle of cochlear implant engineering is that spatial
separation along the multielectrode array is the basis
for frequency discrimination. This tenet relies on
the assumption that stimulation of spatially separated electrodes is perceptually discernible. With the
introduction of HiRes with Fidelity 120 and current
steering, it is now possible hypothetically to stimulate
sites between the physical electrode contacts. As the
spatial density of stimulation increases beyond the
number of physical electrodes, the limit of perceptible differences as a function of stimulation location
likely will be revealed. Understanding these limits of
perception in electrical hearing is important for the
development of sound processing schemes.
1 Johns Hopkins University, Baltimore, MD, USA
2 University of Iowa, Iowa City, IA, USA
3 The Children’s Hospital of Philadelphia, PA, USA
4 Advanced Bionics, Valencia, CA, USA
5 University of Pennsylvania, Philadelphia, PA, USA
“. . .these data showed that
signal-to-noise ratio
is an important cue
for electrode discrimination.”
3
DL (Electrode)
2.5
2
1.5
1
0.5
0
-30
-20
-10
0
10
20
Signal to Noise Ratio (dB)
Figure 1. Interelectrode difference limens (DL) for seven
subjects as a function of signal-to-noise ratio.
30
Our research has sought to quantify the minimum
discernible distance between electrodes in pediatric
implant recipients and to examine how this distance
is affected by sound intensity and signal-to-noise
ratio in both pediatric and adult subjects. Relatively
few basic psycho-electroacoustic studies have been
conducted in children because of the inherently
monotonous nature of psychophysical tasks. In order
to address this challenge, we have developed an interactive and captivating videogame-based platform
that interfaces with the cochlear implant processor.
This game-based software has enabled us to engage
pediatric cochlear implant users in psychophysical
experiments aimed at determining the minimum
interelectrode distance that can be discriminated.
In this study, subjects were children seven years of
age or older who had been implanted with a CII or
HiRes 90K device at either The Children’s Hospital
of Philadelphia or The Hospital of the University of
Pennsylvania. Electrode discrimination was defined
as the minimum discernible interelectrode distance
78 Advanced Bionics® Auditory Research Bulletin 2007
Recent Public at ion
Use of S-Shaped Input-Output Functions
for Noise Suppression in Cochlear Implants
Kalyan Kasturi, Ph.D.
Philipos Loizou, Ph.D.
University of Texas, Dallas, TX, USA
This study evaluated the influence of the shape of
(denoted as difference limen, or DL) that could be
perceived. The DLs were determined using a twointerval forced-choice task with two flanking control
stimuli. Multiple tasks were strung together as a run,
where all stimulus parameters were held constant
except for one, which was varied according to an
adaptive paradigm. With this paradigm, a run would
converge upon the interelectrode distance DL.
the acoustic-to-electric mapping function on speech
Results indicated that older children tended to have
better electrode discrimination than younger children. Moreover, children were able to discriminate
smaller interelectrode distances when the intensity
of the sound was increased. In fact, the majority of
subjects could discern interelectrode distances of
less than the separation between electrodes on a
standard array. The question then was whether the
improved electrode discrimination with increased
stimulus intensity was a result of the improved
signal-to-noise ratio or the absolute intensity of the
stimulus. Therefore, background electrical noise was
added to the stimuli. In general the data indicated
that lowering the signal-to-noise ratio resulted in
worse electrode discrimination—independent of the
absolute stimulus intensity.
for all subjects when compared to the conventional log
recognition in noise for nine experienced CII implant
users who used HiRes sound processing. Subjects were
fit with a new s-shaped mapping function that was
expansive for low input levels up to a knee-point level
and compressive thereafter. The knee point changed
dynamically and was set proportional to the estimated
noise floor level. The s-shaped function did not yield
higher performance on sentence perception in noise
mapping function used in typical programs. However,
significantly higher sentence scores in noise were
obtained when the s-shaped functions were optimized
for each user. The results indicate that the shape of the
nonlinear mapping function can have a significant effect
on speech intelligibility in noise. Optimizing the shape
of the mapping function—if implemented in combination
with the wide input dynamic range and temporal/
spectral resolution offered by HiRes sound processing
implemented with the Harmony processor—may serve to
enhance the spectral contrasts required for understanding
speech in compromised listening environments.
Ear Hear (2007) 28(3):402-411.
Figure 1 shows DLs as a function of signal-to-noise
ratio for seven different subjects when absolute stimulus intensity was held constant. Four of the seven
subjects were able to discriminate between substantially less than the distance between electrodes, even
when the noise was greater than the signal. Five of
the seven subjects were able to discriminate between
less than the interelectrode distance when the
signal-to-noise ratio was increased. Thus, these data
showed that signal-to-noise ratio is an important
cue for electrode discrimination. Future experiments
will focus on determining how electrode discrimination limits affect the discrimination of more complex
stimuli such as speech and music.
Psychophysics, Modeling & Novel Processing 79
Spectral Peak Resolution Assessment of
Cochlear Function
Annelies Vermeiren, M.Aud.1
Andrzej Zarowski, M.D.1
Erwin Offeciers, M.D., Ph.D.1
Filiep Vanpoucke, Dr.Ir.2
Stefaan Peeters, Prof., Dr.Ir.3
1 Medical Center Sint Augustinus, Wilrijk, Belgium
2 Advanced Bionics European Research Center,
Antwerp, Belgium
3 University of Antwerp, Antwerp, Belgium
“. . .this procedure is potentially applicable
to a younger population
whose linguistic skills
may not yet be fully developed.”
Hearing aids and cochlear implants aim to restore
the natural temporal, spectral, and intensity cues
produced by a functional cochlea. However, the
benefit provided by the devices is dependent upon
the auditory experience and language skills of the
listener. That is, speech perception and music appreciation tests may not only reflect the efficacy of the
device but also the maturity of the individual’s brain.
Therefore, speech testing may not be the best way
to evaluate device efficacy in children or adults with
limited auditory experience.
In 2005, Henry and colleagues reported a method
for evaluating the cochlea as a spectral analyzer using
a test that was language independent. Their spectral
ripple task proved a viable method for assessing
spectral discrimination ability, and the results were
highly correlated with vowel and consonant recognition. This study assessed the clinical applicability
of their task by collecting normative data using a
similar spectral ripple procedure and relating those
data to frequency resolution and speech perception.
Participants in the study included 18 adult implant
users and 17 hearing aid users. A variety of devices
and processing strategies were represented. Wideband noise signals were spectrally modulated with
sine or cosine waves so that spectral peaks occurred
80 at different frequencies. The subject’s task was
to discriminate between the sine- and cosinemodulated noises while the number of ripples was
gradually increased. The discrimination (spectral
ripple) threshold was determined using a 3AFC
adaptive procedure where the subject selected the
interval in which the signal was different (one sinemodulated signal, two cosine-modulated signals).
The spectral resolution threshold was defined as the
greatest number of ripples the subject could discriminate. In addition, the acoustic frequency difference
limen was measured for reference frequencies 500
Hz, 1,000 Hz, 2,000 Hz and 4,000 Hz. In contrast
to the spectral ripple tasks, these measurements evaluated spectral resolution at different locations along
the cochlea. Speech performance was assessed with
phoneme discrimination, CVC words, and a vowel
and consonant identification test.
The adult subjects were able to perform the spectral
ripple task. The spectral ripple thresholds averaged
2.7 ripples per octave for the hearing aid users and
1.5 ripples per octave for the implant users. The
highest correlation was found between the spectral
ripple threshold and threshold for words presented
in noise, although all correlations were smaller than
those reported by Henry and colleagues (2005). This
correlation suggested that frequency resolution is
important for hearing speech in noise. Based upon
the ease of testing, this procedure is potentially
applicable to a younger population whose linguistic
skills may not yet be fully developed.
Reference
Henry BA, Turner CW, Behrens A. (2005) Spectral peak resolution and speech
recognition in quiet: normal hearing, hearing impaired and cochlear implant
listeners. J Acoust Soc Am 118(2):1111-1121.
Advanced Bionics® Auditory Research Bulletin 2007
Recent Public at ions
Loudness Growth Obtained Under Partially
Tripolar Stimulation: Model and Data from
Cochlear Implant Users
Asymmetric Pulses in Cochlear Implants:
Effects of Pulse Shape, Polarity, and Rate
Olivier Machery, Ph.D.1
Astrid van Wieringen, Ph.D.1
Robert P. Carlyon, Ph.D.2
John M. Deeks, Ph.D.2
Jan Woulters, Ph.D.1
Leonid Litvak, Ph.D.1
Anthony J. Spahr, Ph.D.2
Gulam Emadi, Ph.D.1
1 Catholic University of Leuven, Leuven, Belgium
2 Arizona State University, Tempe, AZ, USA
1 Advanced Bionics, Valencia, CA, USA
2 University of Cambridge, Cambridge, United Kingdom
Cochlear implants typically use monopolar stimulation
that likely stimulates a broad population of auditory nerve
Contemporary cochlear implants stimulate the auditory
fibers. The spread of excitation may be reduced through
nerve with symmetric biphasic pulses. Recent data suggest
a tripolar stimulation scheme in which compensating
that modifying the pulse shape while maintaining the charge
current of opposite polarity is delivered simultaneously to
balance may reduce power consumption, increase the
two electrodes flanking a center electrode. In this study,
dynamic range, and limit channel interactions. Thresholds and
loudness growth was first modeled and then measured in
most comfortable levels were measured in 12 adult CII and
seven CII and HiRes 90K recipients for varying amounts
HiRes 90K recipients using a “delayed pseudomonophasic”
of compensating current, parameterized by a coefficient σ
(DPS) pulsatile stimulus in which a longer phase of one
that ranged from 0 (monopolar stimulation) to 1 (full tripolar
polarity is presented midway between two short phases of
stimulation). For both the model and the subjects, the
opposite polarity. DPS pulse trains produced thresholds and M
amount of current required for threshold could be described
levels that were more than 10 dB lower than biphasic pulses.
with the equation I(σ) = I0/(1 + σK) where I0 and K were
The reduction in current required for hearing was predicted
fitted constants. Above threshold, equal-loudness contours
by a simple linear model. DPS stimulation might lead to
for some subjects deviated significantly from linear scaled-
considerable power savings if implemented in a wearable
up values of threshold current approximations. The patterns
sound processing strategy. The potential to implement DPS
of deviation were similar to those observed in the model
and other innovative stimulation waveforms is made possible
for conditions in which most of the neurons near the center
by the independent programmable current sources unique to
electrode were excited. The model and data suggest that it is
the CII/HiRes 90K system architecture.
possible to focus neural stimulation using a tripolar scheme.
J Assoc Res Otolaryngol (2006) 7:253-266.
Tripolar stimulation potentially can be implemented in all
CII and HiRes 90K implants because each electrode contact
is powered by a separate programmable current source, a
unique feature of the CII/HiRes 90K system architecture.
J Acoust Soc Am (2007) 122(2):967-981.
Psychophysics, Modeling & Novel Processing 81
The HRStream Research Interface
Peter Bracke, M.Eng.1
Filiep Vanpoucke, Dr.Ir.1
Salomo Murotonen, B.S.2
Leonid Litvak, Ph.D.2
1 Advanced Bionics European Research Center,
Antwerp, Belgium
2 Advanced Bionics, Valencia, CA, USA
“This interface provides the opportunity
to evaluate the benefits of
various stimulation schemes
in actual cochlear implant users.”
The stimulation capabilities of the CII and HiRes
90K implants extend far beyond the standard clinical configuration of monopolar sequential biphasic
pulses. The implants’ 16 independent current sources
provide the opportunity for controlled multipolar
stimulation that can be used to shape the electrical
field. The implants also provide flexibility in the
choice of the pulse shape so that asymmetric or
triphasic pulses, as examples, can be delivered to the
electrode contacts. Some of these parameters already
are being explored (such as bipolar and tripolar
stimulation as well as current steering).
To explore the advanced capabilities of the CII and
HiRes 90K systems, a new open research platform
(HRStream™) has been developed to give investigators control over stimulation parameters and the
ability to test new sound processing algorithms.
This interface provides the opportunity to evaluate
the benefits of various stimulation schemes in actual
cochlear implant users. It allows investigators to
define stimuli on a PC (for example, in Matlab) and
to stream the stimuli to the implant in real time.
Psychoacoustic experiments as well as acute testing
of sound-processing prototypes can be conducted.
Specifically, the hardware interface consists of a
small DSP board that is connected to a PC via a
USB port and to a PSP sound processor through a
modified programming cable. The board is powered
through the USB port. This hardware is much less
bulky and easier to use than the previous generation
CRI-2 system. The streaming software essentially
consists of a relatively low-level software (dll). The
Application Programmer’s Interface (API) has been
designed for flexibility and reliability, and provides a
strategy builder and a stream player. The functions
in the strategy builder allow the user to configure a
stimulation strategy by specifying the temporal and
spatial properties of the stimulation channels and
their relative timing. During strategy compilation,
the strategy builder generates the required binary
objects (for example, the pulse table) to be sent to
the implant based on user specifications. The stream
player provides a software interface to deliver arbitrary stimuli. The input stimuli are prepared and
stored in multichannel .wav files in which values
represent stimulation strength in μA. The software
has a NET interface, a COM, and a C++ interface
so that the system can connect with other applications such as Matlab.
In terms of reliability, the software includes hardware management and a bidirectional status channel.
Communication to the application is based on an
event mechanism. A rich set of events is provided
indicating hardware changes, the actual start and end
of a stimulus, and communication status. Synchronization errors, such as underflow or overflow also are
detected. HRStream will be available in the future.
82 Advanced Bionics® Auditory Research Bulletin 2007
Recent Public at ions
Current Steering Creates Additional Pitch
Percepts in Adult Cochlear Implant Recipients
Using Current Steering to Increase Spectral
Resolution in CII and HiRes 90K Users
Jill B. Firszt, Ph.D.1
Dawn Burton Koch, Ph.D.2
Mark Downing, Ph.D.2
Leonid Litvak, Ph.D.2
Dawn Burton Koch, Ph.D.
Mark Downing, Ph.D.
Mary Joe Osberger, Ph.D.
Leonid Litvak, Ph.D.
1 Washington University, St. Louis, MO, USA
Advanced Bionics, Valencia, CA, USA
2 Advanced Bionics, Valencia, CA, USA
This study investigated the number of spectral channels
This study determined whether cochlear implant users could
(or different pitches) that could be resolved by adult users
hear additional spectral channels (pitches) by using active
of the CII and HiRes 90K cochlear implants when active
current steering. Active current steering involves simultaneous
current steering was applied to three pairs of electrodes
delivery of current to adjacent electrodes where stimulation
along the implanted array. Active current steering delivers
can be “steered” to sites between the contacts by varying the
current simultaneously to pairs of adjacent electrodes—
proportion of current delivered to each electrode of a pair.
hypothetically changing the effective locus of stimulation
Subjects were 106 postlinguistically deafened adults who use
between the contacts by varying the proportion of current
the CII or HiRes 90K device. After loudness balancing and
delivered to each electrode of the pair. Subjects first loudness
pitch ranking three electrode pairs (2-3, 8-9, 13-14), subjects
balanced and pitch ranked electrode pairs (2-3, 8-9, 13-14)
identified the electrode with the higher pitch while current
after which an adaptive paradigm was used to estimate the
was varied proportionally between electrodes in each pair.
number of intermediate pitch percepts that could be heard
The smallest change in proportion yielding a discriminable
for each pair when current steering was implemented. Those
change in pitch was defined as the spectral resolution. Data
data were used to estimate the potential number of spectral
from 115 ears indicate that the number of spectral channels
channels for each electrode pair. Data from 57 implanted
averaged 3.8 for the basal pair, 6.0 for the mid-array pair, and
ears indicated that the numbers of spectral channels per
5.3 for the apical pair. Assuming the number of channels on
electrode pair ranged from one (subjects who could not
these three electrode pairs represents the entire array, the total
tell the electrodes apart) to 52 (an individual who had 52
potential number of spectral channels was calculated and
different pitch percepts for the mid-array pair of electrodes).
ranged from 8 to 451, with an average of 63. There were no
The average number of spectral channels that could be
obvious relationships between age at implantation, duration
distinguished were 5.4 for the basal electrode pair, 8.7 for the
of cochlear implant use, or length of auditory deprivation
mid-array electrode pair, and 7.2 for the apical electrode pair.
and the number of spectral channels in this group. Similarly,
Assuming that the average number of spectral channels for
there were no relationships between two measures of speech
each of these three electrode pairs were representative of the
recognition (CNC words and HINT sentences with a +8 dB
entire 16-contact array, the potential total number of spectral
signal-to-noise ratio) obtained with the subjects’ current
channels could be estimated. For the 57 ears, the number of
clinical strategy (standard HiRes) and the number of spectral
potential channels ranged from 8 to 466, with an average of
channels measured psychophysically. The results indicate
93. The results suggested that the average cochlear implant
that adult cochlear implant recipients are able to perceive
user may have significantly more place-pitch capability than
additional pitch information when active current steering
is exploited presently by cochlear implant systems.
is implemented. The large intra- and intersubject variability
in the perception of intermediate pitches across electrodes
Ear Hear (2007) 28(Suppl 2):38-41.
suggests that multiple factors underlie the ability to resolve
spectral information. Active current steering has been
implemented in the wearable HiRes with Fidelity 120 sound
processing strategy.
Otol and Neurotol (2007) 28(5):629-636.
Psychophysics, Modeling & Novel Processing 83
A Novel Signal Processing Strategy for Current Steering
in Cochlear Implants
Waldo Nogueira, M.Sc.1
Martina Brendel, M.Sc.2
Carolin Frohne-Büchner, Ph.D.2*
Thomas Lenarz, Prof., M.D., Ph.D.2
Bernd Edler, Ph.D.1
Andreas Büchner, Ph.D.2
1 Leibniz University of Hannover, Hannover, Germany
2 Medizinische Hoschschule Hannover, Hannover, Germany
* also with Advanced Bionics, Europe
The main goal of this project is to improve speech
and music perception of cochlear implant users
through an improved signal processing strategy.
Today’s cochlear implant strategies like HiResolution
Sound (HiRes) already provide increased temporal
information through high rates of stimulation. The
challenge is to develop new signal processing strategies that maximize the representation of frequency
information given that only a limited number of
electrodes can be inserted into the cochlea.
In the late 1980s, researchers reported that additional pitch percepts could be created by stimulating
two adjacent electrodes simultaneously (Townshend et al, 1987; Wilson et al, 1994). Simultaneous
stimulation presumably causes a summation of the
electrical fields, thereby creating a stimulation peak
located between the two electrodes. Systematic
adjustment of the proportion of current delivered
simultaneously to two electrodes can elicit multiple
pitch percepts (Donaldson et al, 2005; Koch et al,
2007). This technique, termed “current steering,”
has the potential to improve spectral resolution in
cochlear implant users. Current steering has been
implemented in the HiRes with Fidelity 120 (HiRes
120) sound processing option.
Because implants like the CII and HiRes 90K
potentially can use current steering to more accurately deliver spectral information, accurate spectral
analysis of the audio signal becomes more critical.
Therefore, this project developed a novel processing
algorithm that combines enhanced audio signal
analysis with current steering. The algorithm is based
on a sinusoidal representation of the incoming audio
signals. Sinusoidal modeling has been used successfully in speech analysis and synthesis (McAulay &
84 Quatieri, 1986) as well as for speech (Nishiguchi
& Edler, 2002) and audio coding (Herre & Purnhagen, 2002). This technique in theory provides both
the fine frequency resolution necessary for current
steering and the high temporal resolution necessary for temporal pitch perception. Furthermore,
the selection of the sinusoidal components can be
performed using different criteria (Purnhagen et al,
2002). By combining accurate modeling of the input
signal with appropriately selected current-steered
electrode pairs, it may be possible to improve hearing
performance in cochlear implant users.
The new strategy is termed the sinusoid extraction
strategy, or SineEx. It was built around a research
version of the HiRes 120 strategy, and therefore preserves several elements of the HiRes 120
algorithm. Figure 1 shows a block diagram of the
SineEx strategy. First, the signal picked up by the
microphone is digitized at 17400 samples per second.
Next, the same adaptive-gain-control (AGC) used
in HiRes 120 is implemented. The resulting signal is
sent to a fast Fourier transform (FFT). The linearlyspaced bands of the FFT then are grouped into analysis channels. An analysis channel is defined as the
spectral band that will be delivered between two adjacent electrodes. For each analysis channel, an envelope estimation is made. In order to improve spectral
analysis of audio signals, an analysis-by-synthesis
algorithm is used to model the spectrum with the
most relevant sinusoids for human perception. This
algorithm is based on a matching-pursuit algorithm
combined with an optimal selection of the sinusoidal
components. The outputs of the analysis/synthesis
algorithm are the frequencies of the estimated sinusoids. These frequencies are used both to determine
the place of stimulation and to synthesize the carrier
pulse train. The place of stimulation is determined
by a frequency-weight map, which converts the
frequencies into the proportions of current that will
be delivered to the two electrodes assigned to each
analysis channel. By using the appropriate proportions of current, the frequency of the estimated
sinusoid can be accurately coded through place of
stimulation. The carrier pulse train is synthesized so
that the pulse rate is equal to the estimated sinusoid.
Advanced Bionics® Auditory Research Bulletin 2007
acoustic signal solely with pure sinusoids. These
preliminary findings demonstrate that sinusoidal
coding combined with current steering has the
potential to more accurately represent the sounds
of everyday life and may lead to greater clinical
benefits.
The carrier then modulates the envelope over the
time domain. Thus, the carrier codes the frequency of
the estimated sinusoid through temporal coding. The
electrode pairs for analysis channels not containing a
sinusoid are not stimulated. Finally, the information
from the envelope detector, the carrier synthesis, and
the frequency-weight map is combined with the
subject’s individual program parameters and sent
to the corresponding electrodes. In both SineEx
and HiRes 120, pairs of electrodes assigned to the
same analysis channel are stimulated simultaneously.
However, electrode pairs assigned to different analysis channels are stimulated sequentially to avoid
electrical field overlap.
References
Donaldson GS, Kreft HA, Litvak L. (2005) Place-pitch discrimination of single- versus
dual-electrode stimuli by cochlear implant users. J Acoust Soc Am 118(2):623-626.
Herre J, Purnhagen H. (2002) General Audio Coding. In: Pereira F, Ebrahimi T, eds.
The MPEG-4 Book, New York: Prentice Hall, 530–539.
Koch DB, Downing M, Osberger MJ, Litvak L. (2007) Using current steering to
increase spectral resolution in CII and HiRes 90K users. Ear Hear 28(2):38S-41S.
McAulay RJ, Quatieri TF. (1986) Speech analysis/synthesis based on a sinusoidal
representation. IEEE Transactions on Acoustics, Speech and Signal Processing
34(4):744-754.
Nishiguchi M, Edler B. (2002) Speech Coding. In: Pereira F, Ebrahimi T, eds. The
MPEG-4 Book, New York: Prentice Hall, 461-475.
Details of the study design and clinical results are
presented in a companion summary, which follows
(Büchner et al, at page 86 of this bulletin). To
study speech perception performance, the experimental SineEx strategy was implemented in the
commercially available body-worn sound processor
(PSP). All study participants were able to use the
experimental strategy, which models the incoming
Purnhagen H, Meine N, Edler B. (2002) Sinusoidal coding using loudness-based
component selection. Proceedings of the IEEE International Conference on Acoustics,
Speech and Signal Processing, 13–17 May, 2002, Orlando, Florida, USA.
Townshend B, Cotter N, van Compernolle D, White RL. (1987) Pitch perception by
cochlear implant subjects. J Acoust Soc Am 82:106-115.
Wilson BS, Lawson DT, Zerbi M, Finley CC. (1994) Recent developments with the
CIS strategies. In Hochmair-Desoyer I, Hochmair E, eds. Advances in Cochlear
Implants. Proceedings of the Third International Cochlear Implant Conference, April,
1993, Innsbruck, Austria. Vienna, Austria: Datenkonvertierung, Reproduction and
Druck, 103-112.
Analysis Channel 1
Ts
1
2
Digital
Audio
Grouping
Envelope
Detection
E1
Nonlinear
Map
E2
Analysis Channel 2
Front
End
L-Fast
Fourier
Transform
3
4
5
6
Grouping
Envelope
Detection
E2
Nonlinear
Map
E3
Analysis Channel M
Grouping
L/2
Envelope
Detection
Nonlinear
Map
W
EM-1
EM
1-W
Analysis/
Synthesis
Sinusoidal
Model
Frequency
Weight
Map
Carrier
Synthesis
Figure 1. Block diagram of the SineEx coding strategy.
Psychophysics, Modeling & Novel Processing 85
Comparison of Sound Processing Strategies:
FFT vs. Sinusoidal Approximation (SinEx)
Implemented with Current Steering
Andreas Büchner, Ph.D.1
Martina Brendel, M.Sc.1
Waldo Nogueira, M.Sc.2
Carolin Frohne-Büchner, Ph.D.1*
Bernd Edler, Ph.D.2
Thomas Lenarz, Prof., M.D., Ph.D.1
1 Medizinische Hochschule Hannover, Hannover, Germany
2 University of Hannover, Laboratory of Information
Technology, Hannover, Germany
* also with Advanced Bionics, Europe
Even though the temporal resolution of modern
cochlear implant systems is acceptable, spectral resolution still is not optimal. Until the advent of HiRes
with Fidelity 120 (HiRes 120), the number of
stimulation sites and the number of spectral channels have been limited to the number of physical
electrode contacts. Recently, studies have shown
that it is possible to increase the number of distinct
pitch percepts by sharing the current between adjacent contacts, termed “current steering” (Koch et al,
2007). This study evaluated three different acoustic
signal processing algorithms that were implemented
with the “current steering” technique, two using Fast
Fourier Transform (FFT) analyses and one using
sinusoidal approximation.
SpecRes was an early experimental version of an
FFT current steering strategy that could be implemented on the CII and HiRes 90K devices. In
SpecRes, the acoustic signal is sent through an FFT.
Then the FFT is decomposed into 15 frequency
analysis bands, corresponding to the 15 pairs of electrode contacts in the implanted array. Within each
of the 15 bands, the location of stimulation (that is,
the current distribution between the two adjacent
contacts) is defined as the frequency of the highest
spectral peak in each frequency band. Thus, in each
analysis cycle, 15 stimuli are presented, one in each
band, but with varying focal points. The stimulation
amplitude is determined by the acoustic energy in
each analysis band.
86 This study assessed the SpecRes strategy using
two different resolutions for the stimulation sites
(1) the commercially implemented HiRes 120 with
120 potential stimulation sites and (2) SpecRes
16K with 16,000 potential stimulation sites along
the electrode array. This study also evaluated a third
implementation of current steering using a sinusoid
extraction algorithm. This strategy, called SinEx, was
a cooperative venture between the ENT Department of the Medizinische Hochschule Hannover,
the Laboratory of Information Technology of the
University of Hannover, and Advanced Bionics. The
Sinex algorithm is adapted from the audio coding
used in MP4 signal processing. (See also Nogueira
et al, at page 84 of this bulletin.)
All three strategies were evaluated in eight postlinguistically deafened adults who had at least 12
months of experience with standard HiRes sound
processing. Because the study required a body-worn
PSP processor, subjects who normally used a BTE
processor were converted to the PSP prior to the
study, which they used for one month with their
own clinical HiRes programs. Prior to fitting of an
experimental strategy, subjects were tested with their
HiRes programs using the PSP. The speech perception test battery consisted of the HSM sentence test
in quiet and with various competing noise signals.
Subjects also were administered a sound quality
questionnaire to assess subjective listening benefits.
Following baseline testing, each subject was fitted
randomly with one of the three experimental strategies. After one month of use, each subject was
retested and then fit with another of the three strategies. The study protocol consisted of two cycles with
an ABCBCA strategy order. Clinical results for each
strategy were averaged between the two cycles.
Advanced Bionics® Auditory Research Bulletin 2007
HiRes
100
98.8 99.2
HiRes 120
SpecRes 16K
SinEx
94.9 95.6
90
Percent Correct
80
All subjects adapted easily to all of the new strategies. No significant changes were required in the
program parameters compared to each subject’s
own HiRes program. Even though subjects reported
differences in sound quality, neither the speech
perception nor questionnaire results showed any
significant differences among the three experimental
strategies (Figures 1 and 2). Nonetheless, when the
best current steering program was compared to standard HiRes, a highly significant advantage for the
current steering strategies was clear (Figure 3).
74.2
70
76.6 74.8
65.4
60
50
40
29.7 31.0
30
35.4
26.0
20
10
0
HSM in Quiet
HSM in CCITT Noise
(10 dB SNR)
HSM with CT
(5 dB SNR)
Figure 1. Mean results for the HSM sentence test in quiet and
with competing noise for standard HiRes and the three experimental current-steering strategies.
HiRes
HiRes 120
SpecRes 16K
SinEx
10
Mean Ratings
8
The data in this small group of subjects did not show
significant benefit for any one of the three experimental acoustic signal processing algorithms. Therefore, even though current steering can increase the
number of stimulation sites within the cochlea, the
optimal processing of the acoustic input remains to
be determined. Nonetheless, a significant benefit was
evident when the current-steering strategy yielding
each subject’s best performance was compared to
baseline standard HiRes performance. Future studies
will explore how signal-processing variables can be
optimized for individuals using programs based on
current steering.
6
4
2
0
Overall
Clarity
Pleasantness
Background
Interference
Figure 2. Mean sound quality ratings for standard HiRes and
the three experimental current steering strategies.
HiRes
Best CS
80
70
Koch DB, Downing M, Osberger MJ, Litvak L. (2007) Using current steering to
increase spectral resolution in CII and HiRes90K users. Ear Hear 28:38S-41S.
60
Percent Correct
Reference
50
37.8**
40
29.7
30
20
10
0
P1
P2
P3
P4
P5
P6
P7
P8
AVG
Figure 3. Individual HSM sentence test results in quiet and with
competing noise for each subject’s best performance with a current steering program (“best CS”) and for standard HiRes.
(p < .05)
Psychophysics, Modeling & Novel Processing 87
Current Steering and Current Focusing in
Cochlear Implants: Comparison of Monopolar, Tripolar,
and Virtual Channel Electrode Configurations
Several studies have shown that a monopolar electrode configuration is less than optimal in terms
of channel selectivity. A possible approach to
improving channel selectivity is “current focusing.”
A tripolar (also known as quadrapolar) electrode
configuration results in a more focused current flow
(Figure 1). However, previously we found that to
provide sufficient loudness growth using a tripolar
configuration, it was necessary to return part of
the current to the extracochlear reference electrode (Mens & Berenstein, 2005). The fraction of
current returning to the extracochlear reference is
defined as the remote current fraction (RCF). Direct
measurement (through electrical field imaging) of
the tripolar configuration’s longitudinal electrical
fields indeed confirmed a more confined current
spread compared to monopolar electrode coupling.
However, speech perception in a group of 10 subjects
did not improve significantly with the tripolar-based
strategy, although a tripolar configuration with an
RCF of 50% resulted in a slight improvement of
speech perception in noise.
Carlo Berenstein, B.Sc.1
Lucas Mens, Ph.D.1
Jef Mulder, M.D., Ph.D.1
Filiep Vanpoucke, Ph.D.2
1Radboud University Nijmegen Medical Centre,
Nijmegen, the Netherlands
2Advanced Bionics, Europe
5
6
7
8
Monopole
9
5
6
7
Tripole
8
9
5
6
7
8
Virtual Channel
9
Figure 1. Current configuration (gray) and putative peak of the electrical field (black) for intracochlear electrical fields from monopole (A), tripole (B), and virtual channel (C) electrode configurations. The gray arrow represents the magnitude of the injected
current; the black arrow represents peak position of the corresponding electrical field. In B, each of the return electrodes returns
half of the centrally injected current minus the remote current fraction (RCF) (not shown), which is returned to the extracochlear
reference. The RCF varied between 0% and 100%. Note that a monopolar configuration effectively is the same as a “Q100” configuration. In C, Electrode 7 is stimulated simultaneously but with higher amplitude than Electrode 8. This causes the resulting field
distribution to be shifted slightly toward the right of Electrode 7. The virtual channel strategy created seven channels between each
pair of contacts.
88 Advanced Bionics® Auditory Research Bulletin 2007
As expected from a current focusing strategy, the
behavioral T and M levels were significantly higher
with the tripolar than with either the monopolar or
virtual channel configurations. In other words, the
tripolar configuration was less energy-efficient.
Although the tripolar and virtual channel configuration both resulted in a slightly better spectral
resolution across subjects than the monopolar
configuration, overall spectral resolution remained
extremely poor compared to normal hearing listeners
(Henry et al, 2005). Spectral resolution was significantly higher with the tripolar than with the monopolar configuration. Spectral resolution with the
virtual channel configuration did not differ significantly from the other configurations (Figure 2).
“. . .greater understanding of the effects
of electrical field manipulations
on loudness and channel discrimination
is needed.”
Monopolar
2
Number of Ripples per Octave
In this study (Berenstein et al, in press), we employed
a crossover design to compare the effects of monopolar, tripolar, and “virtual channel” electrode configurations on spectral resolution and speech perception.
Nine experienced adults who received an Advanced
Bionics CII or HiRes 90K cochlear implant used
each of three experimental strategies for two weeks.
Each strategy was a variant of the standard HiRes
processing strategy, using 14 channels and 1105
pulses per second per channel and a pulse duration
of 32 μsec/phase. Spectral resolution was measured
using broadband noise with a sinusoidally rippled
spectral envelope and with peaks evenly spaced on
a logarithmic frequency scale (Supin et al, 1994).
Speech perception was measured for monosyllables
in quiet as well as in steady-state and fluctuating
noise. Subjects were provided with either a RCF of
25% or 75%, depending on which of the two RCFs
resulted in the better spectral resolution test score.
Tripolar
Virtual Channel
1.6
1.2
0.8
0.4
0
1
2
3
4
5
6
7
8
9
Mean
Subject
Figure 2. Spectral ripple resolution per subject for three electrode configurations. The error bars are ± one SD. Subjects are
ordered according to their phoneme scores with the monopolar
configuration (poorest on the left). The mean spectral ripple
resolution is shown on the right side of the graph, with error
bars representing the mean of the individual SD.
Moderate but significant correlations between word
recognition and spectral resolution were found for
speech perception in quiet as well as in fluctuating
noise. Even after individual optimization of the
—continued on next page—
Psychophysics, Modeling & Novel Processing 89
—continued from previous page—
RCF (25% or 75%), no clear advantage in speech
recognition was found across subjects for the tripolar
configuration compared to the monopolar configuration (Figure 3). Similarly, the virtual channel
configuration did not lead to an increase in average
speech understanding. For speech in quiet, word
recognition was best with the monopolar and worst
with the virtual channel configuration. (The tripolar
configuration did not significantly differ from the
Monopolar
100
Percent Correct
.7
75
41
.7
41
Tripolar
Virtual Channel
.9
36
.1
14
50
.6
13
.4
11
.9
12
.1
15
.4
12
25
0
Speech in Quiet
Stationary Noise
0 dB SNR
Fluctuating Noise
0 dB SNR
Figure 3. Monosyllabic word recognition calculated as the
percent of phonemes correct ± one SD across subjects for each
electrode coupling tested in quiet, in stationary noise at 0 dB
SNR, and in fluctuating noise at 0 dB SNR. The numbers above
the bars represent the corresponding mean word scores
(percent correct).
other two.) Pooled across the noise conditions, word
recognition was best with the tripolar and worst with
the virtual channel configuration. (The monopolar
configuration did not significantly differ from the
other configurations.) Although individual results
per strategy did differ, specifically for speech perception in stationary noise, no significant increase in
performance across subjects was found when the best
configuration score for each subject was compared to
results with monopolar coupling.
Subjective comments regarding experience with
music and preference in everyday use were assessed
through questionnaires. Reports showed quite a
large variance. Across subjects, there were no large
differences or clear preferences between strategies.
However, one subject definitely preferred the virtual
channel processor for listening to music.
We conclude that, although spectral resolution was
improved with the tripolar configuration, differences in speech performance were too small (at least
in this limited group of subjects) to justify clinical
application. In addition, overall spectral resolution remained extremely poor compared to people
with normal hearing. It remains to be seen whether
further manipulations of the electrical field will
be more effective. Possibly, individual differences
explain the varying outcomes and a further optimization of the strategies would be beneficial. To do so,
greater understanding of the effects of electrical field
manipulations on loudness and channel discrimination is needed. A more in-depth analysis of the
intracochlear electrical fields and psychophysical
tests may shed further light on this issue.
References
Berenstein CK, Mens LHM, Mulder JM, Vanpoucke FV. (2007) Current steering and
current focusing in cochlear implants: comparison of monopolar, tripolar, and virtual
channel electrode configurations. Ear Hear, in press.
Henry BA, Turner CW, Behrens A. (2005) Spectral resolution and speech recognition
in quiet: Normal hearing, hearing impaired, and cochlear implant listeners. J Acoust
Soc Am 118:1111-1121.
Mens LHM, Berenstein CK. (2005) Speech perception with mono- and quadrupolar
electrode configurations: a crossover study. Otol Neurotol 26:957-964.
Supin A, Popov VV, Milekhina, ON, Tarakanov MB. (1994) Frequency resolving
power measured by rippled noise. Hear Res 78:31-40.
90 Advanced Bionics® Auditory Research Bulletin 2007
Recent Public at ions
The Effect of Gaussian Noise on the
Threshold, Dynamic Range, and Loudness of
Analogue Cochlear Implant Stimuli
Relationship between Perception of Spectral
Ripple and Speech Recognition in Cochlear
Implant and Vocoder Listeners
Robert P. Morse, Ph.D.1
Peter F. Morse
Terry B. Nunn, M.Sc.3
Karen Archer3
Patrick Boyle, M.Sc.4
Leonid Litvak, Ph.D.1
Anthony J. Spahr, Ph.D.2
Aniket Saoji, Ph.D.1
Gene Y. Fridman, Ph.D.3
1 Keele University, Keele, United Kingdom
2 Arizona State University, Tempe, AZ, USA
2 The Open University in the South, Oxford, United Kingdom
3 University of California, Los Angeles, CA, USA
1 Advanced Bionics, Valencia, CA, USA
3 Guy’s and St. Thomas NHS Trust, London, United Kingdom
4 Advanced Bionics, Europe
Spectral resolution has been shown to be closely related
to vowel and consonant recognition in cochlear implant
The addition of Gaussian noise to electrical stimulation has
recipients. A measure of spectral resolution is the spectral
been proposed as a means to enhance the time coding of
modulation threshold (SMT), which is the smallest detectable
signals by the auditory nerve. Specifically, adding inaudible
spectral contrast in a spectral ripple stimulus. The SMT is
noise has the potential to decrease the threshold and
hypothesized to be a measure of the activation pattern in
increase the comfortable levels of electrical stimulation by
electrical stimulation. This study examined the relationships
desynchronizing auditory nerve discharges. The resulting
between spectral resolution and speech perception in CII and
widening of the dynamic range might improve speech
HiRes 90K implant users and normal hearing subjects who
recognition and simplify the choice of compression
listened through a multiband vocoder. Activation patterns
parameters. The effect of adding noise on the thresholds and
were simulated in normal-hearing subjects using a multiband
dynamic ranges for sinusoidal stimuli was tested in eight
vocoder in which stimuli first were decomposed into 15
users of the Clarion S-series implant. Noise decreased the
logarithmically spaced bands and then resynthesized by
thresholds by 2 dB and increased the dynamic range by
multiplying the envelope of each band by matched filtered
0.7 dB. Noise also decreased the loudness of moderate-to-
noise. Current spread was simulated by adjusting the drop-
intense stimuli. These data suggest that loudness is partially
off of the noise spectrum away from the peak. Increasing the
encoded by the degree of phase locking in auditory nerve
simulated current spread increased the SMT and decreased
fibers. Further work will determine if the addition of noise
vowel and consonant scores. The impairments in speech
has functional implications for speech understanding in
understanding were similar to cochlear implant users with
cochlear implant users.
similar SMTs. These data suggest that variability in spread
J Assoc Res Otolaryngol (2007) 8:42-53.
of neural activation may underlie the variability in speech
perception in cochlear implant recipients. SMTs may be
helpful in assessing spectral resolution and understanding
differences in speech perception benefit in implant users.
J Acoust Soc Am (2007) 122(2):982-991.
Psychophysics, Modeling & Novel Processing 91
Processors & Accessories
Through “front-end” innovations—like CD-quality 16-bit audio analysis, wide
programmable input dynamic range, and dual-action AGC—the Harmony™
and Auria® ear-level sound processors deliver HiResolution® Sound, a family
of sound processing options designed to preserve and deliver acoustic inputs
with high fidelity. In addition, several accessories to these sound processors
offer improved connectivity to audio devices such as MP3 and CD players,
cellular phones, FM systems, and other assistive technologies—designed to
make listening easier in the real world.
These accessories include the T-Mic®, which places the microphone in
the external ear canal and provides direct access to cellular telephones and
consumer audio headphones without requiring a connecting cable. The
T-Mic’s location within the concha takes advantage of the sound-filtering
effects of the pinna for better hearing in noise and improved sound localization.
The iConnect earhook option, with its own battery source, provides a cablefree connection to an FM receiver. The T-Coil earhook option is designed to
facilitate access to assistive listening technology such as inductive loop systems
and hearing-aid compatible telephones.
The studies in this section summarize various investigations into the listening
benefits provided by Advanced Bionics’ sound processors and accessories.
93
Evaluation of the Harmony Sound Processor:
Multicenter Study in North America
Previous research has shown that advances in
cochlear implant sound processing technology
improve outcomes in cochlear implant recipients. However, there are few reports that describe
improvements in listening benefits resulting solely
from changing the external sound processor. The
new Harmony behind-the-ear (BTE) processor was
designed to deliver better sound clarity for everyday
hearing, such as listening in noise, listening to music,
and using the telephone. The Harmony processor
implements 16-bit CD-quality front-end signal
processing with a wide programmable input dynamic
range and dual-loop automatic gain control. It
features built-in LEDs to monitor battery status and
system function along with an internal telecoil for
telephone use or for use with other electromagnetic
or induction loop assistive listening devices. In addi-
100
Own Processor
*
80
Rating
1 wk Harmony
60
40
20
0
Voices
Environmental Sounds
Figure 1. Mean ratings for clarity of recorded voices and environmental sounds (0 = very unclear to 100 = very clear) for
subjects’ own processors and after one week of Harmony use.
Clarity of environmental sounds was significantly higher with
Harmony (* p < .05).
94 tion, the Harmony processor is more power-efficient
for longer battery life. To document efficacy, this
study compared the listening and practical benefits
of the Harmony processor with previous generation
sound processors. Early results from this study have
been reported by Séguin and colleagues (Séguin et
al, 2006).
Subjects were adult CII or HiRes 90K recipients.
Age at implantation ranged from 21 to 79 years
and duration of implant use ranged from 2 months
to 5.3 years. Baseline speech recognition data were
obtained from 48 subjects using their own processors. Subsequently, subjects were fit with their
current sound processing strategy on the Harmony
and returned after one week to assess outcomes and
to evaluate their experience with the new processor.
Additional processor data were obtained one and
three months after subjects were fit with HiRes
with Fidelity 120 (HiRes 120) on the Harmony.
Outcome measures included speech recognition in
quiet and noise, ratings of recorded stimuli (voices,
environmental sounds, and music passages) (Tyler et
al, 2004), and a questionnaire that assessed comfort,
ease of use, reliability, and processor preference.
After one week, subjects rated the clarity of recorded
environmental sounds significantly higher with
Harmony compare to their own previous generation
sound processors (p < .05) (Figure 1). After three
months, a majority of subjects reported improved
sound quality, and rated speech as more natural and
easier to understand, environmental sounds as more
distinct, and music as sounding better. Subjects also
rated the Harmony more comfortable than their
own BTE processor, with 96% reporting that the
Harmony was “comfortable” or “very comfortable.”
Ease-of-use ratings for changing programs and
volume adjustments were equivalent to or better
than those reported for their own processors. The
LED features received favorable ratings at both the
one-week and three-month evaluations. After three
months, 100% of subjects rated the Harmony sound
processor as “reliable” or “highly reliable.” In addi-
Advanced Bionics® Auditory Research Bulletin 2007
Participating Centers and Clinics in North America
Carle Clinic, Urbana, Illinois
House Ear Clinic, Los Angeles, California
Houston Ear Research Foundation, Houston, Texas
Indiana University, Indianapolis, Indiana
Johns Hopkins University, Baltimore, Maryland
Mayo Clinic, Rochester, Minnesota
Midwest Ear Institute, Kansas City, Missouri
New York University, New York, New York
Northwestern University, Chicago, Illinois
Ottawa Hospital (Civic Campus), Ottawa, Ontario
University of Massachusetts, Worcester, Massachusetts
University of Texas (Southwestern Medical Center), Dallas, Texas
Washington University, St. Louis, Missouri
tion, PowerCel operating times were longer with
Harmony than with the Auria BTE sound processor.
Average operating time increased from 7 to 14 hours
with the PowerCel Slim and from 15 to 25 hours for
the PowerCel Plus (Figure 2).
This study showed that the majority of subjects
preferred the Harmony sound processor to their own
previous generation processor. Harmony’s advanced
front-end signal processing provided better sound
quality and an overall improvement in listening
benefits. Users also were highly satisfied with
Harmony’s reliability, features, comfort, and ease of
use. The Harmony’s power-efficient design resulted
in longer PowerCel operating times, with users
obtaining a full day of use from a single charge.
Auria
Harmony
24
20
Hours
Of the 48 subjects who participated in this study,
42 (88%) reported that they preferred the Harmony
sound processor to their own previous generation
processor. On a 1–10 scale, (0 = weak preference,
10 = strong preference), the average strength of
preference for the Harmony processor was 7.9 for
subjects who preferred Harmony.
28
16
12
8
4
0
PowerCel Slim
PowerCel Plus
Figure 2. Average operating times for PowerCels used with the
Auria (reported by subjects) and with the Harmony (obtained
from battery logs).
References
Séguin C, Chénier J, Armstrong S, MacDonald E. Evaluation of the Auria Harmony
sound processor. Poster presented at the 9th International Conference on Cochlear
Implants, Vienna, Austria. 14-17 June 2006.
Tyler R, Witt S, Dunn CC. (2004) Trade-offs between better hearing and better
cosmetics. Am J Audiol 13:193-199.
Processors & Accessories 95
HiRes and HiRes 120 on the Harmony:
Results in Experienced Users of Current Steering Strategies
Martina Brendel, M.Sc.
Carolin Frohne-Büchner, Ph.D.*
Andreas Büchner, Ph.D.
Beate Krüger, M.Sc.
Thomas Lenarz, Prof., M.D., Ph.D.
Medizinische Hochschule Hannover, Hannover, Germany
* also with Advanced Bionics, Europe
Recently Advanced Bionics introduced current
steering to increase the number of stimulation
sites beyond the number of physical electrode
contacts. Early implementations of current steering
were implemented on the body-worn Platinum
Sound Processor (PSP) and evaluated in subjects
at Medizinische Hochschule Hannover. This study
evaluated standard HiRes and HiRes with Fidelity
120 sound processing on the new Harmony ear-level
processor in subjects who had participated in some
of the earlier evaluations of current steering with
the PSP. (For example, see Büchner et al, at page
86 of this bulletin.) The study group consisted of 14
postlinguistically deafened subjects who had used
HiRes for more than two years prior to any currentsteering experience. The mean age was 58.8 years
(range: 25.1 to 76.2 years) and the mean duration of
deafness was 5.0 years (range: 0 to 26.0 years). Five
used a CII BTE, three used an Auria, and six used a
PSP as their everyday processors.
Subjects first were evaluated with standard HiRes on
their own processors. They then were fit with HiRes
and HiRes 120 on the Harmony, which they used
for one month. During that month, they were asked
to use HiRes and HiRes 120 each for one week, and
then to use their preferred strategy (HiRes or HiRes
120) for the remaining two weeks. Benefit was
assessed using speech perception tests (Freiburger
monosyllables, HSM sentences in speech-shaped
noise using a 10 dB SNR, HSM sentences with
single competing talker using a 5 dB SNR). Subjects
also completed questionnaires regarding general
sound quality and music perception. At the end of
the study, the “best condition” was defined as the
strategy/processor combination that yielded the
96 highest scores for the HSM sentence test presented
with a competing talker. Subjects were tested again
with their “best condition” at the end of the study.
Speech perception results are shown in Figure 1. The
highest monosyllabic word scores were obtained with
HiRes 120 on the Harmony. For the HSM sentence
test in speech-shaped noise, little differences were
observed between the three processor/strategy
combinations. Significant differences among the
strategies were found only for the HSM sentences
with competing talker. For this test, both Harmony
conditions yielded significantly better results than
HiRes on the subject’s own processor. The difference
between HiRes 120 on the Harmony and standard
HiRes on the subjects’ own processors was highly
significant (p < 0.01).
Figure 2 shows results of the questionnaire reflecting
subjective ratings for HiRes on the PSP as well as
HiRes and HiRes 120 on the Harmony. The ratings
are noticeably higher for the Harmony—especially
for clarity, overall quality, and listening with background interference. HiRes 120 showed an advantage
over HiRes for those same three rating situations.
Results were further improved when subjects rated
their experience in the “best condition.” At the end
of the study, 57% of the participants preferred HiRes
120 on the Harmony, while 29% preferred standard
HiRes on the Harmony. Only 14% preferred their
own processors.
This one-month trial verified the improvement in
benefit with the Harmony compared to previous
sound processors. The improvement can be attributed to both the enhanced front-end audio
Advanced Bionics® Auditory Research Bulletin 2007
“Interestingly, these results overall
are in agreement with the results
of a more elaborate study where subjects
with no current-steering experience
tested the same system for a longer period.”
Percent Correct
73.2
71.6
60.1
60
52.4 52.8
**
*
**
**
56.9
51.2
47.9
48.5
39.6
40
20
0
Monosyllables
HSM 10 dB SNR
HSM CT 5 dB SNR
Figure 1. Mean speech perception scores for four processor/
strategy combinations (n = 14, ** p < .01, * 0.01 < p < 0.05).
10
8
HiRes/Own Processor
HiRes/Harmony
**
*
** *
**
** **
6
HiRes 120/Harmony
Best Condition/Harmony
**
*
**
4
2
O
v
Q era
ua ll
lit
y
Ba
ck
In g
te ro
rfe un
re d
nc
N
e
at
u
O ra
wn ln
Vo ess
ic
e
N
Fe atu
m ra
al ln
e e
Vo ss
ic
es
N
at
M ur
al al
n
e e
Vo ss
ic
es
ea
sa
nt
ar
ne
ity
ss
0
Pl
Van Den Abbeele T, Noel-Petroff N, Arnold L, Crozat-Teissier N, Boyle P. Pediatric
results following switchover from HiRes to HiRes 120. Presented at the 11th
International Conference on Cochlear Implants in Children, Charlotte, North
Carolina, USA. 11-14 April, 2007.
HiRes 120/Harmony
Best Condition/Harmony
81.7
80.0
80
Cl
Reference
HiRes/Own Processor
HiRes/Harmony
100
Mean Rating
processing and the ability to implement HiRes 120.
Performance with HiRes on the Harmony processor
was as good or better than with HiRes on the
subjects’ own processors. In addition, the HSM
sentences with competing talker and questionnaire
data indicate the advantage of HiRes 120 over standard HiRes when implemented on the Harmony.
These study participants already had significant
experience using prototype strategies implementing
current steering prior to participating in this study.
Therefore, it was assumed that one month of experience with HiRes 120 on the Harmony was sufficient
to estimate their performance and preference. Interestingly, these results overall are in agreement with
the results of a more elaborate study where subjects
with no current-steering experience tested the
same system for a longer period. (See Brendel et al,
page 98 of this bulletin.) Similar outcomes also were
observed in a study of HiRes 120 and Harmony
in children (Van Den Abbeele et al, 2007; see also
page 53 of this bulletin).
Figure 2. Mean ratings from the sound quality questionnaire for
four processor/strategy combinations (n = 13,** p < 0.01,
* 0.01 < p < 0.05).
Processors & Accessories 97
Results Using the Harmony Sound Processor
with HiRes and HiRes 120 Processing
Martina Brendel, M.Sc.
Carolin Frohne-Büchner, Ph.D.*
Andreas Büchner, Ph.D.
Beate Krüger, M.Sc.
Thomas Lenarz, Prof., M.D., Ph.D.
Medizinische Hochschule Hannover, Hannover, Germany
* also with Advanced Bionics, Europe
The independent current sources in the CII and
HiRes 90K implants allow full and independent
control over the stimulation delivered to each of
the implanted electrodes. By varying the proportion of current delivered simultaneously to adjacent
electrodes, additional pitch percepts can be created
(Firszt et al, 2007; Koch et al, 2007). This current
steering technique has been implemented in the
HiRes with Fidelity 120 sound processing option
(HiRes 120) and can deliver up to 120 spectral
bands. HiRes 120 is supported by the new Harmony
ear-level processor in which the audio front end
processes an expanded 16-bit (96 dB) dynamic range
(Gault et al, 2006).
HiRes (Own)
100
HiRes 120 (Harmony)
HiRes (Harmony)
92.3 93.8 91.2
Percent Correct
80
60
41.4
40
27.9 28.4 28.4
30.8
36.1
20
6.3 7.9
11.3
0
HSM in Quiet
HSM in Noise
(10 dB SNR)
HSM with CT
(10 dB SNR)
HSM with CT
(5 dB SNR)
Figure 1. Mean sentence perception results by test condition
(n = 9). To reduce the variability and increase the reliability of
the results, data from two evaluations with standard HiRes (at
baseline and the four-month interval) were averaged, as were
data from two evaluations of HiRes 120 (at the one- and threemonth intervals). Note: HSM in Noise condition is with speech
spectrum noise; HSM with CT is with competing talker.
98 In this study, 11 postlinguistically deafened adults
evaluated the potential benefit of HiRes 120 and
Harmony. The study design was a five-month crossover protocol. First, subjects were tested with standard HiRes on their own processors (CII BTE,
Auria, or PSP). Then they were fit with HiRes 120 on
the Harmony and tested after one and three months
of use. Subsequently, subjects were refit with HiRes
on their own processors and evaluated after one
month (four-month test interval). They then were fit
with HiRes on the Harmony and tested after one
month (five-month test interval). Performance was
evaluated using the HSM sentences presented in
four conditions (in quiet, in speech spectrum noise
using a 10 dB SNR, with single competing talker
using a SNR of 10 dB and 5 dB). Subjective data
were collected via questionnaires regarding general
sound quality, music perception, and handling and
comfort of the processors. To reduce the variability
and increase the reliability of the results, data from
both evaluations of standard HiRes (baseline and
four months) were averaged, as were data from both
evaluations of HiRes 120 (one and three months).
Averaged results from the HSM sentence tests
are shown in Figure 1. The results in quiet showed
ceiling effects and the condition with a single
competing talker (5 dB SNR) showed floor effects at
all test intervals. There were no differences between
the test conditions for the HSM sentences in speech
spectrum noise (10 dB SNR). The results for the
condition with a single competing talker (10 dB
SNR) were higher for HiRes 120 as well as for
HiRes on the Harmony processor compared to
HiRes on the subjects’ own processors.
Questionnaire rating data indicated that the
Harmony provided improved benefit for sound
clarity, pleasantness, quality, hearing in background
interference, and naturalness of different voices. The
improved benefit was noted for both standard HiRes
Advanced Bionics® Auditory Research Bulletin 2007
“Notably, only one test
(sentences with competing talker)
was sensitive enough to reveal
the differences between
processing strategies and processors.”
and HiRes 120. The overall average ratings for both
Harmony conditions were significantly higher than
the ratings for HiRes on the subjects’ own processors
(Figure 2). Moreover, 64% of the subjects preferred
HiRes 120 on the Harmony and 18% preferred
HiRes on the Harmony. Eighteen percent of the
subjects had no processor preference.
HiRes (Own)
HiRes 120 (Harmony)
HiRes (Harmony)
10
9
8
7
Mean Rating
*
*
6
5
4
3
2
1
ne
ss
O
ve
r
So al
Q un l
Ba ua d
ck lity
In g
te ro
rfe un
re d
N
nc
at
e
u
O ra
wn ln
e
Vo ss
ic
N
e
Fe atu
m ra
al ln
e e
Vo ss
N
ic
a
es
M tur
al aln
e e
Vo ss
ic
es
Av
er
ag
e
ea
s
Cl
an
t
ar
ity
0
Pl
This study design allowed comparison of previous
processors with HiRes 120 implemented on the
Harmony, as well as a comparison of HiRes 120
and standard HiRes implemented on the Harmony.
Notably, only one test (sentences with competing
talker) was sensitive enough to reveal the differences
between processing strategies and processors. At the
end of the study, 9 of 11 study participants chose
to switch to the Harmony processor because of
improved speech understanding, more natural sound,
or better handling of the processor. The majority of
the study participants preferred HiRes 120 on the
Harmony sound processor, which was reflected clinically in the superior results on the HSM sentence
test with competing talker (10 dB SNR). They also
preferred the extended battery life of the Harmony.
Figure 2. Subjective ratings of listening benefit for HiRes on
the subjects’ own processors, HiRes 120 on the Harmony, and
HiRes on the Harmony (n = 9). Note: * = p < .05.
References
Firszt JB, Koch DB, Downing M, Litvak L. (2007) Current steering creates additional
pitch percepts in adult cochlear implant recipients. Otol Neurotol 28(5):629-636.
Gault A, Frohne-Buechner C, Delande JB, Arnold L, Miller D, Boyle P. Concept and
technical features of the new HiRes Auria+ processor. Paper presented at the 9th
International Conference on Cochlear Implants and Related Sciences, Vienna,
Austria. 14-17 June, 2006.
Koch DB, Downing M, Osberger MJ, Litvak L. (2007) Using current steering to
increase spectral resolution in CII and HiRes 90K users. Ear Hear 28(2):38S-41S.
Processors & Accessories 99
Input Dynamic Range:
Effect on Speech Perception in Noise
Sabine Haumann, M.Sc1
Yassaman Khajehnouri, Dipl.Inform.2
Andreas Büchner, Ph.D.1
Thomas Lenarz, Prof., M.D., Ph.D.1
1 Medizinische Hochschule Hannover, Hannover, Germany
2 Advanced Bionics, Europe
Speech perception typically is assessed using materials delivered at a fixed presentation level. However,
because implant users may optimize their sensitivity
control, fixed level testing cannot assess automatic
gain control (AGC) function and does not represent
real life listening. This study used a novel approach
for assessing speech perception in noise to differentiate AGC and input dynamic range (IDR) contributions to everyday listening (Spahr & Dorman,
2007).
Three groups of 10 experienced adult implant
listeners participated in the study. Group 1 was
composed of Freedom processor users, Group 2
of Esprit 3G users, and Group 3 of Auria users.
Groups 1 and 2 used the ACE sound processing
strategy and Group 3 used HiRes sound processing.
The groups were matched by duration of deafness,
age at implantation, years of implant use, and their
scores on the HSM sentence test at +10 dB SNR
(Table 1).
Table 1. Mean demographic and clinical
sentence perception scores for three study groups
well-matched across these various factors.*
System
Duration of
Age at
Duration of
HSM
Deafness Implantation
CI Use
+ 10 dB SNR
Freedom
1.2 yrs
56 yrs
2.6 yrs
69.8%
Esprit 3G
2.2 yrs
61 yrs
4.9 yrs
66.9%
Auria
1.7 yrs
56 yrs
2.9 yrs
64.8%
* n = 10 in each group
100 Speech perception was assessed using HSM
sentences presented at randomly roved presentation
levels around a nominal 65 dB SPL. Competing
noise was used to estimate the sentence reception threshold (SRT), which was defined as the
SNR at which the subjects understood 50% of the
words across a block of 30 sentences. Each list of
30 sentences was spoken by a male talker and delivered from a loudspeaker placed one meter directly
in front of the subject. Sentences were scored as the
number of words correctly repeated. The sentence
presentation levels were roved randomly around
65 dB SPL by +/-10 dB and +/-15 dB (range
between 55 and 75 dB SPL and between 50 and
80 dB SPL). For each test condition, 10 sentences
were delivered at each of the three presentation
levels. Initial SNR was +20 dB. When a subject
made two or fewer errors in a sentence, the SNR
was decreased for the next sentence. Otherwise,
the SNR was increased. A three-stage adaptive rule
was used to adjust the SNR where 10, 5, and 2.5 dB
changes were made following reversals in the direction of the SNR change. Subjects were tested using
their everyday processor settings. A practice list was
used before testing so that processor settings could
be adjusted.
Group mean scores on the clinical HSM sentence
test using a +10 dB SNR were not significantly
different (p > 0.5). Two-way ANOVA testing
showed that the Auria group performed significantly
better than the Freedom group for the +/-10 dB (t =
2.62, p < 0.025) and +/-15 dB (t = 2.72, p < 0.025)
roving SRT tests (Figure 1). While there was a trend
for the Auria group to be better than the 3G group,
the difference did not reach significance on either
the +/-10 dB or the +/-15 dB roving tests (t = 1.9,
p < 0.08). There were no significant differences
between the Freedom and 3G groups. However,
many of the 3G subjects at MHH already have
converted to the Freedom, so the remaining 3G
users are those who manage particularly well with
their Esprit 3G processor and who have elected not
to convert to the Freedom processor.
Advanced Bionics® Auditory Research Bulletin 2007
In summary, subjects matched by demographics and
clinical speech scores (at a +10 dB SNR) showed
statistically significant differences when evaluated
using a roving presentation level test. The group using
the Auria processor, which has the widest IDR and
a dual-action AGC, demonstrated the best performance. Further evaluation of the Freedom processor
is planned in which the IDR will be expanded
beyond the 45-dB default. The Freedom’s Automatic
Sensitivity Control (ASC) setting also will be evaluated even though it is rarely used. Overall, these data
indicate that front-end signal analysis features like a
wide IDR and dual-action AGC provide significant
advantages when listening in real-life situations. The
wide range of results also suggests that new methods
are required to assess the benefit of contemporary
cochlear implant sound processors.
Reference
Spahr A, Dorman M. (2007) Performance of patients using different cochlear implant
systems: effects of input dynamic range. Ear Hear 28:260-275.
8
7.35
Mean SRT (dB SNR)
7
6.59
6
5.69
5
3.92
4
2.83
3
2
1.62
1
Fr
+/ eed
-1 o
5 m
dB
Fr
+/ eed
-1 o
0 m
dB
+/
-1 3
5 G
dB
+/
-1 3
0 G
dB
+/ A
- 1 ur
5 ia
dB
+/ A
- 1 ur
0 ia
dB
0
Figure 1. Group mean SRT scores for the roving +/-10 and
+/-15 dB SNR test conditions. A lower SRT indicates better
performance.
Auria
3G
Freedom
20
16
SRT (dB SNR)
Given that all subjects scored over 40% correct on
the fixed level HSM sentence test at +10 dB SNR, all
would be expected to do extremely well in everyday
communication situations. However, the distribution of scores for the +4 to +5 dB SNR condition
reveals differences between the groups. This SNR
represents a typical cafeteria or classroom environment. The individual scores in Figure 2 show that
70% of subjects in the Auria group have an SRT
better than +4 dB SNR while 60% of 3G subjects
and 80% of Freedom subjects have SRTs poorer
than +4 dB SNR. This pattern also was observed for
both the +/-10 and +/-15 dB SNR test conditions.
These clear differences in SRTs may represent large
variations in real-life listening situations.
“The group using the Auria processor,
which has the widest IDR
and a dual-action AGC,
demonstrated the best performance.”
12
8
4
0
-4
40
50
60
70
80
90
100
Percent Correct
Figure 2. Individual subject SRTs on the roving +/-10 dB test
plotted against clinical HSM +10 dB SNR test scores. There is
a separation of groups for the roving test scores even though
these subjects were matched for HSM scores at +10 dB SNR.
Individual scores show that 70% of subjects in the Auria group
have an SRT better than +4 dB SNR (at or below the red line)
while 60% of 3G subjects and 80% of Freedom subjects have
SRTs poorer than +4 dB SNR (above the red line).
Processors & Accessories 101
Harmony Processor and T-Mic Benefits in Children:
Speech Perception
Lisa Cowdrey, M.A.
Kristen Lewis, Au.D.
Kristin Lisbona, M.A.
Midwest Ear Institute, Kansas City, Missouri, USA
“. . .results showed that five of nine subjects
demonstrated improved speech recognition
in at least one noise condition. . .”
Own Processor
25
Harmony and T-Mic
SNR (dB)
20
15
10
5
0
S04
S08
S09
S10
S05
Noise from Front
Figure 1. Individual BKB-SIN scores for unilateral subjects with
speech and noise both presented from the front.
Own Processor
25
Harmony and T-Mic
Figures 1 and 2 show the BKB-SIN results for the
unilateral subjects for noise presented from the front
and noise presented on the side of the implant before
and after being fit with the Harmony and T-Mic.
Individual results are as follows:
SNR (dB)
20
15
10
5
• S09 showed improvement in both noise
conditions with the Harmony and T-Mic.
0
S04
S08
S09
S10
S05
Noise from Implant Side
Figure 2. Individual BKB-SIN scores for unilateral subjects with
speech presented from the front and noise presented on the
side of the implant.
102 The Harmony behind-the-ear (BTE) processor was
designed to deliver better sound clarity through its
16-bit CD-quality front-end processing. In combination with the new smaller T-Mic component,
the Harmony has the potential to improve hearing
benefit for pediatric CII and HiRes 90K users. This
study explored the communication advantages of
the Harmony and T-Mic in a total of nine children—five with unilateral implants and four with
sequentially implanted bilateral devices. All children
had CII or HiRes 90K implants and used HiRes
sound processing except for one bilateral subject
(S03) who used simultaneous analog stimulation
(SAS) with his first, previous-generation (Clarion)
implant. Speech perception was evaluated at baseline with the child’s own processor. Then the child
was fit with the Harmony and small T-Mic. After
two weeks of use, each child’s speech perception was
reevaluated. If, prior to the study, the child already
used a Harmony processor, the child was fit with the
new smaller T-Mic and evaluated after two weeks.
Speech perception was measured using the BKBSIN test in which sentences are presented in fourtalker babble where the signal-to-noise (SNR) varies
from +21 to -6 dB. Speech is presented from the
front with either noise from the front or noise from
90 or 270 degrees (toward the side of the implant in
the unilateral users, from either side in the bilateral
users). A lower SNR is a better score.
• S08 showed improvement with Harmony
and T-Mic with spatial separation of speech
and noise (Figure 2) and no difference in
performance with speech and noise from the
front (Figure 1).
Advanced Bionics® Auditory Research Bulletin 2007
Recent Public at ion
Performance of Patients Using Different
Cochlear Implant Systems: Effects of Input
Dynamic Range
• S10 showed essentially equivalent performance
in both noise conditions with Harmony and
T-Mic and his own processor.
• S04 showed comparable performance with
Harmony and T-Mic with spatial separation
of speech and noise (Figure 2) but poorer
performance with speech and noise from the
front (Figure 1).
Anthony J. Spahr, Ph.D.1
Michael F. Dorman, Ph.D.1
Louise H. Loiselle, M.S.2
1 Arizona State University, Tempe, AZ,USA
2 The Mayo Clinic, Scottsdale, AZ, USA
The goal of this study was to determine if cochlear
implant device characteristics affected other hearing
abilities in adult implant users who had identical scores
on a test of word recognition in quiet. Speech perception
• S05 performed better in both noise conditions
with his own processor than with the Harmony
and T-Mic.
in quiet and noise, voice discrimination, and melody
In the bilateral subjects, the three subjects with
CII or HiRes 90K implants in both ears showed
improved speech recognition with Harmony and
T-Mic when speech and noise were presented from
the front, whereas the subject with one Clarion
implant showed equivalent benefit. The scores with
noise presented from each side yielded mixed results
with two of the four children experiencing improved
benefit with the Harmony and T-Mic depending
upon whether the noise was on the side of the first or
second implant. It is notable that two of the subjects
had used their second implant for only a few months
at the time of testing.
were matched for word recognition in quiet, duration
Overall, the BKB-SIN results showed that five
of nine subjects demonstrated improved speech
recognition in at least one noise condition with
the Harmony and T-Mic compared with their
own processor and microphone configuration. The
children had only two weeks to adapt to the new
processor and microphone placement. More experience might yield additional improvements with
the new technology. Interestingly, seven of the nine
children demonstrated BKB-SIN scores of approximately 6 to 16 dB SNR, which are similar to the
BKB-SIN scores of implanted adults.
increasing the IDR had no effect on sentence perception
recognition were evaluated in three matched groups of
implant recipients, each of which used either the CII-BTE,
Esprit 3G, or Tempo+ behind-the-ear processors. Subjects
of deafness, age at testing, and duration of device
experience. Results showed that CII users performed
better on vowel identification than 3G and Tempo+
users. CII and Tempo+ users had higher scores than 3G
users on difficult sentences presented in noise (+10 and
+5 dB SNR). CII users performed higher than 3G users for
difficult sentences presented at a low level (54 dB SPL).
A second set of experiments assessed performance on
sentences presented at conversational levels in quiet and
noise—and at low levels in quiet for CII and 3G recipients
in which the input dynamic range (IDR) was widened.
For CII users, increasing the dynamic range improved
performance in all three conditions. For 3G users,
in quiet, degraded sentence understanding in noise, and
improved performance at low levels. The results indicate
the differences in implant design affect hearing benefit—
especially in more difficult listening situations. IDR and
compression characteristics may be the principle factors
contributing to the higher performance of the CII system.
Ear Hear (2007) 28(2):260-275.
Reference
Buckler L, Lewis K, Lisbona K. Evaluation of the Harmony processor and T-Mic in
children. Poster presented at the 11th International Conference on Cochlear Implants
in Children, Charlotte, North Carolina, USA. 11-14 April, 2007.
Processors & Accessories 103
Harmony Processor and T-Mic Benefits in Children:
Questionnaire Results
Lisa Cowdrey, M.A.
Kristen Lewis, Au.D.
Kristin Lisbona, M.A.
Midwest Ear Institute, Kansas City, Missouri, USA
“In general, parents were highly satisfied
with Harmony’s
features, comfort, and ease of use.”
Table 1. Percent of parents who
agreed or strongly agreed with statements
of Harmony processor benefit.
Statement of Perceived Benefit
with Harmony and T-Mic
Percent Responding
“Agree” or
“Strongly Agree”
My child understands speech better in noise.
67%
My child reported that sound was better.
67%
My child understands speech better in quiet.
56%
My child understands speech better on the
telephone.
44%
My child distinguishes environmental
sounds better.
33%
My child shows more interest in music.
33%
My child speaks with better voice quality.
33%
My child speaks with greater clarity.
33%
104 The Harmony behind-the-ear (BTE) processor was
designed to deliver better sound clarity through its
16-bit CD-quality front-end processing. In combination with the T-Mic component, the Harmony
has the potential to improve hearing benefit for CII
and HiRes 90K users. In addition, the improved
power efficiency, longer battery life, battery- and
microphone-status LEDs, and new smaller T-Mic
features make the Harmony a practical option for
children.
To explore the advantages of Harmony in children,
Buckler and colleagues (2007) compared Harmony
and previous-processor use in nine children (4-10
years of age) who had been implanted with CII
or HiRes 90K devices and who used HiRes sound
processing. Parents completed a questionnaire after
their children used the Harmony and pediatric
T-Mic for two weeks.
The questionnaire results showed that:
• All parents (9 of 9) preferred Harmony and
the T-Mic to their child’s own processor and
microphone. Eighty-nine percent of the parents
preferred Harmony because of their child’s
improved listening performance, while 11%
preferred Harmony because of its ease of use.
• Mean strength of preference for Harmony was
7.9 on a scale of 10 (where weak preference = 1,
strong preference = 10).
• Mean rating of overall satisfaction with
Harmony and T-Mic was 6.0 on a scale of 7
(where 1 = very dissatisfied, 7 = very satisfied).
Parents also rated those aspects of sound that were
better with Harmony and the T-Mic using the 7point scale: strongly disagree, disagree, somewhat
disagree, neutral, somewhat agree, agree, strongly
agree. One- to two-thirds of the parents agreed or
strongly agreed that their child heard better in a
variety of listening situations (Table 1).
Advanced Bionics® Auditory Research Bulletin 2007
LED Function Indicators
7
6
Mean Rating
Parents rated the usefulness of the microphonestatus and battery-status LEDs on a scale from 1 to
7, where 1 = not useful at all and 7 = very useful. The
mean ratings ranged from 5.4 to 6.0 (Figure 1). They
also evaluated wearing comfort and ease of use using
a 7-point scale where 1 = very uncomfortable or very
difficult and 7 = very comfortable or very easy. The
ratings ranged from 5.3 to 6.9 (Figure 2).
5
4
3
2
With respect to power efficiency, 78% of parents
reported that:
1
RF lock
• Their child used the PowerCel Slim.
• The PowerCel used by their child lasted an
entire school day.
Battery
Level
Low
Battery
Overal Ease
of Use
Figure 1. Mean parent ratings of Harmony LED function from
1 (not useful at all) to 7 (very useful).
• The PowerCel used by their child lasted all
waking hours.
Reference
Buckler L, Lewis K, Lisbona K. Evaluation of the Harmony processor and T-Mic in
children. Poster presented at the 11th International Conference on Cochlear Implants
in Children, Charlotte, North Carolina USA. 11-14 April, 2007.
Comfort and Usability Features
7
6
Mean Rating
5
4
3
2
em
en
t
R
e
w t
/o en
Su tio
pp n
Re or
w ten t
/S t
up io
po n
Ch rt
Ea a
rh ng
oo e
k
Ch s
Po a
w ng
er e
Ce
l
Ch
Pr an
og g
ra e
m
Ch
Vo ang
lu e
m
e
ac
Pl
t
pt
fo
r
Ac
ce
m
an
ce
1
Co
In summary, these data show that the Harmony
with T-Mic is a practical processor option for children. All parents (100%) preferred Harmony and
T-Mic to their child’s own processor and microphone configuration. A large percentage of parents
(89%) attributed their preference to their observation of improved listening benefits. One PowerCel
lasted the school day and all waking hours for most
children. In general, parents were highly satisfied
with Harmony’s features, comfort, and ease of use.
Future investigations will examine the benefits of
the Harmony and T-Mic in younger children.
Figure 2. Mean parent ratings of Harmony wearing comfort and
ease of use from 1 (very uncomfortable or very difficult) to
7 (very comfortable or very easy). Some features were not
evaluated by all parents.
Processors & Accessories 105
Benefits of Using Directional Microphones
in a Diffuse Noise Environment
F. B. Van der Beek, M.D.
Wim Soede, Ph.D.
Johan H. M. Frijns, M.D., Ph.D.
Leiden University Medical Centre, Leiden, The Netherlands
In recent years, the speech recognition capability of
adult cochlear implant users has increased rapidly,
particularly for listening to speech in quiet (e.g.,
Firszt et al, 2004; Ramsden, 2004). However, speech
understanding deteriorates rapidly in the presence of
background noise (e.g., Spahr and Dorman, 2004).
Several approaches are being taken to improve the
intelligibility of speech in background noise. One
approach focuses on improving the fidelity of the
signal delivered to the cochlea (such as by optimizing the number of electrodes and rates of stimulation or with the use of bilateral implants). Another
approach is to develop noise reduction algorithms or
to use directional microphones. The second approach
already is used widely with conventional hearing aids
(Luts et al, 2004; Soede, 1993 a and 1993b) and may
be beneficial to cochlear implant recipients.
3.55 m
3.15 m
N
N
N
sphere
S
N
N
N
2.40 m
N
N
Figure 1. Diffuse noise setup with eight loudspeakers emitting background noise (N) and one loudspeaker for speech (S).
The distance between the chair and the loudspeaker providing speech is 1 m. The stand for the handheld microphone is
located 0.75 m from the speech loudspeaker. The sphere illustrates the position of the listener’s head.
106 To explore the applicability of directional microphones in cochlear implants, this study compared
the benefits of two assistive directional microphone
systems to a standard omnidirectional cochlear
implant microphone. Comparisons were made
using test conditions representing real life listening
where the signal was presented in a reverberant
field with multiple competing noise sources. Thirteen postlinguistically deafened adult CII Bionic
Ear users participated in the study. The mean age
was 45.3 years and the mean duration of cochlear
implant experience was 12.3 months (range: 3 to
21 months). All subjects used the continuous interleaved sampling strategy on a body-worn Platinum
Sound Processor.
Experiments were conducted in a sound-treated
room where speech and noise were presented
from identical self-powered loudspeakers (AV110,
Conrad, Germany). Uncorrelated steady-state noise
was delivered through eight loudspeakers placed on
the edges of an imaginary box. Speech was delivered via a loudspeaker placed one metre from the
centre of the box and 1.2 metres from the floor
(Figure 1). The speech signals consisted of CVC
words in Dutch recorded on CD. The CVC test lists
consisted of four sublists of eleven three-phoneme
monosyllables each. Testing was performed in quiet
at 65 dB SPL and with decreasing signal-to-noise
ratios (SNR) down to -15 dB, where the speech
level was held constant at 65 dB SPL and the noise
levels were varied. Speech reception thresholds
(SRTs in dB SNR) and absolute phoneme scores
were measured.
Each subject was tested with a standard headpiece
microphone, and with the two assistive directional
microphones: the handheld FM-system TX3
Handymic (Phonak, Bubikon, Switzerland) and the
Linkit array microphone (Etymotic Research Inc.,
Elk Grove Village, Illinois), which is worn on the
head. The output spectra of the Handymic and the
Linkit were equal within a margin of ± 3 dB within
the frequency range of 500 to 4,000 Hz. To minimize learning effects, the three microphones were
tested in random order using a Latin-square design.
Advanced Bionics® Auditory Research Bulletin 2007
Figure 2 shows the individual speech-in-noise benefits for the directional microphones compared to the
headpiece microphone. The values were calculated
by subtracting the SRTs for the linear interpolated
data for the Handymic, or Linkit, from the SRTs
for the headpiece microphone. SRTs for all subjects
improved with the experimental microphones.
Although there was a trend for greater benefit from
the Handymic compared to the the Linkit (about
2.3 dB), this difference may be explained partly
by the closer positioning of the Handymic to the
speech loudspeaker. However, the Handymic and
the Linkit have been designed for different applications, and the choice of directional microphone
should be made based on the listening situation in
which it will be used. The SRT difference values for
Handymic and Linkit were not correlated with the
SRTs with the headpiece microphone or with the
duration of deafness.
These results suggest that directional microphone
systems can improve speech recognition in noise for
cochlear implant users. In noisy environments such
as restaurants or parties, these assistive microphones
may make conversation easier, less fatiguing, and
more natural for cochlear implant recipients.
“These results suggest that
directional microphone systems
can improve speech recognition in noise
for cochlear implant users.”
Benefit Compared to Headpiece Mic (dB)
Results showed that the directional microphones
did not improve speech recognition in quiet, but
provided significant improvement in background
noise compared to the omnidirectional headpiece
microphone. At 0 dB SNR, the average CVC
scores improved from 42% for the headpiece microphone to 67% and 62% for the Handymic and the
Linkit, respectively. The mean SRT values for the
Handymic and the Linkit were significantly better
than the SRT values obtained with the headpiece
microphone (p < 0.001). Mean SRTs improved by
8.2 dB for the Handymic and by 5.9 dB for the
Linkit. Although the mean SRT was lower for the
Handymic compared to the Linkit, the difference
was not significant (p = 0.3), and the differences
were similar to those reported previously for hearing
aids (Luts et al, 2004).
16
Handymic
Linkit
14
12
10
8
6
4
2
0
-2
Mean
-4
-6
Individual Scores
Figure 2. Individual and group mean SRT benefit of Handymic
and Linkit compared to headpiece microphone expressed in
dB. Benefit was calculated by subtracting the SRT for each of
the two experimental microphones from the SRT for the headpiece microphone.
References
Firszt JB, Holden LK, Skinner MW, Tobey EA, Peterson A, Gaggl W, Runge-Samuelson
CL, Wackym PA. (2004) Recognition of speech presented at soft to loud levels
by adult cochlear implant recipients of three cochlear implant systems. Ear Hear
25:375-387.
Luts H, Maj JB, Soede W, Wouters J. (2004) Better speech perception in noise with an
assistive multimicrophone array for hearing aids. Ear Hear 25:411-420.
Ramsden RT. (2004) Prognosis after cochlear implantation. Brit Med J 328:419-420.
Soede W, Berkhout AJ, Bilsen FA. (1993a) Development of a directional hearing
instrument based on array technology. J Acoust Soc Am 94:785-798.
Soede W, Bilsen FA, Berkhout AJ. (1993b) Assessment of a directional microphone
array for hearing impaired listeners. J Acoust Soc Am 94:799-808.
Spahr AJ, Dorman MF. (2004) Performance of subjects fit with the Advanced Bionics
CII and Nucleus 3G cochlear implant devices. Arch Otolaryngol HNS 130:624-628.
Processors & Accessories 107
Remote Microphone Technology Benefits in Adults
Christiane Séguin, M.Sc.(A)
Elizabeth Fitzpatrick, Ph.D.
David Schramm, M.D.
Josée Chénier, M.O.A.
Shelly Armstrong, M.Cl.Sc.
Ottawa Hospital (Civic Campus), Ottawa, Ontario, Canada
For many adults with cochlear implants, understanding speech in noise is much more difficult
than understanding speech in quiet. However,
implanted adults rarely have taken advantage of
assistive technologies that might help them hear
better in compromised listening situations. As a first
step toward understanding the benefit that might
be obtained using remote microphone technology,
this study documented the listening benefits that
implanted adults receive from using radio frequency
(RF) and infrared (IR) technology while watching
television.
Seven adult users of the Auria or Harmony sound
processors participated in this single-session study.
Age of implantation ranged from 42 to 70 years
and duration of cochlear implant use ranged from
four months to four years. All had open-set HINT
sentence scores in quiet greater than 50% (range:
56-100%) at their most recent clinical evaluations.
Participants were seated in a test room seven feet
from a 20-inch television screen. Using a recorded
television segment, subjects adjusted the volume of
the speech processor and television to a comfortable
level. The test segments consisted of six different
television recordings of approximately 45 seconds
in duration—three news segments by the same
announcer and three talk-show segments with the
same host. The test recordings were presented to
the participants in a randomized manner in each of
three listening conditions, as follows:
• In their everyday listening condition—
unilateral cochlear implant alone or cochlear
implant with hearing aid in the contralateral
ear.
• With an infrared system (Sennheiser 810S)
coupled via the direct-connect earhook to the
Auria/Harmony processor.
• With a radio frequency system (Sennheiser
820S) coupled via the direct-connect earhook
to the Auria/Harmony processor.
Following each television segment, participants were
asked questions to assess their general comprehension of the content. Next, they were asked to repeat
the TV segments in short phrases of 5-10 words
that were scored as percent correct. Participants also
completed three questionnaires that probed general
television habits and experience, perceptions of each
listening condition after the three news segments,
and perceptions of each listening condition after the
three talk show segments.
108 Advanced Bionics® Auditory Research Bulletin 2007
“These preliminary results suggest that
assistive listening devices benefit
adult cochlear implant users
during television viewing.”
Alone
100
with RF
with IR
90
80
Percent Correct
Figure 1 shows the mean percent correct speech
perception scores for the talk-show segments in the
cochlear implant alone (or with hearing aid), cochlear
implant with RF system, and cochlear implant with
IR system conditions. Scores were greater when
using the assistive listening devices compared to
the everyday listening configuration for all subjects.
Questionnaire results revealed that participants
preferred listening to television with one of the assistive listening devices rather than with their cochlear
implant alone. All participants who judged their
comprehension with a cochlear implant alone to be
less than 50% perceived that their comprehension
improved with at least one or the other of the assistive listening devices. Similarly, they indicated that
they had more confidence in listening to television
with an assistive device. Five of seven reported that
considerably less effort was required when listening
with either the RF or IR system.
70
60
50
40
30
20
10
0
1
2
3
4
5
6
7
Subject
Figure 1. Speech recognition results (talk show segments) for
seven adults using a cochlear implant alone, cochlear implant
plus RF system, and cochlear implant plus IR system.
These preliminary results suggest that assistive
listening devices benefit adult cochlear implant users
during television viewing. In particular, patients
reported greater confidence and increased ease of
listening. Future phases of this study will examine
speech perception with and without an RF system
in quiet and noise, as well as evaluate benefits in
everyday listening environments. Moreover, the adult
data have implications for using assistive listening
devices with children in that the benefits of remote
microphone technology may extend beyond classroom listening into social and home activities such
as watching television.
Reference
Séguin C, Fitzpatrick E, Schramm D, Chénier J, Armstrong S. The benefits of
remote microphone technology: learning from adults. Poster presented at the
11th International Conference on Cochlear Implants in Children, Charlotte, North
Carolina, USA. 11-14 April, 2007.
Processors & Accessories 109
iConnect/FM Benefit in Adults
Carolin Frohne-Büchner, Ph.D.1,2
Martina Brendel, M.Sc.2
Alexandre Gault, M.Sc.1
Andreas Büchner, Ph.D.2
Thomas Lenarz, Prof., M.D., Ph.D.2
1 Advanced Bionics, Europe
2 Medizinische Hochschule Hannover, Hannover, Germany
Over the last twenty years or so, assistive communication systems have significantly helped many
people with hearing impairments to understand
speech in difficult listening situations. The most
commonly used devices are frequency modulation
(FM) systems. With advances in miniturization, the
size and weight of FM receivers has been reduced
significantly so that they can be worn at ear level
provided a connection to the hearing aid or cochlear
implant (CI) is available.
Figure 1. Auria with iConnect earhook.
This study examined the benefits that adult Bionic
Ear users gained from using an FM system in
conjunction with the iConnect earhook option
on the Auria behind-the-ear sound processor
(Figure 1). The iConnect earhook has its own battery
source and allows users to connect and power a mini
FM receiver. Thus, the iConnect provides a cablefree connection to the FM receiver without affecting
the battery life of the sound processor itself. For
this study, the iConnect was tested with an MLxS
receiver and the Phonak Campus transmitter.
Ten adult Auria users participated in the study.
Their ages ranged from 29.5 to 67.1 years and duration of implant use ranged from 1.4 to 3.4 years.
Subjects were fit with the iConnect/FM system and
instructed on its use. Speech perception in quiet and
noise was tested with and without the FM system
where the speech signal was delivered from directly
in front of the subject with the noise presented
on the implanted side. The speech signals were
presented at 65 dB SPL in quiet and using a 10 dB
signal-to-noise ratio in the noise condition.
For the iConnect/FM testing, the FM transmitter
microphone was placed in front of the loudspeaker
with a presentation level equivalent to 65 dB (A) at
110 Advanced Bionics® Auditory Research Bulletin 2007
“. . .the iConnect/FM system
provides hearing benefit to
adult implant recipients,
especially when listening in noise.”
one meter. Three mixing ratios were evaluated: “aux
only” (signal only from the FM system), 50/50 (50%
of signal from the ear-level microphone, 50% from
the FM system) and 30/70 (30% of signal from the
ear-level microphone, 70% from the FM system).
Without FM
With FM System
100
80
Percent Correct
Results showed that the sound quality and speech
understanding using the FM system was good
without making any changes to the speech processor
program. There were no difficulties in handling
and using the iConnect, even though the study
participants were tested over a couple of hours
only and had a wide range of experience with FM
systems and accessories. The speech perception data
(Figure 2) showed that there was little advantage of
the FM system for understanding speech in quiet.
Using the “aux only” ratio, speech-in-noise scores
were equivalent to speech-in-quiet results. The
30/70 ratio provided the next highest benefit in
noise although the 50/50 ratio also provided some
advantage. Interestingly, nine of the ten adult users
preferred the “aux only” setting when listening to
speech in noise.
60
40
20
0
Quiet at
65 dB
Noise
from 90°
Quiet
Aux Only
Noise
Aux Only
Noise
50/50
Noise
30/70
Figure 2. Mean sentence recognition scores for 10 adults with
and without the iConnect/FM system for different mixing ratios
(aux only = 100% of signal through FM). For the speech-innoise condition, speech was presented from the front and noise
from a 90-degree azimuth.
Thus, the iConnect/FM system provides hearing
benefit to adult implant recipients, especially when
listening in noise. As a fitting guideline for adults, no
adjustments to the processor programs are required
and the “aux only” mixing ratio is appropriate. The
iConnect earhook provides a cable-free FM connection option that does not affect the battery life of the
Auria processor. The iConnect also is available with
the Harmony processor.
Reference
Frohne-Büchner C, Brendel M, Gault A, Büchner A, Lenarz T. Early experience with
the Auria iConnect. Presentation at the 8th Pediatric Cochlear Implant Conference,
Venice, Italy. 26 March, 2006.
Processors & Accessories 111
Evaluation of the iConnect FM Adapter in Adults
Jace Wolfe, Ph.D.1
Erin C. Schafer, Ph.D.2
1 Hearts for Hearing Foundation, Oklahoma City, OK, USA
2 University of North Texas, Denton, TX USA
Previous research has shown that the use of
frequency-modulated (FM) systems can improve
speech understanding in challenging listening situations for cochlear implant users. However, most of
the research with FM systems has been conducted
with children, presumably because they require
favorable listening conditions to learn language and
to function in school classrooms. Adult cochlear
implant users also have the potential to benefit from
FM technology. Notably, the introduction of smaller
FM systems and wireless adapters to couple the FM
receiver to the implant sound processor makes the
FM-assistive listening technology more cosmetically appealing to adult users. This study examined
the use of the Phonak MicroLink MLxS personal
FM and the iConnect wireless FM adapter in adult
cochlear implant users. In addition to eliminating
cables, the iConnect has no switches or controls
that require adjustments by the audiologist or user,
thereby facilitating ease of fitting and use. Another
CI + FM
100
CI Only
Percent Correct
80
60
40
20
0
CNC in Quiet
65 dB
CNC in Quiet
50 dB
HINT in Noise
+8 dB SNR
Figure 1. Speech recognition at a conversational level (CNC
words at 65 dB SPL), soft level (CNC words at 50 dB SPL) and
with four-talker babble (HINT sentences at 65 dB SPL with +8
dB SNR) with 50/50 mixing ratio for 12 adults with FM off (CI
alone) and FM on (CI plus FM).
112 advantage is the iConnect’s independent power
source (size 10 hearing aid battery), designed to have
a negligible effect on the current drain of the battery
in the sound processor.
A within-subjects study design was used to assess
speech perception benefits with the iConnect in
12 adults who had a CII Bionic Ear or HiRes 90K
implant for at least six months. None of the subjects
had previous experience with a personal FM system.
Performance was examined with the FM transmitter turned off (CI alone) and FM transmitter
turned on (CI plus FM) in the following conditions:
CNC words presented at soft (50 dB SPL) and
conversational (65 dB SPL) levels as well as HINT
sentences presented in four-talker babble (65 dB
SPL presentation level with +5 dB signal-to-noise
ratio). Subjects participated in two test sessions
wherein their performance was evaluated with the
iConnect programmed with a 30/70 or 50/50 mixing
ratio and a 10-dB gain setting. The selection of the
initial mixing ratio was counterbalanced across
subjects who were blinded to the mixing ratio they
were using. Subjects used the iConnect with the
initial mixing ratio for two weeks before returning
for evaluation. They then were fit with the alternate
mixing ratio and evaluated again after two weeks. At
the end of the study, subjects completed a questionnaire about everyday listening benefits, comfort, and
ease of use with the iConnect.
Figures 1 and 2 show results obtained with the
50/50 and 30/70 mixing ratios. There was a significant difference between the FM-off and FM-on
conditions for soft speech (CNC at 50 dB SPL)
and HINT sentences in noise for both mixing ratios
(p <. 001), but not for conversational speech. There
was no significant difference between mixing ratios
in any condition with the FM transmitter turned on
(CI plus FM). With the transmitter turned off, the
CNC score was higher with the 50/50 than 30/70
mixing ratio (p = .002)
Advanced Bionics® Auditory Research Bulletin 2007
“. . .these results indicate that adults
can benefit from using a personal FM
system in difficult listening conditions.”
Questionnaire results obtained from 10 subjects
revealed that all of them reported improved listening
for at least some noisy situations with the MLxS
and iConnect. The greatest benefit was reported for
speech understanding in a car and when listening to
the television. The subjects perceived no differences
between the 30/70 and 50/50 mixing ratios. In addition, the subjects reported that the iConnect was
comfortable and simple to use. The iConnect battery
lasted an average of 30 hours and there was no significant effect on the operating time of the Auria battery
for nine of the ten subjects. (One subject reported a
decrease of two hours (from 12 to 10 hours) in the
operating time of the Auria battery). Nine of the ten
subjects reported that they would recommend use
of the iConnect to other cochlear implant recipients. One subject stated that she preferred to use her
T-Mic over the iConnect because it allowed for
easier telephone use in her job.
In summary, these results indicate that adults can
benefit from using a personal FM system in difficult listening conditions. Four of the 12 subjects
chose to purchase the MLxS system at the end of
the study. Notably, those four subjects demonstrated
the poorest speech recognition scores in noise
without the FM system. Thus, use of the FM technology might be particularly useful for those individuals who experience the greatest difficulty in
noisy environments.
CI + FM
100
CI only
Percent Correct
80
60
40
20
0
CNC in Quiet
65 dB
CNC in Quiet
50 dB
HINT in Noise
+ 8 dB SNR
Figure 2. Speech recognition at a conversational level (CNC
words at 65 dB SPL), soft level (CNC words at 50 dB SPL)
and with four-talker babble (HINT sentences at 65 dB SPL
with +8 dB SNR) with 30/70 mixing ratio for 12 adults with
FM off (CI alone) and FM on (CI plus FM).
CI Only
100
CI + FM (10 dB)
CI + FM (16 dB)
80
Percent Correct
A secondary protocol was employed with five of the
subjects to examine the effect of FM gain setting on
speech recognition performance. Specifically, testing
was completed with the 30/70 mixing ratio with a
10-dB and 16-dB gain setting. Figure 3 shows that
all subjects demonstrated better performance with
the FM system turned on compared to their performance with the FM transmitter turned off. The
performance of the subjects was comparable with
the two different gain settings, except for Subject 2
who actually performed better with the 10-dB gain
setting. This finding contrasts with that reported
for children with the iConnect who demonstrated
improved speech perception performance with a
higher gain setting. (See also Wolfe and Schafer at
page 114 of this bulletin). These results underscore
the importance of optimizing FM fitting parameters, especially for children.
60
40
20
0
1
2
3
Subjects
4
5
Figure 3. Individual results for subjects with 30/70 mixing
ratio with FM off (CI alone) and FM on (CI plus FM system)
using a 10-dB and a 16-dB gain setting.
Processors & Accessories 113
Evaluation of the iConnect FM Adapter in Children
Jace Wolfe, Ph.D.1
Erin C. Schafer, Ph.D.2
1 Hearts for Hearing Foundation, Oklahoma City, OK, USA
2 University of North Texas, Denton, TX USA
Frequency-modulated (FM) systems can help
overcome some of the listening problems created
by distance, noise, and reverberation for cochlear
implant users. Some clinicians, however, have been
reluctant to recommend personal FM systems for
a number of reasons. First, previous technology
required cumbersome cables and adapters to couple
the FM system to the cochlear implant sound
processor. Second, FM systems have adjustable
parameters that should be optimized for each child,
yet little research has been done on determining the
settings that yield the best performance. Notably,
there is no consensus regarding the most appropriate
gain setting on the FM receiver. Many clinicians
use the default setting of 10 dB but there is limited
objective evidence to support this practice. Third, the
operating time of the battery in the speech processor
can be reduced substantially by the power consumption of a personal FM receiver.
Recent technological advances make personal
FM systems a more viable option for cochlear
implant users with CII Bionic Ear or HiRes 90K
implants. Personal FM systems have been reduced
in size (Phonak MicroLink MLxS) and the iConnect™ earhook allows wireless coupling of the FM
receiver to the Auria or Harmony behind-the-ear
sound processors. In addition to eliminating cables,
the iConnect has no switches or controls that
require adjustments by the audiologist or patient,
thereby facilitating ease of fitting and use. Another
advantage is the independent power source of the
iConnect, which uses a size 10 hearing aid battery.
Thus, there is no current drain on the sound
processor battery.
114 This study examined the speech perception benefits
of the miniaturized MLxS FM receiver used with
the iConnect earhook and Auria sound processor in
children. A within-subjects design compared speech
recognition performance with the FM transmitter
turned off (CI alone) and FM transmitter turned on
(implant plus FM system) in the following conditions: speech presented at soft (50 dB SPL) and
conversational (65 dB SPL) levels as well as speech
presented in background noise (speech at 65 dB
SPL). Performance in quiet was assessed with either
the PBK-50 monosyllabic word recognition test
or the Multisyllabic Lexical Neighborhood Test
(MLNT), depending on the child’s age and language
abilities. Speech recognition in noise was assessed
with the Hearing in Noise for Children (HINT-C)
presented with four-talker babble at a +5 dB signalto-noise ratio (SNR).
A secondary goal was to evaluate the effect of FM
gain setting on speech recognition in noise. The
gain settings available on the MLxS receiver range
from -6 to 24 dB. Performance was assessed on the
HINT-C with the default gain setting of 10 dB as
well as a higher setting of 16 dB. Testing in quiet
was conducted only with the 10 dB gain setting. All
testing was conducted using a 50/50 mixing ratio.
This ratio generally is agreed to be most appropriate
for children to enable them to hear their own voices
and the voices of their classroom peers as well as the
signal from the FM receiver.
Five children implanted with the CII Bionic Ear
or HiRes 90K cochlear implant participated in the
study. The children demonstrated some open-set
Advanced Bionics® Auditory Research Bulletin 2007
“. . .these results show that use of
the iConnect with a personal FM system
improves speech recognition
at soft levels and in noise.”
speech recognition and had at least six months of
cochlear implant experience at the time they were
enrolled in the study. None of the subjects had
previous experience with a personal FM system.
The mean results in noise showed significantly
higher performance in the FM-on condition than
the FM-off condition for both gain settings. Mean
performance with the 16-dB gain setting was significantly higher than that with the 10-dB gain setting.
Figure 2 shows the individual sentence recognition
scores obtained in noise for both gain settings.
80
Percent Correct
The mean results in quiet showed significantly
higher performance in the FM-on condition than in
the FM-off condition for soft speech (50 dB SPL)
but not for speech presented at a conversational level
(65 dB SPL). Figure 1 shows the individual word
recognition scores obtained with soft speech with
and without the FM system.
60
40
20
0
1
2
3
4
5
Subject
Figure 1. Word recognition in quiet at a soft level (50 dB SPL)
for each subject with FM off (CI alone) and FM on (CI plus FM
system).
FM Off
100
FM On (10 dB)
FM On (16 dB)
80
Percent Correct
In summary, these results show that use of the
iConnect with a personal FM system improves speech
recognition at soft levels and in noise. In everyday
situations, low-level speech most often originates at
a distance from the listener, thus the results suggest
better distance hearing with a personal FM system.
In addition, performance was significantly improved
with a higher (16 dB) gain setting. Future research
will examine performance with other gain settings.
Questionnaire results revealed that the children and
their parents found the iConnect simple to use and
comfortable to wear.
FM On
FM Off
100
60
40
20
0
1
2
3
4
5
Subject
Figure 2. Sentence recognition in noise (four-talker babble
at 65 dB SPL with +5 dB SNR) for each subject with FM off
(CI alone) and FM on (CI plus FM system) for two gain settings (+10 and +16 dB).
Processors & Accessories 115
iConnect/FM System Benefits in School-Aged Children
Andrea Hedley-Williams, Au.D.
Anne Marie Tharpe, Ph.D.
Vanderbilt University, Nashville, TN, USA
“. . .the current study demonstrated that
speech understanding in noise is improved
when a FM system is used
with a cochlear implant.”
Poor speech perception in the presence of background noise continues to be a problem for cochlear
implant users despite improvements in sound
processing technology, programming techniques,
and auditory training. For school-aged children,
understanding speech in noise is of particular
concern. The signal-to-noise ratio (SNR) in most
classrooms is reported to be +5 dB (Crandell et al,
1998), resulting in adverse listening conditions that
can affect learning and communication for children
with cochlear implants.
A common classroom intervention strategy for children with hearing aids is use of a personal or sound
field frequency-modulated (FM) system. Although
individual results vary, these devices have largely
been successful for improving the SNR and speech
perception performance of children with hearing
loss, particularly in noisy environments (Boothroyd
and Iglehart, 1998; Crandell et al, 1998; Kenna,
2004). Advances in FM technology and the development of wireless FM adapters have reduced the
complexity of fitting and wearing FM systems. The
purpose of this study was to evaluate speech perception in noise in school-aged cochlear implant users
who were fit with the Auria iConnect and a FM
system (Microlink MLxS receiver and Campus S
transmitter from Phonak Hearing Systems).
Subjects were eight children, ages 7-11 years, who
had been implanted unilaterally with a CII or HiRes
90K device. Subjects had at least three months of
experience with their implants and used no hearing
aid in the contralateral ear. Children were fit with
the iConnect FM adapter and a Phonak Microlink
MLxS receiver during a single test session. Half of
the subjects were tested using the implant alone first
followed by the implant with FM. The other half
used the implant with FM first, then the implant
alone. To verify that auditory input to the Auria was
not degraded when coupled to the iConnect, speech
116 Advanced Bionics® Auditory Research Bulletin 2007
Spondee thresholds revealed no signal degradation when the iConnect was coupled to the Auria.
Figure 1 shows the HINT-C results for the eight
children when listening with and without the FM
system. All subjects exhibited improved performance in the FM condition (that is, a lower SNR).
Seven of eight children reported that it was “as easy”
or “easier” to hear in noise with the FM system than
with the implant alone. All subjects described the
iConnect as comfortable. Most subjects thought the
iConnect was cosmetically acceptable and reported
that they would wear it at school (Figure 2).
In summary, the current study demonstrated that
speech understanding in noise is improved when a
FM system is used with a cochlear implant. Furthermore, the school-aged subjects reported subjective
benefits and expressed a desire to wear and use the
iConnect adapter and FM system at school.
CI alone
25
CI + FM
20
dB SNR
15
10
5
0
-5
Individual Scores
Figure 1. HINT-C thresholds in dB SNR for the implant alone
and for the implant plus iConnect/FM system test conditions
(n = 8). Scores are rank-ordered from worst to best for the
implant-alone condition.
Agree/Strongly Agree
Don’t Know
Disagree/Strongly Disagree
8
Number of Responses
detection thresholds were assessed in both conditions (CI alone, CI+FM) with spondees presented
in the sound field from a loudspeaker in front of
the listener. Test materials consisted of the HINTC where speech was presented from the front and
speech-spectrum noise from the rear. An adaptive paradigm was used to determine the SNR at
which a 50% score was attained. Subjective impressions relating to iConnect comfort and FM use and
benefit were obtained via a questionnaire completed
by each child following testing.
7
6
5
4
3
2
1
0
iConnect
looks good.
iConnect
feels comfortable.
Would wear
at school.
Figure 2. Cosmetics and wearability ratings for the iConnect.
References
Boothroyd A, Iglehart, F. (1998) Experiments with classroom FM amplification. Ear
Hear 19:202-217.
Crandell C, Holmes A, Flexer C, Payne M. (1998) Effects of sound field FM
amplification on the speech recognition of listeners with cochlear implants. J Ed
Audiol 6:21-27.
Hedley-Williams A, Tharpe AM. iConnect/FM system benefits in children. Poster
presented at the 11th International Conference on Cochlear Implants in Children,
Charlotte, North Carolina, USA. 11-14 April, 2007.
Kenna M. (2004) Medical management of childhood hearing loss. Pediatr Ann
33(12):822-832.
Processors & Accessories 117
iConnect Benefits in Children
Diane Rafter, B.Sc.1
Elizabeth Kohl, M.Sc.1
Wendy Horler, B.A.1
Marsha Jenkins, M.Sc.1
Jane Bevan, M.Sc.2
Alexandre Gault, M.Sc.2
Deborah Vickers, Ph.D.2
1Guy’s and St. Thomas’ NHS Foundation Trust,
London, United Kingdom
2Advanced Bionics, Europe
FM systems have become increasingly useful in
classrooms where children with cochlear implants
must overcome the hearing difficulties presented
by distance, noise, and reverberation. Until recently,
long cables have been required to couple the child’s
BTE processor to the body-worn FM receiver. These
cables were cumbersome and awkward for children.
They were prone to breakage, increased the potential for interference, and detracted from many of
the advantages of a BTE processor. The iConnect
accessory was designed to overcome these disadvantages by providing a cable-free connection between
the Auria or Harmony BTE processors and the
Phonak MLxS FM receiver. The iConnect contains
an internal battery so that the FM receiver does not
draw power from the processor itself.
This study assessed comfort, ease of use, and retention of the iConnect accessory. Eight children with
the Auria BTE sound processor were studied. Speech
perception was measured to verify the advantages of
an FM system and to determine whether programming changes were required to optimize iConnectFM use.
Questionnaires administered before and after FM
use evaluated comfort and ease. Speech perception
in quiet and noise was tested, with and without the
FM system. Recorded speech signals were delivered
from directly in front of the child, with the noise
presented perpendicular to the BTE from above
the head. The speech material was presented at
65 dB SPL in quiet. The noise condition employed
a 0 dB signal-to-noise ratio with white noise at
65 dB SPL. Three mixing ratios were evaluated—“aux only” (signal only from the auxiliary FM
118 Advanced Bionics® Auditory Research Bulletin 2007
“Testing showed that
speech understanding improved
when using the iConnect-FM system
compared to the implant alone in noise.”
system), 50/50 (50% of signal from the ear-level
microphone, 50% from the FM system) and 30/70
(30% of signal from the ear-level microphone, 70%
from the FM system).
Generally, the iConnect/FM system was easy to
use and beneficial for these school-aged children.
The researchers recommended setting the system
to a 30/70 mixing ratio to optimize hearing with
the FM system, while still allowing some acoustic
signals through the BTE microphone. Using this
setting allows children to monitor their own voices
and receive some environmental information while
listening to the classroom instructor.
Common Phrases
Without FM
With FM System
100
80
Percent Correct
Questionnaire data indicated that the children
generally found the system easy to use and comfortable to wear. Some of the younger children reported
that changing the battery was “sometimes difficult,”
although this task likely would be handled by a
parent or teacher. No changes were required in the
programs, except for the adjustment of the mixing
ratio being studied. Testing showed that speech
understanding improved when using the iConnectFM system compared to the implant alone in noise.
Scores were equivalent for the aux-only and 30/70
test conditions, but poorer for the 50/50 condition
(Figure 1).
60
40
20
0
Quiet at
65 dB
Noise
from 90°
Quiet
Aux Only
Noise
Aux Only
Noise
50/50
Noise
30/70
Figure 1. Mean Common Phrases scores for eight children
with and without the iConnect-FM system for different mixing ratios (aux only = 100% of signal through FM). Significant
differences (p < .01) were found between speech in quiet and
speech in noise (speech in front, noise at 90 degree azimuth)
without the FM, and between the aux-only and 50/50 mixing ratio for speech in noise. No significant difference was
observed between the aux-only and 30/70 mixing ratio for
speech in noise.
Reference
Rafter D, Kohl E, Horler W, Jenkins M, Bevan J, Gault A, Vickers D. Results of the
Auria iConnect study: pediatrics. Poster presented at the 8th Pediatric Cochlear
Implant Conference, Venice, Italy. 26 March, 2006.
Processors & Accessories 119
Optimizing FM Systems for the Auria Speech Processor
Erin C. Schafer, Ph.D.1
Jace Wolfe, Ph.D.2
1 University of North Texas, Denton, Texas, USA
2 Hearts for Hearing Foundation, Oklahoma City,
Oklahoma, USA
“. . .an FM system directly coupled
to the Auria sound processor
provides significant benefits
for listening in noise.”
It is well documented that frequency-modulated
(FM) systems that are directly connected to children’s sound processors significantly improve speech
recognition performance in noise when compared to
a cochlear implant alone condition. Furthermore, a
recent study we conducted using a meta-analytical
approach suggests that FM receivers connected
directly to the implant system provide significantly
greater benefit than classroom sound field or desktop
sound field FM receivers. Recent developments
in FM technology allow audiologists to adjust the
relative inputs from the sound processor and FM
receiver to optimize benefit for the user.
This study determined the effects of FM receiver
(Phonak MLxS) gain on speech perception abilities in adults and older children who use one or two
Auria speech processors. Participants were tested at
several FM gain settings (+6, +10, +14, +20 dB) using
the Bamford-Kowal-Bench Speech-in-Noise (BKBSIN) test. The BKB-SIN automatically adapts the
speech and noise stimuli to yield a speech-in-noise
threshold in dB at the 50% correct performance
level. A lower signal-to-noise ratio (SNR) indicates
better performance. Normative data from adults and
older children using cochlear implants indicate that
a 3.1 dB difference in threshold SNR is significant.
BKB-SIN results for two adults (Subjects 1 and 2)
who use a single Auria speech processor are shown
in Table 1. The data show that using the FM receiver
has a significant influence on speech perception in
noise. Both subjects received significant benefit from
the FM system at all gain settings (+6, +10, +14,
+20 dB), with the greatest benefit at the higher gain
settings. However, despite the better performance
at the +20 dB setting, these adults reported some
discomfort in the highest gain conditions.
BKB-SIN results from three study participants
(Subjects 3, 4, and 5) using sequential bilateral Auria
speech processors are shown in Table 2. All three
120 Advanced Bionics® Auditory Research Bulletin 2007
subjects received significant benefits from using the
FM system. Similar to the unilateral participants,
the FM gain settings influenced speech recognition performance, but results were different among
users. For Subject 3, scores in all the +10 dB FM
conditions were significantly poorer than scores in
the +20 dB FM conditions. For Subject 4, use of an
FM receiver on only one ear yielded significantly
better scores than the bilateral implants alone in all
but two conditions. Subject 4 also represents a user
who performs significantly better with a single FM
receiver on the ear implanted second versus a single
FM receiver on the ear implanted first. Scores for
this subject and for Subject 5 are significantly better
with two receivers when compared to the majority
of conditions with one FM receiver.
These preliminary findings suggest that FM gain
settings can significantly influence speech recognition in noise for children and adults using Auria
speech processors. For users of one Auria speech
processor, settings of +10 and +14 dB provide
optimal performance. Bilateral Auria users also
benefit from the higher FM gain settings, and most
bilateral users received more benefit when using two
FM receivers over one.
In summary, an FM system directly coupled to the
Auria sound processor provides significant benefits
for listening in noise. However, given the variability
in scores in this study group, individualized testing
may be required to determine the optimal FM
settings for each user and for each implant.
Table 1. BKB-SIN scores in dB SNR with no FM versus four FM gain settings for two unilateral Auria users.*
Subject
Age
No FM
+ 6 dB
+ 10 dB
+ 14 dB
+ 20 dB
1
58 yrs
11.5
2.5
1.25
-1.25
-2.75
2
50 yrs
15.5
3.0
-0.5
-1.5
-2.75
* Both subjects implanted with HiRes 90K devices. A lower dB SNR score indicates better performance.
Table 2. BKB-SIN scores in dB SNR with no FM and two FM gain settings for bilaterally implanted Auria users.*
Subject
Age
No FM
1st Implant
+ 10 dB
2nd Implant
+ 10 dB
Both Implants
+ 10 dB
1st Implant
+ 20 dB
2nd Implant
+ 20 dB
Both Implants
+ 20 dB
3
55 yrs
15.0
4.25
7.0
3.25
-1.5
0.25
-2.5
4
32 yrs
21.0
18.75
14.5
7.0
19.5
6.25
5.75
5
14.3 yrs
19.25
7.75
10.25
3.0
6.25
4.75
2.5
* All devices are HiRes 90K except for Subject 5 whose first implant is a CII Bionic Ear. A lower dB SNR score indicates better performance.
Processors & Accessories 121
Dual-Loop AGC Improves Speech Perception in Noise for
CII and HiRes 90K Implant Users
Patrick Boyle, M.Sc.1,2
Michael A. Stone, Ph.D.1
Brian C. J. Moore, Ph.D.1
Yassaman Khajehnouri, M.Sc.3
Andreas Büchner, Ph.D.3
Thomas Lenarz, Prof., M.D., Ph.D.3
1University of Cambridge, Cambridge, United Kingdom
2Advanced Bionics, Europe
3Medizinische Hochschule Hannover, Hannover, Germany
“Although all subjects benefited from
both compression algorithms,
AGC2 proved significantly better than
AGC1 when listening to speech in noise.”
Analogue
pre-emphasis
Audio
IN
multiply
ADC
OUT to
channel
filters
gain
table
rectify
fast
timing
select
largest
interlink
algorithm
slow
timing
Figure 1. Block diagram of the dual-loop AGC system (AGC2).
In both hearing aids and cochlear implants, automatic gain control (AGC) is used to compress the
large dynamic range of everyday sounds into the
reduced dynamic range of a hearing impaired or
implanted listener. Although much research has
focused on optimizing AGC for hearing aids,
little work has been done on optimizing AGC for
cochlear implant recipients. This study compared
two types of single-channel AGC systems—a fastacting syllabic compressor (AGC1) and a dual–time
constant system (AGC2) developed by Moore and
colleagues (Moore et al, 1991). AGC2 uses two
gain control loops—one with a fast attack time to
protect the user from sudden intense sounds and the
other with a slow-acting attack time and recovery
to maintain a constant overall volume (Figure 1).
The loop with the higher control voltage (typically
the slow loop) usually determines the gain, and fast
loop acts only when needed to reduce the effects of
sudden increases in sound level. Most often the gain
changes slowly, thereby avoiding distortion of the
temporal envelope and minimizing cross-modulation of different input signals, and thus theoretically
maximizing speech intelligibility when background
sounds are present (Stone & Moore, 2004). AGC2
is the standard compression algorithm used in the
Auria and Harmony sound processors, with AGC1
available as a programmable option. (AGC1-type
compression is used in most other implant systems.)
Six experienced CII or HiRes 90K implant users
who wore the Auria sound processor participated in
this study. Baseline speech perception was assessed
with their clinical programs, which used the standard
AGC2. Subjects then were fitted with either AGC1
or AGC2 (using research software), which they used
for one month. Subsequently, subjects were fitted
with the alternate AGC algorithm, which they used
for one month. Subjects were tested immediately
after being fitted with each compression algorithm
(“acute” testing) and after one month of listening
experience (“long-term” testing). Two sentence
perception tests were used to assess benefit. In one
122 Advanced Bionics® Auditory Research Bulletin 2007
Figure 2 shows the individual results for AGC2
plotted against the results for AGC1 obtained from
the fixed SNR testing. All subjects scored better with
AGC2 than with AGC1 at both the acute and longterm test intervals. Long-term test results from the
roving SRT test are shown in Figure 3. Two twoway within-subjects ANOVAs were applied to both
the fixed and roving SRT test results. Factors were
AGC type and test interval (acute or long term).
Performance was found to be significantly better for
AGC2 than for AGC1 on the fixed level test [F(1,4)
= 243.3, p < 0.001] and on the roving SRT test
[F(1,4) = 94.8, p < 0.001]. As expected, there were
no learning effects between the acute and long-term
test intervals for AGC2. Learning effects for AGC1
were marginally significant on the fixed level test but
not significant on the roving SRT test.
Acute Scores
Percent Correct AGC2
100
80
60
40
20
0
20
40
60
80
100
Percent Correct AGC1
Figure 2. Individual AGC2 vs. AGC1 percent correct scores for
HSM sentences presented at a fixed SNR. Squares represent
“acute” scores; triangles represent “long-term” scores.
10
8
6
4
2
0
AGC1
Although all subjects benefited from both compression algorithms, AGC2 proved significantly better
than AGC1 when listening to speech in noise.
Notably, subjects reported very little difference
between AGC1 and AGC2 in quiet, thereby demonstrating the need for testing compression systems in
conditions more representative of everyday life. The
superior performance of the AGC2 system indicates
that further development should be aimed at optimizing the dual-loop system for cochlear implant
users. Moreover, a roving-level test may be a better
tool than a fixed-level test for assessing benefit in
everyday listening situations.
Long-term Scores
0
dB SNR
test, HSM sentences were presented in unmodulated speech-shaped noise (Hochmair-Desoyer et
al, 1997) using a fixed signal-to-noise ratio (SNR),
chosen individually for each subject to yield a score
between 30% and 70% correct. The other test used
an adaptive procedure to measure the speech reception threshold for HSM sentences in noise (SRT in
dB SNR) while the overall level was varied over a
20 dB range.
AGC2
Sentence Reception Thresholds
Figure 3. Mean long-term sentence reception thresholds (SRTs
in dB SNR) for AGC2 and AGC1 (lower SNR representing
better performance).
References
Boyle P, Stone MA, Moore BCJ. Evaluation of dual and fast-acting AGC systems
in cochlear implants: initial results. Poster presented at the British Cochlear Implant
Group, Dublin, Ireland. 12-13 April, 2007.
Hochmair-Desoyer I, Schulz E, Moser L, Schmidt M. (1997) The HSM sentence
test as a tool for evaluating the speech understanding in noise of cochlear implant
users. Am J Otol 18:S83.
Moore BCJ, Glasberg B, Stone MA. (1991) Optimization of a slow-acting automatic
gain control system for use in hearing aids. Brit J Audiol 25:171-182.
Stone MA, Moore BCJ. (2004) Side effects of fast-acting dynamic range compression that affect intelligibility in a competing speech task. J Acoust Soc Am
116:2311-2323.
Processors & Accessories 123
Bilateral Implantation
Providing binaural hearing through bilateral cochlear implantation has become
an accepted medical treatment for both adults and children with hearing loss.
The degree to which cochlear implant patients can benefit from binaural
implantation is related to the way in which sound is processed and delivered to
the two ears. The HiResolution® Bionic Ear System and HiResolution sound
processing options may offer particular advantages to patients with two devices
because the system preserves the intensity, spectral, and timing cues important
for binaural hearing.
Based upon the benefits experienced by adults and older children with two
implants, several new studies are exploring the benefits of simultaneous bilateral implantation in very young children. It is anticipated that children who
use two devices at an early age will develop auditory, speech, and language
skills earlier and faster than children with only one device.
The studies in this section summarize various investigations exploring and
documenting the benefits of bilateral implantation in adults and children.
125
Simultaneous Bilateral Cochlear Implantation in Adults:
Comparison of Bilateral and Unilateral Benefit
Multicenter Study (1) in North America
Because the safety and efficacy of unilateral cochlear
implantation is well established, there is a growing
trend to provide two devices to patients rather than
just one. Most studies have examined benefit in
adults and children with sequentially placed devices,
often with a substantial period of time between
placement of the first and second implant. Relatively
few studies have examined performance in individuals who received two implants at the same time.
The primary purpose of this prospective study is to
examine bilateral benefit in postlinguistically deafened adults who receive a cochlear implant in each
ear during the same operation.
There are two phases in the ongoing multicenter
study. In the first phase, a prospective counterbalanced between- and within-subjects design is used
to evaluate bilateral listening benefits and to compare
sound processing modes—CIS vs. HiRes. The study
uses a six-month crossover design (three months
Right
100
Bilateral
Left
Percent Correct
80
60
40
20
0
3 months
6 months
8 months
Figure 1. Mean CNC word scores in quiet for the left ear alone,
right ear alone, and both ears together for 15 subjects. Results
are shown for three, six, and eight months following initial
stimulation.
126 with each processing mode) with an additional onemonth period in which subjects reevaluate the two
processing modes (two weeks with each mode) and
indicate a preference. Subjects then are evaluated at
one month after using their preferred mode (eight
months total). The second phase of the study evaluates the bilateral benefit of HiRes with Fidelity 120
(HiRes 120) sound processing. Subjects are fit with
HiRes 120 and evaluated after three months.
Thus far, 15 subjects have reached the eight-month
test interval, and their results are combined across
sound processing algorithms to show overall bilateral benefit. Word and sentence recognition scores
in quiet show that the bilateral scores were higher
than either ear alone at all test intervals and that
bilateral benefit improved over time (Figure 1).
Binaural speech-in-noise benefit was determined by
measuring reception thresholds for sentences (RTS)
using the HINT adaptive paradigm. The RTS was
measured for nine conditions. For the left implant
alone, the right implant alone, and for both implants
together, the speech signal originated from the
front and the noise signal originated from the front,
the right, and the left. Binaural listening benefits
(binaural redundancy, head shadow, and binaural
squelch) were determined by comparing RTSs
between various pairs of conditions. Eleven subjects
were able to undergo the HINT adaptive testing.
Binaural redundancy represents the benefit of two
implants in noise environments where speech and
noise are not spatially separated. After eight months,
9 of 11 (82%) demonstrated a bilateral redundancy
effect compared to at least one unilateral condition,
and 2 (18%) showed better bilateral scores than both
unilateral scores. Head shadow is the advantage
a listener experiences when listening with the ear
having the better signal-to-noise ratio (SNR) when
speech and noise are spatially separated. All subjects
(100%) demonstrated a head shadow advantage for
all the conditions in which they were tested. Binaural
squelch is the additional advantage above the head
Advanced Bionics® Auditory Research Bulletin 2007
Participating Centers and Clinics in North America
California Ear Institute/Let Them Hear Foundation, Palo Alto, California
Carle Clinic Association, Urbana, Illinois
Dallas Otolaryngology, Dallas, Texas
House Ear Clinic/House Ear Institute, Los Angeles, California
Midwest Ear Institute, Kansas City, Missouri
New York University, New York, New York
Spectrum Health, Grand Rapids, Michigan
Tampa Bay Hearing and Balance Center, Tampa, Florida
University of Miami, Miami, Florida
An adaptive sector localization task was used to
provide functional descriptions of directional hearing
ability. Sound source locations corresponding to 12
equally spaced loudspeakers (sectors) in the horizontal plane in the rear half field were simulated
using head-related transfer functions and delivered to the auxiliary input of the sound processor.
The target sounds consisted of a wideband impulse
stimulus (a “gunshot”) presented from a randomly
selected sector and randomly roved (+/- 3 dB range)
about a comfortable level. Subjects listened to two
presentations of the stimulus and then indicated
the source location. Subjects heard two blocks of
12 pairs each. At the end of the two blocks, if the
subject’s 4-sector score indicated an accuracy of 60%
or greater, two more blocks were administered. If
the subject’s 4-sector score was less than 60%, the
testing was stopped. If the subject’s 12-sector score
was 60% or greater, four additional blocks were
administered. Figure 2 shows the mean localization
accuracy scores for the 15 subjects at eight months.
Subjects perform at chance levels when using only
one implant. However, the mean accuracy improves
when using both implants, although the accuracy is
still below that of normal listeners. Individual data
for all subjects also show a positive improvement
when using two implants rather than just one.
The results to date indicate that bilateral implantation continues to be beneficial to postlinguistically deafened adults. The effect of enhanced sound
processing, which may preserve additional acoustic
information essential for binaural hearing, will be
addressed in future reports. Preliminary HiRes 120
results for 12 subjects indicate that eight of them
prefer the enhanced strategy to either CIS or standard HiRes.
Right Alone
Left Alone
Normal Listeners
100
Localization Accuracy
shadow effect when adding in the contribution of
the ear with the poorer SNR when speech and noise
are spatially separated. For 10 subjects with eightmonth results, 7 (70%) exhibited a squelch effect for
at least one of the noise-right or noise-left conditions. Overall, the binaural squelch effect was small
and the variance across subjects was large.
Bilateral
Chance
80
60
40
20
0
2
3
4
Number of Sectors
6
12
Figure 2. Localization accuracy as a function of number of
sectors for the right implant alone, left implant alone, and
both implants together at eight months post initial stimulation.
Chance levels and data from normal listeners are shown for
comparison.
Bilateral Implantation 127
Simultaneous Bilateral Cochlear Implantation in Adults:
Everyday Listening Benefits
Multicenter Study (2) in North America
“. . .these data illustrate that
bilateral cochlear implant recipients report
considerable everyday listening benefits
in a variety of
difficult listening environments.”
Normal hearing listeners gain significant everyday
benefits from binaural hearing, such as the ability to
localize sounds and to understand speech in noisy
environments. Traditional audiological assessments
typically evaluate perception of sound or voice
stimuli where spatial position is static and the spectral and temporal characteristics are fixed or predictable (Gatehouse & Noble, 2004). With the increased
spectral and temporal resolution implemented in
HiRes sound processing, cochlear implant recipients
have the potential to obtain better speech perception in noise and improved overall sound quality in
everyday listening situations.
To capture the real-world listening experiences of
cochlear implant users, subjects participating in
the study of simultaneous bilateral implantation in
adults completed the Speech, Spatial and Qualities
of Hearing Scale (SSQ) (Gatehouse & Noble, 2004)
and the Glasgow Health Status Inventory (GHSI)
(Robinson et al, 1996). Subjects completed the
questionnaires at baseline (prior to implantation)
and after eight months of device use.
The SSQ measures listening experiences in three
domains: speech hearing, spatial hearing, and sound
quality—each domain defined as follows:
• Speech Hearing Rating Scale: This domain
assesses speech understanding in a variety of
listening environments, such as one on one, in
a small group, or in a large group with various
types of noise or attentional interference.
• Spatial Rating Scale: This domain assesses
spatial listening, such as the ability to identify
distance, movement, and direction of sounds
in a variety of indoor and outdoor locations
with various types of competing noise or
environmental sounds.
• Sound Qualities Rating Scale: This domain
rates sound recognition and clarity, such as
128 Advanced Bionics® Auditory Research Bulletin 2007
The GHSI assesses quality of life in individuals with
hearing disorders and evaluates subjects’ perception of their psychological, social, and physical
well-being. The GHSI includes self reports of how
often hearing loss affects involvement in everyday
activities, self-confidence in social interactions, and
the impact of hearing loss on job opportunities.
Responses are rated on a 1-5 scale.
There are a total of 18 items in the GHSI. Nine
items (50%) showed significant improvement
(p < .05) between baseline and eight months.
Figure 2 shows a representative sample of GHSI
results.
Overall, these data illustrate that bilateral cochlear
implant recipients report considerable everyday
listening benefits in a variety of difficult listening
environments.
8 months
9
8
Mean Rating
7
6
5
4
3
2
gr
ou
p
co
in nve
re rs
st at
ph
au io
ra ns
on
nt
e
s
co
nv
er
sa
di
re
tio
ct
ns
io
n
of
so
di
un
st
d
an
ce
of
so
di
u
re
nd
ct
io
n
of
tra
so
un
ffi
c
ds
ar
e
ex wh
pe e
ct re
ed
fa
m
ili
ar
m
us
ev
ic
er
yd
ay
co
so
nv
un
er
ds
sa
tio
ns
in
ca
vo
r
ice
an s ar
d ec
na le
tu ar
ra
l
1
0
Figure 1. Mean ratings for 10 sample test items from the SSQ at
baseline and after eight months of device use.
5
baseline
8 months
4
Mean Rating
There are a total of 51 items in the SSQ. For the
15 subjects, 49 items (96%) showed significant
improvement (p < .05) between baseline and eight
months. Figure 1 shows results from a representative
sample of SSQ items.
baseline
3
2
1
0
qu
e
pa ncy
rti ac
cip ti
at vity
io
n
aff
ec
te life
d
by no
HL t
no se
t a lf c
ffe on
ct fid
ed e
by nce
jo
HL
no b o
t a pp
ffc ort
te un
d it
by ie
no
HL s
ts
el
fc
on
ab sc
ou iou
no
tH s
ti
L
nc
on
ve
ni
en
by ced
HL
pa
so rti
cia cip
l a ati
ct on
iv
do
iti in
es
n’
tw
so ith
cia d
l s raw
itu f
at rom
io
ns
Subjects read a short description of an everyday
listening situation and rate their performance on a
0–10 scale.
10
fre
evaluating the distinctness of two different
sounds occurring at the same time and the
naturalness of specific sounds or voices with
no visual cues in various difficult listening
environments.
Figure 2. Mean ratings for eight sample test items from the
GHSI at baseline and after eight months of device use.
References
Gatehouse S, Noble W. (2004) The speech, spatial and qualities of hearing scale
(SSQ). Intl J Audiol 43:85-99.
Robinson K, Gatehouse S, Browning G, Glasgow S. (1996) Measuring patient
benefit from otorhinolaryngological surgery and therapy. Ann Otol Rhinol Laryngol
105:415-422.
Bilateral Implantation 129
Bilateral Benefit in Twins Implanted Simultaneously
with Two Devices before Twelve Months of Age
Michael A. Novak, M.D., Project Director
Danielle Edmonson, M.A.
Amy Fiscus, M.S.
Shawnda Fuller, M.A.
Angie Hallam, M.S.
Dianne Hammes, M.A.
Lee Ann Rotz, M.A.
Jean Thomas, M.S.
Mary Willis, M.S.
Julie Wirth, M.S.
Carle Clinic Association and Carle Foundation Hospital,
Urbana, IL, USA
Table 1. Rates of language growth following
implantation for the twins and for the two cohorts
implanted before 12 months of age and
between 13 and 18 months of age.
13-18 Mos
12 Mos & Younger
Twins C1 & K2
TEST
INTERVAL
CHRON
AGE
(mos)
LANG
AGE
(mos)
LANG
GAP
(mos)
LANG
QUOTIENT
Most
Recent
35 & 35
40 & 41
5&6
1.14 & 1.17
1.5 yr post
27 & 27
28 & 28
1&1
1.03 & 1.03
1 yr post
23 & 22
22 & 22
-1 & 0
.95 & 1
Most
Recent
55.6
58.4
2.8
1.06
1.5 yr post
28.4
27.8
-0.6
0.97
1 yr post
22.6
21.5
-1.1
0.94
Most
Recent
79.9
73.9
-6.0
0.92
1.5 yr post
36.1
28.7
-7.4
0.80
1 yr post
29.6
27.1
-2.4
0.97
130 The aim of this ongoing study is to assess the benefit
of two implants on the development of communication skills in twin boys implanted under 12 months
of age. The identical twins were identified at Carle
Foundation Hospital with profound sensorineural
hearing loss and fit with hearing aids at two months
of age. They were enrolled in an intensive auditoryoral program at three months of age. Both boys
received bilateral HiRes 90K devices at ten months
of age and have reached two years postimplantation as of March 2007. Postoperative assessments
with age-appropriate auditory, speech, and language
measures have taken place at six-month intervals.
The twins’ development on these measures is being
compared to a larger cohort of infants implanted
by 12 months of age (n = 8, 4 of 8 bilateral—the
twins included in this group), a cohort of children
implanted at 13-18 months of age (n = 7, 1 of 7
bilateral), and a cohort of children implanted at 1930 months of age (n = 20, 4 of 20 bilateral). These
groups consist of children who have hearing loss
as their only disability and with a minimum of 12
months of cochlear implant experience.
Figure 1 shows the language quotients (LQ) after 24
months of implant use for the twins (designated C1
and K2) relative to the mean most recent scores for
the three comparison groups. A language quotient
is derived by dividing language age by chronological
age. A language quotient near 1 is typical for twoyear-old children with normal hearing. Separate
analyses show similar language quotients for both
the unilaterally and bilaterally implanted children.
On the whole, the two groups with the youngest ages
at implantation have language quotients commensurate with normal-hearing peers. The twins also have
met this level of achievement.
Table 1 illustrates the rate of language growth
following implantation for the twins and for the
two cohorts implanted before 12 months of age and
between 13 and 18 months of age. The twins’ rate
of language growth closed the language gap within
12 months of implantation, kept pace with that of
normal-hearing peers, and closely matched that of
the cohort implanted by 12 months of age. Interest-
Advanced Bionics® Auditory Research Bulletin 2007
“. . .this study continues to support the
efficacy of early implantation
and demonstrates that implantation
in infancy can foster
age-appropriate speech and
spoken language skill development.”
ingly, children implanted between 13 and 18 months
demonstrate an average language gap of six months,
whereas the twins and the youngest cohort demonstrated no gap.
At their most recent evaluation, the twins’ speech
recognition skills had reached ceiling performance
(90-100%) for closed-set measures (GASP-W, ESP
Standard-Spondees, ESP Standard-Mono, Mr.
Potato Head–Words and Sentences). The exception
was their performance on the GASP sentences test
(30% score for each twin). This lower score may have
been impacted by their poor speech intelligibility,
thereby resulting in an underestimation of their
speech recognition for connected speech.
The WIPI test (in noise) also was administered to
remove biasing of test scores resulting from poor
speech intelligibility. Twin C1 demonstrated a
binaural advantage in noise with a signal to noise
ratio (SNR) of +10 dB. In contrast, K2, who was
tested with a SNR of +15 dB, did not show a binaural
advantage. Testing could not be completed at +10 dB
SNR with K2 because he became too distracted and
could not complete the task. It should be noted that
equivalent performance in noise is reported anecdotally by the children’s therapists and parents.
In summary, the twins demonstrate a rate of language
development commensurate with that expected in
normal-hearing children at 24 months after implantation. Early implantation and rapid development in
auditory and language skills allowed for early identification of the twins’ apraxia, minimizing additional
delays that might occur as a result of their speech
disorders. The twins’ developmental progress is in
good agreement with the cohort implanted at less
than 12 months of age. Overall, this study continues
to support the efficacy of early implantation and
demonstrates that implantation in infancy can foster
age-appropriate speech and spoken language skill
development (Hammes et al, 2002).
1.2
1
Language Quotient
In sharp contrast to their rapid gains on language
and speech perception measures, the twins’ speech
development has progressed much more slowly. As
the gap between their language and speech skills
widened, it became clear that use of auditory-based
therapy methods alone was not yielding the expected
rate of advancement in the twins’ speech production.
Their therapists began to suspect that hearing loss
was not the primary reason for their speech delays.
Evaluation with the Kaufman Speech Praxis Test
for Children (KSPT) revealed that both twins had
childhood apraxias of speech. Because each of the
twins had excellent speech perception and otherwise
age-appropriate language skills, this secondary diagnosis was made at an age commensurate with that
for normal-hearing peers with apraxia. The twins
now are responding favorably to therapy modified
for this disorder.
0.8
0.6
0.4
0.2
0
C1
K2
≤ 12 mos
13–18 mos
19–30 mos
Figure 1. Language quotients for the twins at 24 months
postimplant (C1 and K2) and mean language quotients for the
three comparison groups at their most recent test intervals.
Language quotient is derived by dividing language age by
chronological age.
References
Hammes D, Novak MA, Rotz LA, Willis M, Edmondson D, Thomas J. (2002) Early
identification and cochlear implantation: critical factors for spoken language
development. Ann Otol Rhinol Laryngol 111(Suppl 189):74-78.
Thomas J, Novak M, Edmondson D, Fiscus A, Fuller S, Hammes D, Rotz LA, Wirth
J. Bilateral benefit in twins implanted simultaneously with two devices before 12
months of age. Poster presented at the 11th International Conference on Cochlear
Implants in Children, Charlotte, North Carolina, USA. 11-14 April, 2007.
Bilateral Implantation 131
New Study
Study Init
Initiat
iatiive
ve
New
Auditory Skill Development in Young Deaf Children with Bilateral HiRes 90K Implants:
Background and Justification for Study Design in the United States
B. Robert Peters, M.D.1
Jeanne Black, Ph.D.2
medical or psychological contraindications. However, the
1 Dallas Otolaryngology, Dallas, TX, USA
literature has been criticized by some reviewers. The low
2 Advanced Bionics, Valencia, CA, USA
rate of cochlear implantation has resulted in publication
design of pediatric bilateral studies in the peer-reviewed
of findings based on relatively small sample sizes or
There is an increasing trend to implant profoundly deaf
retrospective analysis. Some have assumed incorrectly that
children bilaterally so that they may experience the numerous
only randomized, controlled trials (RCTs) provide reliable
advantages of binaural hearing in everyday situations.
evidence of treatment effectiveness. In fact, the majority
For example, both normal-hearing listeners and bilateral
of surgical research consists of non-randomized studies,
cochlear implant users hear speech in noise and localize
especially for low incidence conditions.
sounds much better when using two ears versus one ear.
Moreover, educational research demonstrates that children
The U.S. Centers for Medicare and Medicaid Services
with single-sided deafness—even with normal hearing in the
(CMS) acknowledges this in its Decision Memo for Cochlear
other ear—experience significant difficulties hearing in the
Implantation, which stated that “strength of evidence
classroom and fall behind their peers in numerous indicators
generally refers to 1) the scientific validity underlying study
of educational success.
findings regarding causal relationships between health care
interventions and health outcomes; and 2) the reduction
In addition to the general advantages of binaural hearing,
of bias. . .regardless of whether the design of a study is a
there may be other compelling reasons for considering
randomized controlled trial, a non-randomized controlled
bilateral implantation in very young children. First, a child
trial, a cohort study or a case-control study, the primary
who is born deaf must use the information conveyed by the
criterion for methodological strength or quality is the extent
implant to learn spoken language. The rate of incidental
to which differences between intervention and control groups
language learning is likely to be accelerated in children who
can be attributed to the intervention studied.” (CMS, 2005.)
receive two implants rather then just one because they will
In the case of cochlear implantation for severe-to-profound
be better able to hear speech in everyday situations (most of
hearing loss there is no uncertainty regarding the causal
which are noisy) and to localize to relevant communication
relationship between implantation and auditory outcomes,
signals in their environment. Moreover, children who have
and there is no placebo effect in profound deafness.
two implants are likely to experience greater academic
Therefore, RCTs are not necessary (Glasziou et al, 2007).
success as they progress through school and encounter
increased challenges in their listening and learning
Moreover, a poorly conducted RCT may provide less useful
environment. Second, providing sound input to both ears in
information than a well-designed, prospective observational
a young deaf child assures that sound is processed on both
study. Low incidence conditions require multicenter studies
sides of the brain. Thus, the right and left auditory cortices can
to accrue a sufficient number of patients. An RCT thus may
develop in a more normal sequence. If a child is implanted
introduce bias as a result of unmeasured differences between
on only one side, the parts of the brain that would have been
centers. In any controlled trial, whether randomized or not,
stimulated by the non-implanted ear will not develop, and
the control subjects must not cross over into the treatment
eventually plasticity will be greatly diminished (Sharma et al,
group. For this reason, a study comparing unilateral versus
2005). Thus, bilateral implants have important implications
bilateral implantation in children with prelinguistic deafness
for the development of the auditory system in a young deaf
is not feasible in the United States. Because of research
child with congenital or early acquired deafness.
findings regarding the “sensitive period” for maximal neural
plasticity in young children (Peters et al, 2007), parents
Based upon the current research evidence, the medical
are unwilling to risk losing the potential benefit of bilateral
community has accepted that facilitating binaural hearing
implantation for the sake of research. For the same reason,
through bilateral cochlear implantation in children should
some cochlear implant surgeons believe that it is unethical to
be considered medically necessary in the absence of other
withhold bilateral implantation from appropriate candidates.
—continued on next page—
132 Advanced Bionics® Auditory Research Bulletin 2007
New Study Init iat i ve
Early Language and Spatial Hearing Skills in Young Implanted Children
Ruth Litovsky, Ph.D.
Tina Grieco, Ph.D.
Jenny Saffran, Ph.D.
•Do children who are implanted bilaterally at a young age
learn language differently than children who use a single
implant?
University of Wisconsin, Madison, WI, USA
•Do children who are implanted bilaterally at a young
age develop age-appropriate spatial hearing sooner than
children who receive bilateral implants at older ages?
For children who are born with profound hearing loss, the
electrical stimulation provided by a cochlear implant (CI)
To address these questions, children who were born
facilitates the development of oral language as well as other
with profound hearing loss and received unilateral or
auditory skills. Recently, there has been a clinical trend to
bilateral cochlear implants by 2.5 years of age will
provide CIs to children with severe-to-profound hearing
be enrolled in the study. There will be three groups:
loss at very young ages (around 12 months) with the goal of
(1) users of one CI, (2) users of bilateral CIs who received
providing electrical input to these children at the age when
their second CI at the same time or within several
language typically develops. Although many standardized
months of the first device, and (3) users of bilateral CIs
measures have shown that children who receive cochlear
who received the second CI within a year of the first
implants earlier have better speech and language outcomes,
CI. All subjects will be evaluated between the ages of
the mechanisms by which these children learn language
24-36 months, after having at least 12 months of auditory
has not been studied extensively. In addition, many of these
input. Children with bilateral CIs will have at least six months
children are receiving a second CI either simultaneously
of bilateral experience at the time they are evaluated. The
or shortly after receiving their first CI. One impetus for this
tests include psychophysical measures of spatial hearing/
latter trend is to stimulate the development of binaural
localization, and two different psychophysical measures used
processes (including spatial hearing) at a young age. Some
extensively in typically-developing children to assess word-
of the research questions that are currently being addressed
learning skills as well as speech recognition abilities (in quiet
include:
and in noise). The results will provide insight into how early
•Do children who have a history of deafness and who are
learning oral language through the use of cochlear implants
learn language similarly to children who have normal
hearing?
implantation affects the development of spatial hearing and
language-learning skills in an attempt to guide clinicians and
parents in making appropriate treatment choices for young
children with profound hearing loss.
—continued from previous page—
Previous
bilateral
of children with one implant. This study also will examine
implantation also have been limited by the wide range of
observational
studies
of
pediatric
predictors of bilateral performance, similar to analyses
ages at which subjects were implanted as well as variability
conducted in studies where children receive one implant
in the period between implants. In addition, the age at which
(e.g., age at implant, preimplant hearing variables, and
infant hearing loss is identified and cochlear implantation
family variables).
is first performed has been decreasing. There have been
no prospective multicenter studies involving very young
References
deaf children who received two implants simultaneously
Centers for Medicare and Medicaid Services. Decision Memo for Cochlear
Implantation (CAG 00107N). April 4, 2005. http://www.cms.hhs.gov/mcd/
viewdecisionmemo.asp?id=134 (accessed 25 October, 2007).
or separated by a short period of time. Taking these issues
into consideration, the pediatric bilateral study sponsored
by Advanced Bionics (described at page 134 of this bulletin)
will use a prospective within-subjects design in which the
development of auditory skills is tracked over a period of
two years. Relying on historical data, this study will compare
performance of these bilaterally implanted children to that
Glasziou P., Chalmers I., Rawlins M, McCulloch P. (2007) When are randomized
trials unnecessary? Brit Med J 334:349-350.
Peters BR, Litovsky R, Parkinson A, Lake J. (2007) Importance of age and
postimplantation experience on speech perception measures in children with
sequential bilateral cochlear implants. Otol Neurotol 28:649-657.
Sharma A, Dorman MF, Kral A. (2005) The influence of a sensitive period on
central auditory development in children with unilateral and bilateral cochlear
implants. Hear Res 203:134-143.
Bilateral Implantation 133
New Study Init iat i ve
Auditory Skill Development in Young Deaf Children with Bilateral HiRes 90K Implants:
Prospective Multicenter Study in North America
Participating Centers and Clinics in North America
Based upon the preceding considerations presented by Peters
Beth Israel-New York Eye and Ear Cochlear Implant Center,
New York, New York
and Black, the Advanced Bionics-sponsored study will be
Boys Town National Research Hospital, Omaha, Nebraska
months at time of surgery, who receive two implants in the
California Ear Institute/Let Them Hear Foundation, Palo Alto,
California
same operation or in two different surgeries with the initial
Carle Foundation Hospital, Urbana, Illinois
Children’s Healthcare of Atlanta, Atlanta, Georgia
Children’s Hospital of Alabama, Birmingham, Alabama
Children’s Hospital of Eastern Ontario, Ottawa, Ontario
Dallas Otolaryngology, Dallas, Texas
Glenrose Rehabilitation Hospital, Edmonton, Alberta
tracking outcomes in a cohort of 60 children, ages 12 to 36
fitting of the devices separated by no more than six months.
Acquisition of auditory milestones and speech recognition
skills, which underpin the development of spoken language,
will be assessed on a battery of outcome measures typically
used to quantify implant benefits. For example, tests will
include the Infant-Toddler Meaningful Auditory Integration
Scale (IT-MAIS), Early Speech Perception Test (ESP), and
House Ear Institute CARE Center, Los Angeles, California
Multisyllablic Lexical Neighborhood Test (MLNT), and the
Massachusetts Eye and Ear Infirmary, Boston, Massachusetts
Children’s Realistic Intelligibility and Speech Perception
Medical College of Wisconsin, Milwaukee, Wisconsin
Test (CRISP) (Garadat & Litovsky, 2005). Historically, these
Midwest Ear Institute, Kansas City, Missouri
measures have been sufficient to demonstrate the large
New York University, New York, New York
benefits associated with unilateral implantation.
Phoenix Children’s Hospital, Phoenix, Arizona
Tampa Bay Hearing and Balance Center, Tampa, Florida
The development of measures that can capture the functional
The Children’s Memorial Hospital, Chicago, Illinois
benefits of bilateral implantation is in its early stages. In this
University of California, San Francisco, California
study, communicative performance will be assessed through
University of Minnesota, Minneapolis, Minnesota
parental proxy with the FAPCI (Functioning After Pediatric
Vanderbilt University, Nashville, Tennessee
Cochlear Implantation), a tool developed by Frank Lin and
Washington University, St. Louis, Missouri
colleagues at Johns Hopkins University (Lin et al, 2007). In
addition, health-related quality of life will be assessed by
asking the parents to rate their feelings about their child’s
overall health and development status using visual-analog
scales.
Performance will be tracked after 3, 6, 12, 18, and 24 months
of device use. The study also will identify variables that may
predict the degree of implant benefit (e.g., age at implant,
pre-implant hearing thresholds, communication mode,
family socioeconomic status, post-implant aided thresholds,
simultaneous initial fitting of both devices versus sequential
fitting separated by up to six months).
References
Garadat S, Litovsky R. Methods for evaluation of bilateral advantages in children
with cochlear implants and/or hearing aids. Poster presented at
the Conference on Implantable Auditory Prosthesis, Pacific Grove, CA, USA.
30 July-4 August, 2005.
Lin FR, Ceh K, Bervinchak D, Riley A, Miech R, Niparko JK. Development
of a communicative performance scale for pediatric cochlear implantation.
Paper presented at the 11th International Conference on Cochlear Implants in
Children, Charlotte, North Carolina, USA. 11-14 April, 2007.
Peters BR, Litovsky R, Parkinson A, Lake J. (2007) Importance of age and
postimplantation experience on speech perception measures in children with
sequential bilateral cochlear implants. Otol Neurotol 28:649-657.
Ryugo DK, Kretzmer EA, Niparko JK. (2005) Restoration of auditory nerve
synapses in cats by cochlear implants. Science 310:1490-1492.
134 Advanced Bionics® Auditory Research Bulletin 2007
New Study Init iat i ve
Bilateral Cochlear Implant Benefit in Young Deaf Children
Craig Buchman, M.D.
Carolyn J. Brown, M.S.
Hannah Eskridge, M.S.P.
Holly Teagle, Au.D.
Carlton Zdanski, M.D.
John Niparko, M.D.
Stephen Bowditch, M.A.
Jill Chinnici, M.A.
Jennifer Yeagle, M.Ed.
The Listening Center at Johns Hopkins, Baltimore, MD, USA
University of North Carolina, Chapel Hill, NC, USA
An ongoing multicenter longitudinal study (John Niparko,
Mark 3 (HUI-3), the EQ-5D, and the Quality of Well-Being
Principal Investigator) is using a hierarchical test battery to
(QWB). The HUI-3 is a 15-item multiattribute measure for
assess acquisition of preverbal communication skills, speech
classification of health status based on the domains of vision,
recognition, speech production, and language skills in
hearing, speech, ambulation, dexterity, emotion, cognition,
children with normal hearing and children with one cochlear
and pain (Horsman, et al, 2003). Each domain has five or six
implant (Childhood Deafness after Cochlear Implantation
levels of ability or disability. Used as a profile or descriptive
[CDaCI] [Eisenberg et al, 2006]). This new collaborative study
measure, the HUI-3 has been found to be responsive to
between the University of North Carolina and Johns Hopkins
change in health status. The EQ-5D is a generic instrument
University is examining bilateral benefit in young children
developed by the EuroQol Group (Brooks et al, 2003). It is a
ages 12-30 months who receive two cochlear implants at the
15-item measure of five domains of health: mobility, self-care,
same time or separated by no more than six months. Children
usual activities, pain/discomfort, and anxiety/depression,
will be followed for at least two years following implantation.
with three response levels available for each domain. It also
The hierarchical test battery from the CDaCI study will be
includes a Visual Analog Scale (VAS). The VAS is presented
used as well as several parental proxy tools that will evaluate
as a 20 cm vertical “thermometer” scale anchored at 0
functional communication, general health, and quality of life.
(worst imaginable health) and 100 (best imaginable health).
The performance of the children with bilateral implants will
Parents mark the scale at a point corresponding to their
be compared to the two cohorts in the CDaCI study. A novel
children’s current health. The Quality of Well-Being (QWB)
aspect of the study is inclusion of a video analysis procedure
is a comprehensive instrument that was developed for use as
to assess the acquisition of preverbal language skills (Tait and
a preference measure. The QWB includes questions about
Luttman, 1994; 1997). These preverbal skills are precursors
functioning in four domains: self-care, mobility, physical
of early language development in children with normal
activity, and social activity (Kaplan et al, 1998).
hearing and have been shown previously to predict later
speech production and perception skills in young children
with cochlear implants.
The development of communicative performance will be
assessed with the FAPCI (Functioning after Pediatric Cochlear
Implantation) (Lin et al, 2007). This new parental proxy
instrument is a 23-item paper-and-pencil questionnaire
that measures communicative performance after cochlear
implantation in children 2 to 5 years of age. It is designed
to complement speech and language measures. The
questionnaire items incorporate situational and behavioral
contexts in order to capture the child’s communicative
competence in the real world, thus providing a measure of
implant effectiveness.
Relatively little is known about changes in quality of life
following bilateral implantation in young children. Therefore,
the general well-being, development, and health of the
children will be evaluated using the Health Utilities Index
References
Brooks R, Rabin RE, de Charro F, eds. (2003) The measurement and valuation of
health status using EQ-5D: a European perspective. Dordrecht, The Netherlands:
Kluwer Academic Publishers.
Eisenberg LS, Johnson KC, Martinez AS, Cokely CG, Tobey EA, Quittner AL, Fink
NE, Wang NY, Niparko JK, CDaCI Investigative Team. (2006) Speech recognition
at 1-year follow-up in the childhood development after cochlear implantation
study: Methods and preliminary findings. Audiol Neurotol 11:259-268.
Horsman J, Furlong W, Feeny D, Torrance G. The Health Utilities Index (HUI):
concepts, measurement properties and applications. (2003) Health and Qual
of Life Outcomes http://www.hqlo.com/content/1/1/54 (accessed 24 October
2007).
Kaplan RM, Ganiats TG, Sieber WJ, Anderson JP. (1998) The Quality of WellBeing Scale: critical similarities and differences with SF-36. Int J Qual Health
Care 10:463-465.
Lin FR, Ceh K, Bervinchak D, Riley A, Miech R, Niparko JK. Development
of a communicative performance scale for pediatric cochlear implantation.
Presented at the 11th International Conference on Cochlear Implants in
Children, Charlotte, North Carolina. 11-14 April, 2007.
Tait M, Luttman ME. (1997) The predictive value of measures of preverbal
communicative behaviors in young deaf children with cochlear implants.
Ear Hear 18:472-478.
Tait M, Luttman ME. (1994) Comparison of early communicative behavior
in young children with cochlear implants and with hearing aids. Ear Hear
15:352-361.
Bilateral Implantation 135
New Study Init iat i ve
ABC Pilot Study for a Randomized Controlled Trial comparing Bilateral and Unilateral
Cochlear Implantation in Children
A. Quentin Summerfield, Prof., Ph.D.
University of York, York, United Kingdom
Health care in the United Kingdom is one of increasingly
of the acquisition of spoken language, (3) enhanced quality
stringent governance. Consequently, new treatments and
of life, and (4) an acceptable cost-effectiveness ratio? The
relaxations of candidacy criteria for existing treatments
aim is to conduct as rigorous a study as possible because
must be supported by demonstrated effectiveness and cost-
experience indicates that NICE attaches it highest priority
effectiveness if they are to be provided routinely. The approval
to high quality evidence (Dakin et al, 2006). In order to
process is overseen by the National Institute for Health and
conduct a rigorous study, the following steps will be taken:
Clinical Excellence (NICE). First, treatments are referred by
the Department of Health to NICE for appraisal. Thereafter,
NICE commissions a systematic review of the evidence for
clinical effectiveness and also commissions the synthesis of
data into a cost-effectiveness model. Finally, NICE weighs
the review together with the model and issues a mandatory
guidance on the provision of the treatment in England and
Wales.
Currently, NICE is undertaking an appraisal of both unilateral
and bilateral cochlear implantation (NICE, 2006). NICE likely
will find sufficient published evidence to form a judgement
about the cost effectiveness of unilateral implantation
for adults and children (e.g., UKCISG, 2004; Barton et al,
2006). However, data on the cost effectiveness of bilateral
implantation, particularly for children, are sparse. In previous
situations where robust evidence is not available, NICE has
deferred judgement, pending the conclusion of appropriately
powered randomized controlled trials.
In anticipation of that outcome, a consortium consisting
of clinicians in Birmingham, Kilmarnock, Manchester,
and Nottingham and researchers in Health Sciences and
Psychology at the University of York has designed a series
of randomized controlled trials to reduce uncertainty about
the costs and benefits of bilateral compared with unilateral
implantation. Advanced Bionics is supporting the first of
these trials, a pilot study designed to confirm the feasibility of
methods and to inform the power calculation for a full clinical
trial. In the pilot study, 40 children will be allocated randomly
to undergo either unilateral implantation (a single implant
in one ear with an acoustic hearing aid in the contralateral
ear) or bilateral implantation (two implants, one in each ear).
The children will be 12-36 months of age, the usual age for
implantation in the United Kingdom. They will be followed
for two years to answer four questions—when compared with
•Children will be allocated to each group by a random
process, akin to tossing a coin. This is the only process
that can control all of the variables, both known and
unknown, that might otherwise vary between patients
who seek, or are offered, one treatment rather than
another (Kunz & Oxman, 1998).
•To avoid bias, randomization will be performed by a webbased computer system that is managed independently
of the clinicians who provide implantation and the
researchers who evaluate outcomes. Randomization
will be stratified according to the average hearing level
in the “second ear”—that is, the ear that would receive
the acoustic hearing aid if the child were randomized to
the unilateral group. That step will eliminate the bias that
would arise if a disproportionate number of children with
aidable hearing were allocated by chance to one group
or the other.
•The trial is pragmatic in that it addresses the potential
effects of different policies for treatment rather than the
effects of the treatment itself. Accordingly, data will be
analyzed in accordance with “intention to treat”—that
is, all children will be followed and crossovers from one
group to the other will be ignored.
•Because implants and acoustic hearing aids have visible
external components, patients, clinicians, and researchers
cannot be blinded to the treatment received by each child.
Thus, a potential exists for expectation bias, in which,
knowingly or unknowingly, better outcomes are recorded
for the group that is expected to do better. To avoid this
bias, video recordings of the tests of spatial hearing
and language will be edited to disguise the treatment
received by each child. Performance then will be scored
by independent panels of judges. In addition, parents of
hearing children will judge the quality of life of the study
children from anonymous vignettes that describe the
functional status, but not the treatment received.
•Key outcome measures will be specified in advance, to
avoid the bias inherent in selecting the outcomes that
happen to be significant retrospectively.
•Adherence to ethical and research protocols will be
scrutinized by independent steering and data-monitoring
committees.
unilateral implantation, does bilateral implantation lead to
First results for the pilot study are anticipated within 18 to
(1) better spatial hearing, (2) an improved rate and/or extent
30 months of commencement in Autumn 2007.
—continued on next page—
136 Advanced Bionics® Auditory Research Bulletin 2007
New Study Init iat i ve
A Longitudinal Study of Young Children who Receive Bilateral and Unilateral
Advanced Bionics Cochlear Implants
Ruth M. Reeder, M.A.1
Jamie H. Cadieux, Au.D.2
Jerrica L. Kettel, Au.D.2
Jill B. Firszt, Ph.D.1
1 Washington University School of Medicine, St. Louis, MO, USA
2 St. Louis Children’s Hospital, St. Louis, MO, USA
studying (1) the ability to understand words and sentences
at reduced presentation levels in quiet and in noise, (2) the
rate of verbal language acquisition, and (3) the ability to
attend to an auditory stimulus. Two groups of children will
be implanted with Advanced Bionics devices between 12
For children with significant hearing loss, the bilateral fitting
and 30 months of age. One group will receive two implants
of hearing aids has been standard clinical care for at least
(simultaneous surgeries or within a short period of time) and
two decades. There are several strong reasons for providing
one group will be implanted with one device. The children
acoustic input to both ears. Bilateral input contributes to
will be matched for age at implant and socioeconomic status.
improved abilities to listen in noise and to localize sound in
All children will have rehabilitation plans that emphasize
the environment. Moreover, providing input to only one ear
development of communication and language using audition
modifies auditory cortex response patterns in ways that differ
and speech. Assessments will be administered preimplant
from bilateral hearing. For many years, unilateral cochlear
with developmentally appropriate speech recognition,
implantation has been the primary recommendation for
speech production, and language measures repeated after
bilateral severe-to-profound hearing loss. Early on, clinicians
1, 3, 6, 9, 12, 18, 21, and 24 months of device use. Additional
were hesitant to recommend two devices because of the
assessments will occur thereafter every six months until the
unknowns in cochlear implantation, such as limited data
child is 5 to 5 ½ years of age. Speech recognition measures
concerning the development of auditory, speech, and
as well as reaction time measures will be administered in the
language abilities as well as ongoing improvements in
unilateral right, unilateral left, and bilateral conditions for the
technology that perpetuated a “wait and see” mind set.
bilaterally implanted children and in the implanted ear for
However, at this time, the benefits of listening with two
the unilaterally implanted children.
implants have been documented in adults. On the other hand,
the number of bilaterally implanted children is still relatively
We predict that children who receive bilateral cochlear
small and results to date are limited even though worldwide,
implants will demonstrate improved auditory and attention
a greater number of children have received bilateral implants
skills in the bilateral compared to unilateral conditions and
than adults (approximately 3000 recipients in total, of which
that their speech and language skills will be better and will
two-thirds are children). In addition, little is known about
develop faster than the unilaterally implanted children.
the effects of bilateral implantation in very young children,
Results from this study will be used (1) to provide new
particularly those who receive devices simultaneously or
information about the outcomes of bilateral implantation in
sequentially within a short time period.
young children for families, clinicians, and surgeons who
are considering bilateral implantation, and (2) to identify
The primary aim of this longitudinal study is to evaluate the
measures of bilateral benefit in children that could be
effects of bilateral implantation in very young children by
incorporated into clinical protocols.
—continued from previous page—
References
Barton GR, Stacey PC, Fortnum HM, Summerfield AQ. (2006) Hearing-impaired
children in the United Kingdom IV: Cost-effectiveness of paediatric cochlear
implantation. Ear Hear 27:575-88.
Dakin HA, Devlin NJ, Odeyemi IA. (2006) “Yes,” “No,” or “Yes, but”?
Multinomial modelling of NICE decision-making. Health Policy 77:352-67.
National Institute for Health and Clinical Excellence. (2006) Cochlear Implants:
Final Scope. http://guidance.nice.org.uk (accessed 3 August 2007).
United Kingdom Cochlear Implant Study Group. (2004) Criteria of candidacy
for unilateral cochlear implantation in post-lingually deafened adults II: costeffectiveness analysis. Ear & Hear 25:336-60.
Kunz R, Oxman AD. (1998) The unpredictability paradox: a review of empirical
comparisons of randomised and non-randomised clinical trials. Brit Med J
317:1185-1190.
Bilateral Implantation 137
Pediatric Bilateral Implantation: Two Case Studies
Donald M. Goldberg, Ph.D.
Cleveland Clinic Foundation, Cleveland, OH, USA
“. . .these two children illustrate
that auditory skills and spoken language
competency can develop quickly,
independent of the time interval
between implantation of each ear.”
With the adoption of Universal Newborn Hearing
Screening programs, hearing impairment is being
identified at much younger ages. The prompt fitting
of personal hearing aids and FM systems and the
implementation of early intervention programs
that emphasize the parents’ role as a child’s first
and most important teacher (Pollack, 1970) have
resulted in impressive outcomes for many children.
And for children for whom personal hearing aids do
not amplify the entire speech sound spectrum sufficiently, cochlear implants have become an efficacious
option. Cleveland Clinic has been providing cochlear
implant services since the early 1980s. As of early
June 2007, 79 individuals have been implanted bilaterally, including 53 children. Most of the surgeries
have been sequential, although 9 adults and 10 children received both devices during the same surgery.
For children, the postsurgical intervention philosophy is to provide auditory-verbal and auditorybased therapy that stresses the development of
spoken language through listening (Pollack et al,
1997). Typically, implanted children participate in
periodic (weekly, biweekly, or monthly) hour-long
auditory-verbal or auditory-based therapy sessions.
In addition, the implant team focuses on “empowering” parents so that they are in the position to reinforce listening for spoken language development at
home. Two children who were bilaterally implanted
with Advanced Bionics devices provide some insight
into the benefits of sequential bilateral implantation
supported by continuous auditory-based training
and parental reinforcement.
Case 1. The first child is an example of one of the
many “super” listeners and spoken language users
who began therapy as an infant and who progressed
remarkably well with her first device. She was 18
months old when her first implant was activated.
Because she experienced “unusual” sensations with
her first implant after about five years of use, the
Cleveland Clinic team recommended that she
receive a second cochlear implant in the contralateral
138 Advanced Bionics® Auditory Research Bulletin 2007
ear, with the expectation of eventually explanting
and reimplanting the first ear. She was implanted
with the second device at 7 years, 7 months of age.
The time interval between the two implants was
just under six years. By four to six months after
activation of the second device, the performance
in the second ear was equivalent to the performance with the first ear based upon Ling sound
identification, Early Speech Perception subtests
1 through 3, auditory connected discourse tracking,
and other measures. For example, her Test of Auditory Comprehension (TAC) score with the first
implant just before she received the second device
was 132/139 (September 2005). Six months after
the second surgery, the TAC score with the second
implant alone was 135/139 (August 2006). Her most
recent TAC score with both implants was 136/139
(April 2007). On the Listening Test conducted
in June 2006, her score was 73/75 (percentile rank
= 99). This youngster just completed the third grade
in a mainstream placement.
subject-dependent variables certainly contribute to
their remarkable progress, the team at Cleveland
Clinic believes that auditory-verbal and auditorybased therapy play a significant role in allowing
these children to reach their full hearing potential.
The team will continue to provide such therapy as
more children are bilaterally implanted so that children who are deaf can achieve even greater benefits
from their cochlear implants.
Acknowledgments
The Hearing Implant Program and Cochlear Implant Team at the Cleveland Clinic
includes Peter Weber, M.D. (Co-Director); Michael Scott, Au.D.; Joyce Crawford,
M.S.; Keiko Hirose, M.D.; Gordon Hughes, M.D.; Craig Newman, Ph.D.; Sharon
Sandridge, Ph.D.; Cynthia Gensur, Au.D.; Janet Fraser, M.A.; Lisa Wodzisz, Au.D.;
Amy Aylward, Au.D.; and Nancy Adamson, M.A.
References
Pollack D. (1970) Educational audiology for the limited hearing infant and
preschooler (Third edition). Springfield, IL: Charles C. Thomas Publishers.
Pollack D, Goldberg D, Caleffe-Schenck N. (1997) Educational audiology for the
limited hearing infant and preschooler: An auditory-verbal program. Springfield, IL:
Charles C. Thomas Publishers.
Case 2. The second child also was sequentially
implanted. In contrast to the first case, she was
implanted bilaterally with surgeries separated by
less than one year—at ages 2 years, 3 months and
3 years, 2 months. This child also is another “superstar” spoken language user whose performance
between the two ears was equivalent only one
month after the second device was activated (again
based upon Ling sound identification, Early Speech
Perception subtests 1 through 3, and tracking
data). Her TAC score with the first implant alone
was 81/139 (August 2005). After one year and two
months of using both implants together, her TAC
score was 114/139 (March 2007). This youngster now is attending a local public preschool with
hearing peers.
Taken together, these two children illustrate that
auditory skills and spoken language competency can
develop quickly, independent of the time interval
between implantation of each ear. Although other
Bilateral Implantation 139
Music Studies
With the introduction of HiResolution® Sound and HiRes Fidelity 120™
(HiRes 120)* processing options, CII Bionic Ear® and HiRes 90K™ implant
recipients now have access to greater temporal and spectral resolution of input
signals. Thus, the potential exists for improved music appreciation and overall
sound quality. Early results indicate that HiRes 120 improves the satisfaction
and frequency of listening to music, that music is significantly more pleasant
and distinct with HiRes 120, and that HiRes 120 may allow some users to
discriminate and identify musical instruments. New studies are underway
to investigate the psychophysical, perceptual, emotional, and developmental
benefits of music listening with HiRes 120 in adults and children.
The studies in this section summarize various investigations into the music
listening and sound quality benefits enjoyed by HiRes and HiRes 120 users.
*In the United States, an optional feature for adults only. See package insert for details.
141
Music Benefits in Adults Using HiRes with Fidelity 120:
Multicenter Study in North America
Percent responding “often” or “very often”
and “satisfied” or “very satisfied”
The majority of adult postlinguistically deafened
cochlear implant users can understand speech
remarkably well in quiet. Because the speech signal is
redundant, only limited spectral resolution is necessary (Shannon et al, 2004). On the other hand, the
ability to understand speech in noise and to enjoy
HiRes (baseline)
100
HiRes 120 @ 3 mos
80
60
43.4
41.2
40
28.2
19.2
20
0
Frequency
Satisfaction
Figure 1. Fifty-five subjects rated their music listening experience with standard HiRes and after using HiRes 120 for three
months. After three months using HiRes 120, there was a
13% shift in the proportion of adults reporting that they listen
to music “often” or “very often” compared to their baseline
ratings with standard HiRes. There was a 24% shift in the
proportion of subjects stating that they are “satisfied” or “very
satisfied” when listening to music using HiRes 120 compared
to standard HiRes.
10
HiRes (baseline)
HiRes 120 @ 3 mos
HiRes @ 3 mos
Mean Rating
8
6
6.0
6.8
6.1
5.8
5.0
5.2
4
2
0
Pleasantness
Distinctness
Figure 2. Mean ratings of music pleasantness and distinctness for 55 subjects with baseline HiRes, after three months
using HiRes 120, and after being reprogrammed with standard HiRes. Subjects rated the pleasantness and distinctness
of recorded instrumental music passages on a scale from 0
(extremely unpleasant/indistinct) to 10 (extremely pleasant/distinct). The music samples consisted of a variety of styles including classical, easy listening, and solo instrument passages.
The HiRes 120 ratings were significantly higher than baseline
HiRes (p < .05).
142 music remains poor (e.g., Mirza et al, 2003). This
outcome is not surprising, given the limited spectral
resolution delivered by previous generation implant
technology. With the introduction of HiResolution
(HiRes) and HiRes with Fidelity 120 (HiRes 120)
sound processing, CII and HiRes 90K recipients now
have access to greater temporal and spectral resolution. Thus, the potential exists for improved music
appreciation and overall sound quality. Results from
55 subjects who have used HiRes 120 indicate that
HiRes 120 improves the satisfaction and frequency
of listening to music (Figure 1), that music is significantly more pleasant and distinct with HiRes 120
(Figure 2), and that HiRes 120 may allow some
users to discriminate and identify musical instruments (Advanced Bionics, 2007; Koch et al, 2007).
The multicenter study is assessing both music appreciation and perception in a large sample of adults with
postlingual onset of severe-to-profound hearing loss
who have been implanted with CII or HiRes 90K
implants. There are two study groups. One group
consists of users of standard HiRes sound processing
who are evaluated at a baseline visit and subsequently
fit with HiRes 120. After three months of HiRes
120 experience, they are evaluated again. The second
group consists of HiRes 120 users with at least three
months of HiRes 120 experience who are evaluated
during a single test session. Music appreciation is
assessed with the Iowa Musical Background and
Appreciation Questionnaire (IMBAQ) (Gfeller and
Witt, 1998), a tool that has been used extensively
by Gfeller and colleagues at the University of Iowa.
This questionnaire evaluates musical training as well
as music enjoyment. Music perception is assessed
with the University of Washington-Clinical Assessment of Music Perception (UW-CAMP) test (Kang
et al, 2006; Nimmons et al, in press). This computerbased test battery evaluates pitch discrimination,
melody recognition, and instrument identification
(timbre). Music perception also is evaluated using
two subtests from the Appreciation of Music by
Cochlear Implantees (AMICI) test battery developed by Spitzer and colleagues (Cheng et al, 2007;
Spitzer et al, 2007). The results are analyzed to
compare music appreciation and perception and
to identify variables that predict the ability of CI
users to hear and enjoy music (for example, HiRes
Advanced Bionics® Auditory Research Bulletin 2007
Participating Centers and Clinics in North America
Arkansas Children’s Hospital, Little Rock, Arkansas
Beth Israel-New York Eye and Ear Cochlear Implant Center, New York, New York
Cleveland Clinic, Cleveland, Ohio
Columbia University Medical Center, New York, New York
Hearts for Hearing, Oklahoma City, Oklahoma
House Ear Clinic, Los Angeles, California
Houston Ear Research Foundation, Houston, Texas
Johns Hopkins University, Baltimore, Maryland
Mayo Clinic, Rochester, Minnesota
Medical College of Wisconsin, Milwaukee, Wisconsin
Medical University of South Carolina, Charleston, South Carolina
New York University, New York, New York
Ottawa Hospital-Civic Campus, Ottawa, Ontario
Stanford University, Palo Alto, California
Sunnybrook & Women’s College Health Sciences Centre, Toronto, Ontario
University of California, San Francisco, California
University of Michigan, Ann Arbor, Michigan
University of North Carolina, Chapel Hill, North Carolina
University of Pennsylvania, Philadelphia, Pennsylvania
University of Virginia, Charlottesville, Virginia
Vanderbilt University, Nashville, Tennessee
Washington University, St. Louis, Missouri
Yale University, New Haven, Connecticut
vs. HiRes 120 sound processing, duration of deafness, age at onset of deafness, music training, speech
perception ability).
The study is unique in that music benefit is evaluated
using measures that assess a range of music experience. Assessment of music appreciation and “realworld” music perception (IMBAQ and AMICI)
provides evidence of the music benefits realized
by implant users in everyday life. Psychophysical
assessment of the perception of musical elements
(UW-CAMP) indicates how well current implant
technology represents the basic acoustic characteristics of music, as well as providing insight into how
that representation might be improved in future
sound coding strategies. Assessing music benefit
across a range of experience also is important because
some subjects might report significant enjoyment
in listening to music even though they have difficulty on the perception tests. Other subjects might
perform well on the psychoacoustic tests, but derive
little everyday pleasure from listening to music. The
data from this study will indicate the relationships
among subjective and objective music assessments,
speech recognition, and demographic variables.
References
Advanced Bionics. (2007) HiRes with Fidelity 120 Clinical Results (white paper).
Valencia, CA: Advanced Bionics.
Cheng M-Y, Mancuso DM, Spitzer JB. Reliability of Appreciation of Music by
Cochlear Implantees (AMICI). Poster presented at the American Academy of
Audiology, Denver, CO, USA. 18-21 April, 2007.
Gfeller K, Witt S. Iowa Musical Background and Appreciation Questionnaire
(IMBAQ). Iowa City: IA, University of Iowa, 1998.
Kang R, Liu G, Drennan W, Lognion J, Ruffin C, Worman T, Yueh B, Rubinstein J.
Development and validation of the University of Washington (UW) Music Test for
cochlear implantees. Poster presented at the American Academy of Otolarynology,
Toronto, Canada. 17-20 September, 2006.
Koch DB, Osberger MJ, Quick A. Increased spectral resolution enhances listening
benefits in cochlear implant users. Poster presented at the Association for Research in
Otolaryngology, Denver, CO, USA. 10-15 February, 2007.
Mirza S, Douglas SA, Lindsey P, Hildreth T, Hawthorne M. (2003) Appreciation of
music in adult patients with cochlear implants: a patient questionnaire. Coch Impl
Int 4(2):85-95.
Nimmons G, Kang R, Drennan W, Lognion J, Ruffin C, Worman T, Yueh B, Rubinstein
J. Clinical assessment of music perception in cochlear implant listeners. Otol
Neurotol, in press.
Shannon RV, Fu QJ, Galvin J. (2004) The number of spectral channels required
for speech recognition depends on the difficulty of the listening situation. Acta
Otolaryngol (Suppl)552:1-5.
Spitzer JB, Mancuso DM, Cheng M-Y. Beta version of Appreciation of Music
by Cochlear Implantees (AMICI). Poster presented at the American Academy of
Audiology, Denver, CO, USA. 18-21 April, 2007.
Music Studies 143
Music Perception in Adult HiRes 120 Users
Roberto Filipo, Prof., M.D.
Deborah Ballantyne, Ph.D.
Patrizia Mancini, M.D.
Chiara D’Elia, M.D.
University of Rome La Sapienza, Rome, Italy
The aim of this study was to verify the hypothesis
that the increased spectral resolution delivered by
HiRes with Fidelity 120 sound processing (HiRes
120) leads to improvement in the perception of
music. The increased spectral resolution provided by
HiRes 120 is implemented using current steering
where multiple pitch precepts may be created by
stimulating adjacent electrode contacts simultaneously with proportional distribution of current to
each contact. The study group consisted of 12 postlinguistically deafened adults who were participating
Standard HiRes
10
9
Mean Quality Rating
HiRes 120
BTE
PSP
8
7
6
5
4
3
2
1
0
Classical
Jazz
Soul
Classical
Jazz
Soul
Figure 1. Mean music quality ratings for standard HiRes and
after four months of HiRes 120 experience for three musical
genres.
Jazz
Quality Rating Difference
6
Classical
Soul
5
4
3
2
1
0
0
1
2
3
4
5
6
7
8
9
Mean Number of Spectral Channels per Electrode Pair
Figure 2. Difference in music quality ratings (four months HiRes
120 minus standard HiRes) as a function of average number of
spectral channels per electrode pair for three musical genres.
144 in a study comparing standard HiRes to HiRes 120
sound processing. All subjects were experienced
HiRes users. Five were programmed with HiRes 120
on the body-worn PSP processor, and seven were
upgraded from an earlier BTE processor to HiRes
120 on the Harmony processor. Mean age was 47
years, mean duration of deafness was 74 months,
and the duration of standard HiRes use prior to the
study was 38 months. An evaluation of the number
of spectral channels perceived by each subject was
made before switching to HiRes 120, and at one,
two, and four months after being fit with HiRes
120. After loudness balancing and pitch ranking
electrode pairs (3-4, 8-9, 13-14), an adaptive paradigm was used to estimate the number of intermediate pitch percepts that could be heard for each pair
when current steering was implemented. Those data
were used to estimate the potential number of spectral channels for each electrode pair. Assuming that
the average numbers of spectral channels for these
three electrode pairs were representative of the entire
16-contact array, the potential total numbers of
spectral channels could be estimated. Using three
tests, music perception was assessed with standard
HiRes and after being fit with HiRes 120. First,
subjects listened to three musical tracks—Classic
(dynamic), Jazz (mellow), and Soul (vocal)—and
rated the quality of perception on a scale from
1 to 10, including an appreciation of loudness and
rhythm. Second, subjects underwent a test of recognition of musical instruments (timbre) based on
wind (trumpet, clarinet, flute), string (cello, guitar),
and percussion (piano) instruments. The last test
was pitch ranking in which subjects were asked
to identify tones and semitones in ascending and
descending order of a C-major scale played on a
keyboard instrument.
The ratings questionnaire showed a consistent
tendency for improvement in perception of music
quality when using HiRes 120 compared to standard HiRes (Figure 1). After four months of HiRes
120 experience, the differences were statistically
significant for the PSP users (p < 0.001), but not
for the BTE users. No significant differences were
found between HiRes (89% correct) and HiRes 120
(92% correct) for discrimination between the three
types of music (Classic, Jazz, Soul). For the individual musical genres, the Classic tract (dynamic)
showed the greatest improvement in quality with
Advanced Bionics® Auditory Research Bulletin 2007
HiRes 120. This track was the most complex of
the three musical genres and may have been most
enhanced by increased spectral resolution. Figure 2
shows relationships between mean number of spectral channels and the difference in ratings between
the four-month HiRes 120 rating and the standard
HiRes rating for the three musical genres. These
data suggest that the greater the number of spectral
channels, the better the music perception, particularly for the classical music.
The ability to hear rhythm (temporal processing)
seemed to be fairly easy for all three genres of
music, and no significant differences were found
between the two strategies. Scores on the timbre
test were initially relatively low with HiRes 120 but
improved slightly over time. However, these findings
may have been biased by subjective variables such
as knowledge of music and by familiarity with the
task, particularly by the end of the protocol. Further
studies of music perception in cochlear implant users
will be based on the Perceptual Attributes of Timbre
rating scales (instead of recognition tasks) to reduce
the bias associated with previous musical knowledge
or experience. In summary, improvement in music
perception with HiRes 120 was reflected primarily
in the ratings questionnaire. However, more specific
tests are required to assess appreciation of timbre,
preferably by using rating scales that would not be
biased by knowledge of music.
New Study Init iat i ve
Music Perception Using HiResolution Sound Processing
Jay Rubinstein, M.D., Ph.D.
Ward R. Drennan, Ph.D.
Clinical Assessment of Music Perception (UW-CAMP)
University of Washington, Seattle, WA, USA
administerable test that can measure many levels of music
suggest that this test may be a sensitive, efficient, and selfperception in cochlear implant users (Kang et al, 2006;
Music plays important psychological and social roles in
Nimmons et al, in press). The UW-CAMP is a computerized
everyday life. However, despite its importance to quality
test comprised of pitch discrimination, melody identification,
of life, music listening has received far less attention than
and timbre identification assessments. The pitch subtest
speech perception in cochlear implant research. Moreover,
uses a two-alternative forced-choice adaptive procedure to
while performance on standard speech perception tests has
determine a threshold interval for discrimination of complex
improved with advancements in sound processing, studies
pitch direction changes. The melody subtest assesses
of music perception and appreciation typically have shown
recognition of 12 isochronous melodies, and the timbre
poor results among implant recipients (for example, Leal et
subtest evaluates recognition of eight musical instruments.
al, 2003; McDermott, 2004).
UW-CAMP test results will be compared with psychophysical
measures of spectral resolution and speech perception in
With the introduction of HiRes and HiRes Fidelity 120
various noise conditions in the same subjects. This research
(HiRes 120) sound processing, CII and HiRes 90K recipients
is part of a larger effort to validate the UW-CAMP for clinical
now have access to greater temporal and spectral resolution.
use.
While speech can be understood in the presence of severe
degradation of spectral and temporal cues, music recognition
References
and appreciation are compromised by even mild degradation
Kang R, Liu G, Drennan W, Lognion J, Ruffin C, Worman T, Yueh B, Rubinstein
J. Development and validation of the University of Washington (UW) Music
Test for cochlear implantees. Poster presented at the American Academy of
Otolarynology, Toronto, Canada. 17-20 September, 2006.
(Shannon, 2005). Initial results from subjects who have used
HiRes 120 indicate that, in fact, it improves the satisfaction
and frequency of listening to music and that it may allow
some users to discriminate and identify musical instruments.
Leal MC, Shin YJ, Laborde ML, Calmels MN, Verges S, Lugardon S, Andrieu
S, Deguine O, Fraysse B. (2003) Music perception in adult cochlear implant
recipients. Acta Otolaryngol 123(7):826-835.
McDermott HJ. (2004) Music perception with cochlear implants: a review.
Trends Amplif 8(2):49-82.
This study will evaluate in greater detail music perception
in standard HiRes and HiRes 120 users with the music
perception test battery developed by Rubinstein and
colleagues. Early data from the University of Washington
Nimmons G, Kang R, Drennan W, Lognion J, Ruffin C, Worman T, Yueh B,
Rubinstein J. Clinical assessment of music perception in cochlear implant
listeners. Otol Neurotol, in press.
Shannon RV. (2005) Speech and music have different requirements for spectral
resolution. Int Rev Neurobiol 70:121-134.
Music Studies 145
Development of Musical Behavior in Children with
Cochlear Implants
Deborah Vickers, Ph.D.1,3
Hilal Dincer, M.Sc.2
Patrick Boyle, M.Sc.1
Gunay Kirkim, Ph.D.4
Levent Olgun, M.D.5
Armagan Incesulu, M.D.6
Esra Yucel, Ph.D.7
Sevginar Onder, M.Sc.8
Nurdan Aygener, M.Sc.9
1 Advanced Bionics, Europe
2 Erisci Elektroniks, Ankara, Turkey
3 University College London, London, United Kingdom
4 Dokuz Eylul University Hospital, Izmir, Turkey
5 SB Izmir Bozyaka Hospital, Izmir, Turkey
6 Osman Gazi University Hospital, Eskisehir, Turkey
7 Hacettepe University Hospital, Ankara, Turkey
8 SB Ankara Numune Hospital, Ankara, Turkey
9 SB Ankara Diskapi Hospital, Ankara, Turkey
Adult cochlear implant users commonly complain
that they do not enjoy listening to music (Fujita
and Ito, 1999; Leal et al, 2003; McDermott, 2004;
Mirza et al, 2003). Although rhythm perception
usually is quite good, cochlear implant users generally have difficulty hearing pitch or melody (Kong
et al, 2004). Nonetheless, some parents comment
that implanted children do, in fact, enjoy listening
to music and singing. It may be the case that children who acquire hearing with a cochlear implant
may encode sounds differently than adults who are
implanted after losing their hearing. Consequently,
implanted children may be able to use acoustic cues
that are not perceptible or available to implanted
adults to hear and appreciate music.
This study examined the development of musical
behaviour in children with cochlear implants and
compared the results to normal-hearing counterparts. The aims were to:
• Observe musical development in children with
normal hearing;
• Develop a questionnaire based on behavioural
observations;
• Evaluate musical development in children with
cochlear implants;
• Measure normative values and identify
developmental milestones;
• Identify differences in music development
between children with normal hearing and
children with cochlear implants.
A Musical Stages Profile (MSP) questionnaire was
developed using a parental interview format similar
to the format of the Meaningful Auditory Integration Scale (MAIS) questionnaire (Robbins et al,
1991). Questions were categorized into key areas in
146 Advanced Bionics® Auditory Research Bulletin 2007
“It is particularly encouraging to observe
development in behaviours
relating to melody perception
since this is a skill not often seen
in implanted adults.”
The MSP was evaluated with 15 normal-hearing
children and 25 children implanted with either a
CII Bionic Ear or HiRes 90K device. For analyses,
the normal-hearing children were divided into three
age ranges: 25-36 months (n = 6), 37-48 months
(n = 8), and older than 49 months (n = 1). The
implanted children were divided into four age ranges
according to their age at implantation: younger
than 24 months (n = 3), 25-36 months (n = 8),
37-48 months (n = 6), and older than 49 months
(n = 8). For the implanted children, MSP scores
were collected before implantation and at one, three,
and six months after implantation.
Preliminary results for four of the key areas of musical
skills development are presented as mean scores for
each age group. Figure 1 shows that all children with
cochlear implants were developing good awareness of sounds in their environment. This skill was
particularly apparent for children implanted at the
youngest ages, whereas the children implanted at the
oldest age (older than 49 months) were developing
sound awareness more slowly. It is encouraging to
see that melody perception was developing in the
implanted children (Figure 2). This skill was particularly noticeable for the two middle age ranges. For
the children implanted older than 49 months and
younger than 24 months, melody awareness was
slower to develop.
preimplant
1 month
3 months
6 months
Normal Hearing
5
Mean Rating
4
3
2
1
0
<24 months
25–36 months
37–48 months
>49 months
Age at Implantation
Figure 1. Mean ratings for the development of Sound Awareness for four groups of children implanted at different ages and
compared to ratings for normal-hearing children within the two
age groups shown.
preimplant
1 month
3 months
6 months
Normal Hearing
5
4
Mean Rating
which hierarchical development of listening skills
was assessed. The areas included General Information, Sound Awareness and General Reaction to
Sound, Exposure to Music, Melody and Dynamic
Changes, Rhythmical Changes, and Emotional
Aspects. Parental observations of behaviour were
categorized on a five-point scale where 1 = never,
2 = rarely, 3 = occasionally, 4 = frequently, and 5 =
always.
3
2
1
0
<24 months
25–36 months 37–48 months
>49 months
Age at Implantation
Figure 2. Mean ratings for the development of Melody and
Dynamic Changes for four groups of children implanted at
different ages and compared to ratings for normal-hearing children within the two age groups shown.
—continued on next page—
Music Studies 147
—continued from previous page—
Rhythm development was fairly slow for all
implanted children and had not reached the level
of the age-matched normal-hearing listeners by six
months after implantation (Figure 3). It may be
that this skill is more difficult to achieve. Moreover,
although the youngest group did not show much
progression, their preimplant scores generally were
higher than the other subjects. As to development
of the emotional aspects of music, there was some
preimplant
1 month
3 months
6 months
Normal Hearing
5
Mean Rating
4
3
2
1
0
<24 months
25–36 months 37–48 months
>49 months
Age at Implantation
Figure 3. Mean ratings for Rhythm Development for four groups
of children implanted at different ages and compared to ratings
for normal-hearing children within the two age groups shown.
preimplant
1 month
3 months
6 months
In summary, the MSP may become a useful and efficient tool for assessing the development of musical
behaviours. It is sensitive enough to highlight small
changes in music-related behaviour. Moreover, the
MSP is valuable for making parents aware of the
importance of listening and encouraging music
appreciation in their implanted children. The small
sample sizes in this first study make it difficult to
draw strong conclusions at this point although
general trends are becoming apparent. Preliminary
results show that implanted children can develop
music listening skills although the rate of development may be dependent upon the age at implantation. It is particularly encouraging to observe
development in behaviours relating to melody
perception since this is a skill not often seen in
implanted adults. Continuing long-term evaluation
with the MSP is required to assess the progress in
this area. Further work will continue to establish
normative values, particularly for younger and older
normal-hearing children.
Normal Hearing
5
Acknowledgement
Professor Graham Welch (University College London) and Christine Rocca (Mary
Hare School for the Deaf) contributed significantly to the development of the
Musical Stages Profile.
4
Mean Rating
improvement in the three youngest groups, but
only a small trend for improvement for the oldest
group (Figure 4). This pattern suggests that children
implanted later in childhood take longer to tune into
the emotional nuances of music.
3
2
References
1
Fujita S, Ito,J. (1999) Ability of Nucleus cochlear implantees to recognize music. Ann
Otol Rhinol Laryngol 108:634-640.
0
Kong Y-Y, Cruz R, Jones A, Zeng F-G. (2004) Music perception with temporal cues in
acoustic and electric hearing. Ear Hear 25:173-185.
<24 months
25–36 months 37–48 months
>49 months
Age at Implantation
Figure 4. Mean ratings for Emotional Development for four
groups of children implanted at different ages and compared to
ratings for normal-hearing children within the two age groups
shown.
Leal MC, Shin YJ, Laborde M-L, Lugardon S, Andrieu S, Deguine O, Fraysse B. (2003)
Music perception in adult cochlear implant recipients. Acta Otolaryngol 123(7):826835.
McDermott HJ. (2004) Music perception with cochlear implants: a review. Trends
Amplif 8(2):49-82.
Mirza S, Douglas S, Lindsey P, Hildreth T, Hawthorne M. (2003) Appreciation of
music in adult patients with cochlear implants: a patient questionnaire. Coch Impl
Intl 4(2):85-95.
Robbins AM, Renshaw JJ, Berry SW. (1991) Evaluating meaningful auditory
integration in profoundly hearing impaired children. Am J Otol 12(Suppl):144-150.
148 Advanced Bionics® Auditory Research Bulletin 2007
Video Assessment of the Development of Musical Behavior
in Children with Cochlear Implants
Armagan Incesulu, M.D.1
Cem Kecik, M.D.1
Hilal Dincer, M.Sc.2
1 Osman Gazi University, Eskisehir, Turkey
2 Erisci Elektroniks, Istanbul, Turkey
1 month
6 months
5
4
3
2
1
Ge
Be M ner
ha us al
vi ic
or
Em
Be oti
ha on
vi al
or
Pe Ry
rc th
ep m
tio n
n
& M
e
Ch Dyn lod
an am y
ge ic
s
po Mu
su sic
re
0
Ex
The development of musical behavior is assessed
using video analysis together with a questionnaire
to map developmental stages. The key areas of
development are categorized into sound awareness,
exposure to music, melody and dynamic changes,
rhythm perception, emotional behavior, and general
musical behavior. For the video analysis, the clinician rates the child’s development in each category
of music behavior on a five-point scale ranging from
1 (never) to 5 (always). Results from the video analysis are compared to clinical outcomes such as the
IT-MAIS, CAP (Categories of Auditory Performance), SIR (Speech Intelligibility Rating) profile,
and standard speech perception tests, whenever
possible. The study will determine if these behaviors and skills develop at different rates compared to
normal-hearing children.
preimplant
Aw S
ar oun
en d
es
s
Thus far, five children with prelinguistic, bilateral,
severe-to-profound sensorineural hearing loss are
participating in the study. Clinical sessions using
music activities have been designed in a hierarchy of
tasks to explore detection, discrimination of instruments and melodies, and music enjoyment. Each
assessment includes observation of social skills such
as spontaneous vocal behavior, turn-taking, and
imitation. The clinic sessions are to be conducted
over a one-year period to track changes in musical
behavior. All sessions are recorded on videotape.
When the rating scale was converted to percentage
values (1 = 0%, 2 = 20%, 3 = 40%, 4 = 80%, 5 =
100%), preliminary results showed that musical
behavior in general improved 8% on average in the
first month after implantation and 36% on average
at six months after implantation when compared
to preimplant ratings (Figure 1). These results were
consistent with improvements in speech perception
and production skills. The video analyses and development of a transcription method is ongoing. It is
anticipated that the video recordings will provide
objective, detailed, and permanent records of the
development of musical behaviors in the implanted
children. The data then can be analyzed by a variety
of professionals so that transcription and scoring can
be compared to determine test reliability.
Mean Rating
The aim of this ongoing study is to use video analysis to observe the stages of musical behavior development in children using the Advanced Bionics
HiRes 90K implant. An additional aim is to educate
families on the benefits of using music to encourage
children’s listening and communication skills and to
promote their speech and language development.
Figure 1. Clinicians’ ratings of music behavior development in
five children before cochlear implantation, after one month of
implant use, and after six months of implant use on a five-point
scale (1 = never; 2 = rarely; 3 = occasionally; 4 = frequently;
5 = always).
Music Studies 149
Effects of Pitch, Rhythm, and Melody Training
on Speech and Music Development in Children
with Cochlear Implants
Esra Yucel, Ph.D.1
Deborah Vickers, Ph.D.2
Gonca Sennaroglu, Ph.D.1
Hilal Dincer, M.Sc.3
Bilgehan Budak, Ph.D.1
Erol Belgin, Prof., Ph.D.1
1 Hacettepe University Hospital, Ankara, Turkey
2 University College London, London, United Kingdom
3 Erisci Elektroniks, Ankara, Turkey
Research shows that listeners with cochlear implants
do not hear music well (Leal et al, 2003; McDermott, 2004; Mirza et al, 2003). It is believed that the
main reason for this poor perception is that cochlear
implant users have difficulty perceiving pitch. Poor
pitch perception may occur because (1) implants
do not deliver stimulation associated with the
fundamental frequency (50-300 Hz), (2) implant
sound processing preserves envelope cues but does
not retain fine structure cues associated with good
pitch perception, and (3) neural survival limits rate
discrimination in some implant users.
However, it may be that young children who learn
to perceive sound using an implant can learn to
perceive pitch more effectively than postlinguistically
deafened adults, especially when using strategies
such as HiResolution Sound (HiRes) that aim to
deliver some fine structure information to the low-
frequency channels (apical electrodes). Pantev and
colleagues (1998) have shown that if children with
normal hearing learn to play a musical instrument
below the age of nine years, they develop a different
cortical representation for music and pitch than nonmusicians. This observation—taken together with
the fact that Anvari and co-workers (2002) showed
a relationship between musical skills, phonological
processing, and reading ability—provides a strong
case for exposing young children with cochlear
implants to a musical environment at an early age to
allow for optimal processing of pitch.
This study aimed at training young implanted children in pitch, rhythm, and melody perception with
the goal of improving music discrimination as well
as speech perception. Two groups of nine subjects
each participated in the study. The first group were
children who used HiRes but did not receive musical
training. The second group were children who used
HiRes and received musical training. Demographic
data show that the mean audiometric threshold
averages were similar for the two groups (Table 1).
The ages for hearing aid fitting and cochlear implantation were slightly lower for the control group than
for the trained group, potentially giving an early
learning advantage to the control group.
At the initial stimulation, families with children in
the trained group were provided with a curriculum
consisting of note discrimination, rhythm discrimi-
Table 1. Average pure tone thresholds and demographics for two study groups—
one receiving music training and a control group that did not receive music training.
Age at Hearing Aid Fitting
(months)
Age at Implantation
(months)
Group
PTA*
Min
Max
Mean
Min
Max
Mean
Trained
104.2 dB
8
49
28.67
± 13.29
39
96
55.22
± 17.57
Control
102.4 dB
12
36
23.11
± 8.89
22
96
49.33
± 20.83
* PTA for .5, 1, 2, and 4 kHz.
150 Advanced Bionics® Auditory Research Bulletin 2007
“. . .speech perception development
was more rapid in the early months for
children who received music training. . .”
Results showed that there were no significant differences between the two groups on the MAIS, ITMAIS, and MUSS questionnaires. The ESP and
Mr. Potato Head test results showed that by 12
months, the mean results for both groups were nearly
at ceiling levels. However, children receiving music
training typically improved their speech perception scores more rapidly than those in the control
group (Figures 1 and 2). The most noticeable difference was observed at the three-month test interval.
The Musical Stages Profile results were the same for
both groups except that the musically trained children were exposed to music more than the children
who did not receive music training.
Although the Musical Stages Profiles were similar
between the two groups, the study had a positive
effect by making families aware of the importance of
music in everyday life. After 12 months experience
with the music training program, families have found
the experience rewarding and enjoyable. Although
these results are preliminary, it is encouraging to see
that speech perception development was more rapid
in the early months for children who received music
training, thereby suggesting that the discrimination
training program did help to improve speech perception abilities even though it did not have an effect on
the development of musical behavior.
Trained Group
Control Group
Mean Percent Correct
100
80
60
40
20
0
baseline
1 month
3 months
6 months
12 months
Test Interval
Figure 1. Mean Early Speech Perception (ESP) test results over
time for the children receiving musical training and the control
group.
Trained Group
Control Group
100
Mean Percent Correct
nation, and melody identification tasks. Each family
was given an electronic keyboard with color-coded
notes and a hierarchical set of tasks to perform. The
children returned to the clinic for assessment after
1, 3, 6, 12, and 18 months of implant use. Children
in the control group were assessed at the same postimplant intervals. Assessment consisted of parental
questionnaires (MAIS, IT-MAIS, MUSS) and
standard speech tests for young children (ESP, Mr.
Potato Head). At the 12-month test interval, each
child was evaluated with the Musical Stages Profile.
(See also Vickers et al, page 146 of this bulletin.)
80
60
40
20
0
baseline
1 month
3 months
6 months
12 months
Test Interval
Figure 2. Mr. Potato Head Test results over time for the children
receiving musical training and the control group.
References
Anvari SH, Trainor LJ, Woodside J, Levy BA. (2002) Relations among musical skills,
phonological processing, and early reading ability in preschool children. J Exp Child
Psych 83:111-130.
Leal MC, Shin YJ, Laborde M-L, Lugardon S, Andrieu S, Deguine O, Fraysse B. (2003)
Music perception in adult cochlear implant recipients. Acta Otolaryngol 123(7):826835.
McDermott HJ. (2004) Music perception with cochlear implants: a review. Trends
Amplif 8(2):49-82.
Mirza S, Douglas S, Lindsey P, Hildreth T, Hawthorne M. (2003) Appreciation of
music in adult patients with cochlear implants: a patient questionnaire. Coch Impl
Int 4(2):85-95.
Pantev C, Oostenveld R, Engellen A, Ross B, Roberts LE, Hoke M. (1998) Increased
cortical representation in musicians. Nature 392(23):811-813.
Music Studies 151
The CIRCLE Game: A Music Education Program for
Pediatric Cochlear Implant Users
Patricia M.H. Gulpen, M. A.1
Jeroen J. Briaire, M.Sc.1
Ruud Mourik2
Johan H. M. Frijns, M.D., Ph.D.1
1 Leiden University Medical Centre, Leiden, The Netherlands
2 Musician and Teacher, Nieuwveen, The Netherlands
“. . .67% of the parents noted benefits
from The CIRCLE Game
on one or more aspects
of their children’s development.”
Children love to make music, thus making music
a powerful educational tool. Studies have shown
that good musical abilities facilitate phonological
and syntax processing as well as reading abilities
in normal-hearing children (Anvari et al, 2002;
Bastian, 2000; Jentschke et al, 2005). However,
most studies of cochlear implant users indicate that
music perception and appreciation is poor (e.g.,
Gfeller et al, 2000; Vongpaisal et al, 2006). Nonetheless, training in specific musical aspects is known
to improve music perception and enjoyment in
adult implant recipients (Gfeller et al, 2002). Thus,
there may be opportunities for enhancing the rehabilitation of pediatric implant recipients through
music training, especially when taking into account
the greater plasticity of their developing brains.
Moreover, newer strategies with higher stimulation rates might facilitate improved representation
of music information for current pediatric cochlear
implant recipients.
152 For this reason, the Leiden University group has
developed and piloted The CIRCLE Game, a music
training program for Dutch pediatric cochlear
implant users. The main purposes of the program are
to introduce children to music, to engage children
in musical activities, and to support overall auditory
development. The program uses a range of musical
instruments, activities, and newly composed songs
to teach children about rhythm, melody, pitch, and
singing. There are activities for different age groups
and levels of auditory development, making The
CIRCLE Game suitable for a wide range of children, including children with additional disabilities.
The first edition of The CIRCLE Game was
comprised of six 45-minute music lessons with one
lesson given every three weeks. Eighteen children
participated in the evaluation of the first edition. At
the start of the program, the mean age of the children was 43 months (range: 14–71 months) and
the mean duration of cochlear implant use was 23
months (range: 2–52 months). All children were
using a CII or HiRes 90K device and five of them
were implanted bilaterally. They were divided into
four groups based on development and hearing age.
The effects of The CIRCLE Game on musical skills
were evaluated by using the Musical Stages Questionnaire (MSQ). This questionnaire, developed
by Advanced Bionics Europe, evaluates a child’s
exposure to music and the awareness of sounds in
general, awareness of changes in melody, dynamics
and rhythm, and awareness of the emotional aspects
of music. The Dutch versions of the Categories of
Auditory Performance (CAP-NL) and Speech Intelligibility Rating (SIR-NL) were used to assess how
the program affected auditory and speech production skills. Parents completed all questionnaires
before and after the series of lessons. The parents
of 15 children returned the questionnaires at both
time points. In addition, parents completed an extra
Advanced Bionics® Auditory Research Bulletin 2007
The MSQ, CAP-NL, and SIR-NL results before
and after the program were compared using the
Wilcoxon Signed Ranks Test. No significant differences were found. Detailed analyses of the data also
revealed no clear trends for improvement of musical,
auditory, or speech production skills. In contrast, the
results of the parents’ appraisal questionnaire were
more positive. First, 93% of the parents reported
that their children enjoyed the music lessons. All
of the parents, with the exception of one, declared
that they would enroll their child for a new edition
of The CIRCLE Game. Second, 67% of the parents
noted benefits from The CIRCLE Game on one or
more aspects of their children’s development.
As shown in Figure 1, the greatest benefits were
observed in the area of musical development (with
67% of the parents reporting “moderate” to “very
much” improvement), whereas the smallest effects
were seen in language development (with 40% of
the parents reporting “moderate” to “very much”
improvement). Some parents also noted that they
saw their children develop more interest in music
and start to recognize songs, but that the number of
lessons and frequency should be increased for those
gains to be realized more fully.
In summary, The CIRCLE Game resulted in little
improvement on the musical, auditory, and speech
perception skills evaluated by the MSQ, CAP-NL
and SIR-NL. Nonetheless, a majority of the parents
reported added benefits from the program on one
or more aspects of their child’s development. If the
parents’ observations are correct, it may be that the
improvements after this small number of music
lessons is yet too small to be detected by the MSQ,
CAP-NL, and SIR-NL, or that the benefits of The
CIRCLE Game are not revealed by these measures.
In either case, it is evident that the children enjoyed
and benefitted from the music lessons, which in itself
underscores the value of using music as a rehabilitation tool for pediatric cochlear implant recipients.
50
None
Little
Moderate
Much
Very Much
40
Percent of Parents
appraisal questionnaire at the end of the program
that consisted of 18 questions regarding parents’
opinions and experiences concerning the benefit of
the lessons, the child’s enjoyment of the experience,
and the duration of the program.
30
20
10
0
Enjoyment
of Music
Music
Development
Auditory
Development
Language
Development
Figure 1. Distribution of parent ratings of benefit from The
CIRCLE Game on various aspects of their child’s behavior and
development.
References
Anvari S, Trainor L, Woodside J, Levy,B. (2002) Relations among musical skills,
phonological processing, and early reading ability in preschool children. J Exp Child
Psychol 83(2):111-130.
Bastian HG. (2000). Musik(erziehung) und ihre Wirkung. Mainz: Schott.
Gfeller K, Christ A, Knutson J, Witt S, Murray KT, Tyler RS. (2000) Musical
backgrounds, listening habits, and aesthetic enjoyment of adult cochlear implant
recipients. J Am Acad Audiol 11:390-406.
Gfeller K, Witt S, Adamek M, Mehr M, Rogers J, Stordahl J, Ringgenberg S. (2002)
Effects of training on timbre recognition and appraisal by postlingually deafened
cochlear implant recipients. J Am Acad Audiol 13:132-145.
Jentschke S, Koelsch S, Friederici AD. (2005). Investigating the relationship of music
and language in children: influences of musical training and language impairment.
Ann New York Acad Sci 1060:231-242.
Vongpaisal T, Trehub SE, Schellenberg EG. (2006). Song recognition by children and
adolescents with cochlear implants. J Speech Lang Hear Res 49:1091-1103.
Music Studies 153
Objective Measures
In addition to programming the HiResolution® Bionic Ear System, the
SoundWave™ Professional Suite software offers tools for testing the
implanted electrode array and for objectively assessing hearing nerve function.
For example, Neural Response Imaging (NRI) measures the response of the
hearing nerve to electrical stimulation, yielding single-channel NRI recordings. Also, unique to SoundWave are Speech Bursts—stimuli that can be used
to measure electrical stapedius reflexes thresholds, which relate closely to
everyday most comfortable listening levels.
In addition to the clinical SoundWave software, other research software
applications can measure banded-NRI responses whereby multiple electrodes
are stimulated simultaneously. The banded-NRI technique allows rapid
assessment of the entire electrode array and yields input-output functions
closely parallel to the psychophysical loudness growth functions obtained with
Speech Bursts.
And, the new Smart NRI™ software implements an autodetection algorithm
to identify neural responses. Clinical studies indicate that Smart NRI has
an error rate of less than 3% when compared to subjective clinician response
classifications. Smart NRI is the first truly objective method for assessing
auditory nerve function.
The studies in this section summarize various investigations into the clinical
efficacy of these objective measures.
155
European Multicenter Objective Measures Study
Laure Arnold, M.Sc.
Patrick Boyle, M.Sc.
Advanced Bionics, Europe
“. . .when no intraoperative
objective measures are available,
measuring NRI responses on
only one electrode during the first fitting
may be sufficient to create
an initial usable program.”
Single Electrode
100
n=37
n=37
E 3/1–4
E 7/5–8
Speech Burst
n=37
n=35
80
60
Percent Response Rate
40
20
0
E 11/9–12
Intraoperative
100
n=70
n=73
E 15/12–16
Mean
6 months
n=69
n=67
80
60
40
20
0
E3
E7
E 11
E 15
Mean
Location on Electrode Array
Figure 1. Measurement response rate for (A) ESRT and (B) NRI
recordings, depending on location along the electrode array.
For both ESRT and NRI measurement, the response rate was
slightly lower at the most basal electrode tested compared to
the rest of the array.
156 The objective of this multicenter study was to
develop guidelines for the use of electrically evoked
compound action potentials (ECAPs) and electrical
stapedius reflex thresholds (ESRTs) to optimize
HiRes fittings by investigating the relationship
between those objective measures and psychophysical data. The principle investigator was Thierry Van
Den Abbeele from Hôpital Robert Debré in Paris,
France.
All subjects used a CII Bionic Ear or a HiRes 90K
implant unilaterally. ECAPs were measured using
the Neural Response Imaging (NRI) module within
the SoundWave fitting software. The stimulating/
recording electrode pairs typically were 3/1, 7/5,
11/9, 15/13. The measurements were made intraoperatively, at first fitting, and after 3, 6, and 12 months
of implant use. Two measures are defined within
SoundWave, (1) the threshold of NRI (tNRI), the
stimulation level at which the NRI response should
have zero amplitude and (2) the “1st NRI,” the
lowest stimulation level at which the NRI waveform can be identified visually. Both measures were
collected. ESRTs were determined intraoperatively
by visual detection of the stapes movement. Both
single-electrode stimulation and Speech Bursts, where
four electrodes are stimulated as a sequential group,
were used to elicit ESRTs. Subjects were fitted using
the SoundWave default parameters—that is using
Speech Bursts and automatic calculation of thresholds
(Ts) (10% of M levels).
Data from 118 subjects across 14 clinics were evaluated. As seen in Figure 1, for ESRTs, speech burst
stimulation elicited a significantly higher response
rate (84%) than single-channel stimulation (64%)
(sign test: p < 0.05 for electrodes 3, 7, 11, p < 0.01 for
electrode 15). The NRI response rate was 81% intraoperatively and increased to 96% after six months
of implant use. This difference was significant for
all locations tested (sign test: p < 0.05 for electrodes
3 and 7, p < 0.005 for electrode 11, and p < 0.01
for electrode 15). All subjects showed at least one
NRI response.
Advanced Bionics® Auditory Research Bulletin 2007
Principal Study Site
Hôpital Robert Debré, Paris, France
Participating Study Sites
Belgium: University of Ghent, Ghent
France: Hôpital Edouard Herriot, Lyon; Hôpital Charles Nicolle, Rouen
Germany: Unfallkrankenhaus Charité Medical School, Berlin
India: DeSa’s Hospital, Mumbai
Israel: Hadassah Medical Organisation, Jerusalem; Chaim Sheba Medical Center, Tel Hashomer;
Schneider Children’s Medical Centre, Petah Tikva
Italy: Cattedra di Audiologia, Ferrara
Morocco: Clinique Rachidi, Casablanca
Turkey: Hacettepe University Hospital, Ankara; SB Ankara Diskapi Hospital, Ankara ; SB Izmir Tepecik Hospital, Izmir
A significant correlation was found between the
intraoperative 1st NRI on electrode 3 and the sixmonth M levels for all four measured locations
(Electrode 3: r = 0.37, p = 0.002, n = 70; Electrode 7:
r = 0.345, p = 0.003, n = 70; Electrode 11: r = 0.3,
p = 0.012, n = 70; Electrode 15: r = 0.262, p = 0.03,
n = 69). Because there were no significant M level
differences across electrodes, these data suggested
that usable “flat” M level programs might be predicted
by the intraoperative 1st NRI levels. An average M
level was calculated on the basis of the correlation
between intraoperative 1st NRI (IO1stNRI_3) and
six-month M level on Electrode 3 (M6) (calculation:
M6 = 98.91 +0.44 * IO1stNRI_3) and was compared
to the average behavioural M level at six months. A
highly significant correlation was found (r = 0.338,
p = 0.005, n = 69). A paired sample t-test showed
that there was no significant difference between the
calculated M and the average M (t = -0.745, df = 68,
p = 0.459), hence showing that the use of a flat
program based on an apical intraoperative 1st NRI
value may be a good objective indicator of where
program levels might be set. A similar result was
obtained using the first fitting 1st NRI (FF1stNRI_
3), which was significantly (p <0.01) correlated to the
300
Clinical Units (Charge)
M levels were highly correlated across electrodes
(p < 0.01) and there were no significant differences
between the M levels across electrodes. As indicated
in Figure 2, M levels were significantly different
between the first fitting and all subsequent sessions
(3, 6, 12 months) but not between the latter sessions
themselves, suggesting that M levels stabilized after
six and perhaps even three months.
250
200
150
100
50
0
First fitting
3 months
6 months
12 months
Figure 2. Mean M levels over time, averaged across four electrodes. The error bars show the standard deviation. The variability across subjects was relatively large. After three months, the
changes in M levels were no longer significantly different from
each other.
—continued on next page—
Objective Measures 157
—continued from previous page—
“The data suggest that a usable
“flat” program could be fit initially
based on only one intraoperative
NRI measurement on an apical electrode.
When ESRT recordings were also
available, an even better model could
be defined . . . based on intraoperative
ESRTs and first fitting NRI responses.”
six-month M level on Electrode 3 (calculation: M6
= 42.6 + 0.94 * FF1stNRI_3) and also to the other
three locations. Thus, when no intraoperative objective measures are available, measuring NRI responses
on only one electrode during the first fitting may be
sufficient to create an initial usable program.
Analyses of ESRTs showed that the electrodes could
be split between apical (Electrodes 3, 7) and basal
(Electrodes 11, 15). A multiple regression analysis
generated a predictive model for the apical M levels
(Map) in a congenitally deaf subgroup of subjects,
according to the following equation: Map = -43.37 + 0.14 * IOESRT + 1.39 * FF1stNRI, where
IOESRT is the intraoperative ESRT obtained with
speech burst stimulation and FF1stNRI is the first
fitting 1st NRI. The model was tested on the basal
electrodes by determining the correlation between
the calculated basal Ms and the psychophysical
basal Ms (R = 0.638, p = 0.014). A paired-sample ttest showed that there was no significant difference
between calculated and psychophysical basal Ms,
suggesting that the model may be applied across the
entire electrode array. The model also was tested on
the entire group of subjects and proved appropriate
for predicting M levels at the apical end of the electrode array, but not for the basal part of the array.
These results provide useful insights into the behaviour of objective measures with respect to time, electrode, and programming levels. The data suggest that
a usable “flat” program could be fit initially based
on only one intraoperative NRI measurement on an
apical electrode. When ESRT recordings were also
available, an even better model could be defined for
the congenitally deaf subgroup, based on intraoperative ESRTs and first fitting NRI responses. The next
steps in this study will evaluate SmartNRI™, a new
NRI measurement method likely to reduce variability and to improve objectivity and quality.
158 Advanced Bionics® Auditory Research Bulletin 2007
New Study Init iat i ve
Comparison of Two Artefact Suppression Methods for Neural Response Imaging (NRI)
Martin Westhofen, Prof, M.D.1
Wolfgang H. Döring, Dr. Ing.1
Herbert Dujardin, Dipl. Ing.1
Alexander Elsholz, Dipl.Ing.1
Petra Ambrosch, Prof., M.D.2
Joachim Müller-Deile, Dipl. Phys.2
Andre Morsnowski, Dipl. Phys.2
1Universitätsklinikum Aachen, Aachen, Germany
2Universitätsklinikum Schleswig Holstein, Kiel, Germany
The
Advanced
Bionics
software,
key question is how well TECAPs obtained during surgery are
introduced in 2002, incorporates Neural Response Imaging
able to predict the subjective psychophysical loudness levels
(NRI), a noninvasive diagnostic tool to evaluate electrically
measured at the initial stimulation. Another key question is
evoked
telemetric
whether TECAP-M level relationships in children differ from
measurement (TECAPs). With NRI, the stimulus is delivered
those in adults. It is reasonable to predict that changes would
by the implant and the auditory nerve’s response is registered
be more noticeable in children because of the maturation
via a measuring amplifier within the implant. The data are
process.
compound
SoundWave
action
potentials
fitting
by
sent through the skin by telemetry to the fitting system for
further processing.
Two groups of both children and adults will participate in
the study. One group will consist of experienced implant
It has been shown that the threshold values for TECAPs
users with at least six months of device use. The other group
typically fall within the electrical dynamic range of the
will consist of newer users with less experience to track
cochlear implant user. Thus, NRI has the potential to serve
developments during the early phase of device use.
as an objective tool to help audiologists in programming
implants in people unable to provide reliable behavioural
The results of this study should (1) establish the advantages
feedback. The challenge of measuring TECAPs with the
and disadvantages of both artefact reduction methods with
implant system is distinguishing the neural responses from
respect to measurement time, reproducibility, latency, and
the stimulus artefact. The amplitude of electrical stimulus
steepness of amplitude growth, threshold, and success rate,
amounts to several volts, while the amplitude of the neural
(2) compare various elements of both NRI methods with
response is only a few microvolts. Therefore, it is essential
clinical program parameters (e.g., differences between clinical
that the stimulus artefact be suppressed so that the true neural
M levels and NRI thresholds with respect to the steepness
response can be measured.
of amplitude growth functions, differences between the M
levels and NRI thresholds with respect to the clinical success
In the proposed study, two different suppression procedures
of implantation, measured by age-appropriate tests of speech
will be investigated in detail using the RSPOM software
understanding or questionnaires), and (3) describe changes
with the HiRes 90K implant: (1) alternating stimulus polarity
in TECAP responses and M levels over time (intraoperative,
(Zimmerling & Hochmair, 2002) and (2) forward masking
first fitting, one month, and three months after first fitting).
using a masker-probe-stimulus technique (Brown & Abbas,
1990). The study will ascertain to what extent each artefactsuppression method affects the results of TECAP measurements
and their correlations with the implant’s program parameters.
In addition, an attempt will be made to establish whether
and to what extent the relationship between M levels and
TECAPs changes over time with implant experience. One
References
Brown CJ, Abbas PJ. (1990) Electrically evoked whole-nerve action potentials:
parametric data from the cat. J Acoust Soc Am 88(5):2205-2210.
Zimmerling MJ, Hochmair ES (2002). EAP recordings in Ineraid patientscorrelations with psychophysical measures and possible implications for patient
fitting. Ear Hear 23:81-91.
Objective Measures 159
Smart NRI Multicenter Study in Europe:
Preliminary Results
Bruno Frachet, Prof., M.D.1
Yves Ormezzano, M.D.1
Laure Arnold, M.Sc.2
Patrick Boyle, M.Sc.2
1 Hôpital Avicenne, Bobigny, France
2 Advanced Bionics, Europe
“It is anticipated that SmartNRI
will become a quick and straightforward
clinical method for identifying
valid NRI responses. . .”
Neural Response Imaging (NRI) is a software tool
for measuring auditory neural responses in the
HiResolution Bionic Ear System. If NRI is to be
useful clinically in setting program levels in individuals who cannot provide behavioral feedback, then
the neural responses must be easily identifiable by all
clinicians, experienced and inexperienced. However,
the inherent overlap of a large stimulus artifact with
the much smaller neural response makes the accurate determination of neural response thresholds a
task requiring considerable skill and experience.
A rigorous and automatic, statistically based method
(Smart NRI™) was developed for determining
the stimulation level at which a neural response
is present in a set of NRI responses (Litvak et al,
2005). The statistical method assumed that all
measurements consisted of neural responses, noise,
and residual artifact. First, noise was reduced using a
principle-component analysis. Then the NRI traces
were reconstructed using a linear sum of seven basic
functions. A best-fit artifact model was constructed.
An analysis of over 1,000 NRI traces verified a
model that accurately represented residual artifact.
Any trace that differed significantly from the artifact
model was considered to be a real neural response. A
strength-of-response (SOR) metric was computed
to quantify how far a response deviated from the
artifact.
The aims of the in-progress European multicenter
study are (1) to validate the automatic detection
algorithm and (2) to compare behavioral fittings to
SmartNRI-based programs. In the pilot phase of the
study, NRI measurements made with the SoundWave fitting software are classified as “responses”
or “no responses” by experienced clinicians. The
160 Advanced Bionics® Auditory Research Bulletin 2007
Principal Study Site
Hôpital Avicenne, Bobigny, France
Participating Study Sites
Belgium: Cliniques Universitaires Saint-Luc, Brussels; University Hospital, Ghent
France: Hôpital Edouard Herriot, Lyon; Hôpital Gui de Chauliac, Montpellier
Germany: Unfallkrankenhaus Charité Medical School, Berlin; Universitätsklinikum, Kiel
India: Jaslok Hospital and DeSa’s Hospital Cochlear Implant Centre, Mumbai
Israel: Schneider Children’s Medical Center Hospital, Petah Tikva
Turkey: Hacettepe University Hospital, Ankara; Osman Gazi University, Eskisehir; SB Ankara
Diskapi Hospital, Ankara; SB Ankara Numune Hospital, Ankara; SB Izmir Tepecik Hospital, Izmir
classified responses then are processed through the
auto-detection algorithm. Outcomes are compared
to validate the system. Preliminary results have
been collected at Hôpital Avicenne in France. Four
experienced clinicians examined 452 NRI traces
using a specific classification interface. The subset
of measurements that were agreed by all reviewers
to be “responses” or “no responses” (268/452: 59%)
was assessed against the performance of the automatic classification system. Ninety-nine percent of
the “responses” were correctly classified as responses
by the algorithm and 96% of the “no responses”
were identified. Thus, there were seven false positives
and one false negative, yielding an error rate of less
than 3%.
The pilot phase now will continue with a multicenter
validation of the algorithm on a larger number of
NRI traces. The algorithm has been incorporated into
the Advanced Bionics Research Platform for Objective Measures (RSPOM, Van Immerseel et al, 2007),
which has reduced the NRI acquisition time to six
minutes from the twenty minutes typically required
in current clinical applications. The second phase of
the study will compare behaviorally based programs
and programs created using the SmartNRI principle
within RSPOM. It is anticipated that SmartNRI
will become a quick and straightforward clinical
method for identifying valid NRI responses, thereby
potentially providing a faster and more reliable tool
for setting program levels in implant users who are
unable to make loudness judgments, particularly the
pediatric population.
References
Frachet B, Ormezzano Y, Arnold L, Boyle P. Results of the SmartNRI study. Presented
at the Fifth International Symposium on Objective Measures in Cochlear and
Brainstem Implants, Varese, Italy. 9-12 May, 2007.
Litvak L, Emadi G. Automatic estimate of threshold from Neural Response Imaging
(NRI). Poster presented at the Conference on Implantable Auditory Prostheses, Pacific
Grove, CA, USA. 30 July-4 August, 2005.
Van Immerseel L, Vanpoucke F, Dykmans P, Boyle P. Flexible NRI measurements
with the HiRes 90K implant. Poster presented at the Fifth International Symposium
on Objective Measures in Cochlear and Brainstem Implants, Varese, Italy 9-12
May, 2007.
Objective Measures 161
European Banded Neural Response Imaging (NRI) Study:
One-Year Results
Istemihan Akin, M.D.1
Murad Mutlu, M.D.1
Gokhan Kuran, M.D.1
Hilal Dincer, M.Sc.2
Laure Arnold, M.Sc.3
1 SB Ankara Diskapi Training and Research Hospital,
Ankara, Turkey
2 Erisci Elektroniks, Istanbul, Turkey
3 Advanced Bionics, Europe
The SoundWave programming software uses
Speech Bursts to set hearing thresholds (T levels) and
most comfortable levels (M levels) for the HiResolution Bionic Ear System. Speech Bursts consist of
white noise that is passed through the same FFT
algorithm applied in HiRes processing of sound in
everyday use. Speech Bursts typically are delivered to
four groups of four adjacent electrodes with a stimulation rate of approximately 3,000 pulses per second
per contact. In contrast, the electrical compound
action potential (ECAP) is measured by stimulating
one electrode at a low stimulation rate and recording
from two more apical electrodes using the Neural
Response Imaging (NRI) software. Thus, the stimulation parameters used for behavioral programming
and NRI are quite different.
Banded NRI is a procedure in which neural
responses are elicited by stimulating simultaneously
three or four contiguous electrodes at an overall rate
that approximates speech burst stimulation. Banded
NRI can be implemented using research software
and may yield neural responses that more closely
approximate behavioral fitting parameters. Preliminary results have shown that banded NRI ECAP
growth functions are steeper than single-channel
NRI growth functions. Also, the thresholds are lower
for banded vs. single-channel stimulation (Akin et
al, 2006; Guiraud et al, 2005).
The objectives of this study were to investigate the
relationship between single-channel and banded
NRI responses and to evaluate whether banded
NRI measurements are better correlated to fitting
parameters than single-channel NRI measurements. Single-channel NRI responses were recorded
from four individual electrodes with the SoundWave fitting software. Banded NRI responses were
recorded from four bands of four electrodes each
using the research software. In most cases, data were
collected intraoperatively, at first fitting, and after 3,
6, and 12 months of implant use. The tNRIs were
extrapolated and the slopes of the growth functions
were calculated. Postoperative behavioral programs
were created using Speech Bursts with an automatic
calculation of T levels (10% of M levels).
162 Advanced Bionics® Auditory Research Bulletin 2007
Thus far, data have been collected for 19 subjects
(adults and children). Banded NRI responses could
be recorded reliably, and the procedure was not
more time consuming than the single-channel NRI
procedure. However, because of time limitations, it
was not always possible to complete all eight NRI
measurements at each session. Thus, only 15 (of
19) subjects have complete 12-month data sets.
The data revealed steeper growth functions and
lower thresholds for the banded NRI compared to
the single-channel NRI responses. The mean ratio
between banded and single-channel ECAP growth
function slopes was 4.3. The mean thresholds for
single-channel stimulation were 4.1 times higher
than the mean thresholds for banded NRI stimulation. As summarized in Table 1, correlations were
observed between the first fitting tNRIs (singlechannel and banded) and 12-month M levels on
electrodes 3, 7, 11 (bands 1-4, 5-8, 9-12), but not
for the basal electrode 15 (band 13-16). The correlations with behavioral M levels were more significant
and slightly higher for the banded NRI responses
than for the single-channel NRI responses on all
tested electrodes.
These study results indicate that banded NRI
responses can be recorded reliably and require
minimal test time. The correlations between NRI
and program parameters obtained in this sample
indicate that banded NRI may be a more appropriate fitting tool than single-channel NRI. More
data from a larger group of subjects are necessary to
confirm these results and to refine the correlations in
order to develop clinical guidelines for using banded
NRI to aid in programming the HiRes system.
“The correlations between
NRI and program parameters
obtained in this sample
indicate that banded NRI
may be a more appropriate fitting tool
than single-channel NRI.”
Table 1. Summary of correlations and
significance levels between
first fitting tNRI (single and banded)
and M levels at 12 months (n = 9).
Electrodes
Pearson R:
single-channel
tNRI
Pearson R:
banded
tNRI
3/1 (1-4)
7/5 (5-8)
0.81**
0.70*
p = 0.008 p = 0.038
0.85**
0.73*
p = 0.004 p = 0.025
11/9 (9-12) 15/13 (13-16)
0.83**
0.37
p = 0.006
p = 0.321
0.84**
0.62
p = 0.005
p = 0.075
Note: * = significant; ** = highly significant.
A significant correlation was observed for electrodes 3, 7, 11
(bands 1-4, 5-8, 9-12, respectively) but not for electrode 15 (band
13-16). Overall, the correlations were higher and more
significant with banded NRI compared with single-channel NRI.
References
Akin I, Mutlu M, Kuran G, Dincer H, Arnold L. (2007) One year results of the Banded
Neural Response Imaging (NRI) study. Paper presented at the Fifth International
Symposium on Objective Measures in Cochlear and Brainstem Implants, Varese,
Italy. 9-12 May, 2007.
Akin I, Yardimci S, Arnold L. (2006) Relationship between banded NRI and program
parameters. Paper presented at the 9th International Conference on Cochlear
Implants, Vienna, Austria. 14-17 June, 2006.
Guiraud J, Arnold L, Boyle P, Truy E, Collet L. (2005) Preliminary results with banded
neural response imaging (NRI). Paper presented at the 10th Symposium on Cochlear
Implants in Children, Dallas, Texas, USA. 15-19 March, 2005.
Objective Measures 163
Banded Neural Response Imaging (NRI) Measurements:
Comparison to Single-Channel NRI Responses and
Psychophysical Measurements Over Time
Gül Caner, M.D.1
Laure Arnold, M.Sc.2
Levent Olgun, M.D.1
Gürol Gültekin, M.D.1
1 Dokuz Eylul University, Izmir, Turkey
2 Advanced Bionics, Europe
“These data indicate that banded NRI
is a potentially useful tool
for clinical programming.”
Studies have shown that objective measures such
as single-channel Neural Response Imaging (NRI)
and electrically evoked stapedius reflex thresholds
(ESRTs) can provide initial fitting guidelines, especially for children. The electrically evoked compound
action potential (ECAP) also can be recorded using
banded-electrode stimulation implemented with
research NRI software. Banded NRI is a procedure
in which neural responses are elicited by stimulating
three to four contiguous electrodes at the same time
with an overall rate that approximates speech burst
stimulation. Banded NRI responses may be more
related to clinical programming levels because Speech
Bursts are used by the SoundWave fitting software
to program the implant.
The aims of this study were to investigate the relationship between single-channel and banded-NRI
responses over time and to determine the relationships among the two NRI measurements and behavioral measures of loudness. Nine subjects (eight
adults and one child) were included in the study. All
subjects were implanted with a HiRes 90K device.
164 Single-channel NRI responses were obtained by
stimulating four electrodes at four locations (typically electrodes 3, 7, 11, 15) and by recording from
the corresponding more apical electrodes (1, 5, 9,
13). For banded-NRI measurements, four electrode
bands (1-4, 5-8, 9-12, 13-16) were stimulated. NRI
measurements were made intraoperatively, at first
fitting, and after 3, 6, and 12 months of implant
use. NRI thresholds (tNRI) were extrapolated and
the growth function slopes were calculated for both
banded and single-channel NRI stimulation. Most
comfortable levels (M levels) were adjusted in live
speech mode at first fitting and at all postoperative
intervals. At the three-month test interval, loudness
perception was evaluated in the adults using a tenstep loudness scale.
Banded NRI responses could be recorded in all
subjects and were identified in 96.8% (93/96) of
the recordings. Loudness perception was evaluated in seven adult subjects. No significant correlations were found between intraoperative banded
tNRIs and first fitting M levels. On average, the
first fitting M levels were 3.24 times higher than the
intraoperative banded tNRIs. Banded tNRIs and
single-channel tNRIs were lower at the apical half
of the array compared to the base. Detection levels
(T levels) averaged 38 CU (clinical units) and
banded tNRI levels averaged 42 CU. Banded tNRI
values were within the behavioral dynamic range and
were closer to T levels than to M levels. Globally, the
T level/banded-tNRI ratio was 1.09 throughout the
array. M levels and M level/banded-tNRI ratio were
higher at the apical part of the array. Over time, both
T levels and banded-tNRI levels increased. At the
basal part of the array, both levels stabilized earlier.
Advanced Bionics® Auditory Research Bulletin 2007
Some subjects were asked to compare behavioral
programs to programs based on single-channel
tNRIs. Most subjects preferred the behavioral
programs and performed better with them, which
may be explained by the difference in profile illustrated in Figure 1. In other words, the subjects
preferred less high-frequency stimulation. As stated
above, the 12-month M-level profile was almost
parallel to the banded-tNRI profile. As a result, it
may be possible to use banded NRI as a fitting tool.
For example, at first fitting, M levels could be set by
doubling the banded-NRI thresholds and decreasing
the current at the basal electrodes. Overall M levels
then could be adjusted in live speech mode.
These data indicate that banded NRI is a potentially
useful tool for clinical programming. Banded-tNRI
levels may provide a guide for setting the profile of a
program, especially for pediatric first fittings. Nevertheless, to optimize fitting, all objective measures
should be combined with behavioral information
whenever possible.
Initial M
6-month M
1-year M
Intraop NRI 1-year tNRI
Intraop Banded NRI
Initial Banded NRI
6-month Banded NRI
1-year Banded NRI
Band 1
Band 2
Band 3
Band 4
180
Clinical Units
160
140
120
100
80
60
40
20
0
Figure 1. Relationships over time among objective measures
(banded and single-channel tNRI) and psychophysical M levels
throughout the electrode array.
Band 1
Band 2
Band 3
Band 4
Single 1
Single 2
Single 3
Single 4
0.8
0.7
0.6
Amplitude (µV)
Across the array, the banded-tNRI profile was
roughly parallel to both M levels and T levels. In
contrast, the single-channel tNRI profile was significantly lower at the apical end of the array (Figure 1).
Single-channel tNRI growth functions were steeper
than banded-tNRI growth functions by a ratio of
2.34 (first fitting M level/intraoperative banded
tNRI). The tNRI values were lower for bandedNRI measurements compared to single-channel
NRI responses by a ratio of 3.13 (Figure 2). Loudness growth functions were parallel to banded-NRI
growth functions, but not to single-channel NRI
growth functions.
0.5
0.4
0.3
0.2
0.1
0
0
100
200
300
400
Clinical Units
Figure 2. Growth functions for banded- and single-channel NRI
measurements in a single subject.
Objective Measures 165
Relationship between Electrical Stapedial Reflex
Thresholds and Speech Perception in Adult Users
of the HiResolution Bionic Ear
Jace Wolfe, Ph.D.
Heather Kasulis, Au.D.
Hearts for Hearing Foundation, Oklahoma City, OK, USA
“These data indicate that
individuals can obtain
very favorable benefit
from cochlear implants
using programs with stimulation levels
based on objective measures.”
Optimizing a user’s outcome with a cochlear implant
is at least partially dependent upon creating an effective signal processing program for the individual user,
a task requiring the user to make loudness judgments
to determine the amount of electrical stimulation
that is most comfortable. However, some cochlear
implant users (such as infants, young children,
persons with cognitive disabilities, and persons with
age-related dementia) may not be able to the make
reliable loudness judgments necessary to determine
the amount of electrical stimulation that is deemed
to be most comfortable. Consequently, clinicians
have turned to objective measures of a user’s physiologic response to electrical stimulation to help in
creating appropriate program settings.
In particular, two objective measures have proven
useful in programming cochlear implants, the electrically evoked stapedial reflex threshold (ESRT)
and the electrically evoked compound action potential (ECAP). To date, there have been no published
reports describing speech recognition performance
for subjects using ESRT-based programs with the
HiResolution Bionic Ear System. Given the results
reported by Buckler and colleagues (Buckler et al,
2003), it is possible that the particularly strong relationship between the ESRT and most comfortable
listening levels (M levels) with the Bionic Ear may
result in even better speech recognition performance
for subjects using ESRT-based programs compared
to previous reports using other cochlear implant
systems. Therefore, the aims of this study were
(1) to characterize ESRTs in a group of adult Bionic
Ear users, (2) to characterize ECAPs in the same
group of subjects, (3) to determine the relationships
among ESRTs, ECAP thresholds, and M levels, and
(4) to evaluate speech recognition performance for a
subgroup of adults who were fit with programs based
solely on ESRTs.
166 Advanced Bionics® Auditory Research Bulletin 2007
These data indicate that individuals can obtain very
favorable benefit from cochlear implants using
programs with stimulation levels based on objective
measures (Wolfe and Kasulis, in press).
References
ESRT-Based Program
250
200
150
100
50
1-4
5-8
9-12
12-16
M Levels by Electrode Group
Figure 1. Mean M levels for conventional behaviorally based
programs and ESRT-based programs (n = 11).
Table 1. Speech perception scores (percent correct)
for 11 subjects using conventional behavioral
programming techniques and after
programming with ESRT-based parameters.
HINT in
CNC
CNC
CNC
Noise
Words
Words
Words
60 dB
50 dB SPL 60 dB SPL 70 dB SPL
SPL
SNR = 8 dB
Conventional Mean
Program
Buckler L, Dawson K, Overstreet E. (2003) Relationship between electrical stapedial
reflex thresholds and HiRes program settings: a potential tool for pediatric cochlear
implant fitting (white paper). Valencia, CA: Advanced Bionics Corporation.
Wolfe J, Kasulis H. Relationships among objective measures and speech perception
in adult users of the HiResolution Bionic Ear. Coch Impl Int (in press).
Conventional Program
300
Clinical Units
ESRTs and ECAP thresholds were measured in 19
postlinguistically deafened adults using either a CII
or HiRes 90K cochlear implant. The results showed
that ESRTs could be measured easily in a majority
of subjects (79% of participating subjects), whereas
ECAPs were measureable in 100% of the group.
There were close agreements between ESRTs and
M levels in the subjects’ behaviorally based programs.
In contrast, the correspondence between M levels and
ECAP thresholds was not as strong. New programs
were created with M levels based on the ESRTs for
11 subjects for whom ESRTs did not match behavioral M levels. (Figure 1 shows differences between
behavioral and ESRT-based M levels.) Programs
created using ESRTs as a guide for setting M levels
yielded better speech recognition than programs
using conventional behavioral measures of M levels
(Table 1).
ESRT-Based
Program
75%
35%
56%
60%
SD
17%
20%
19%
22%
Mean
78%
42%
62%
70%
SD
16%
21%
19%
15%
Note: SD = standard deviation
Objective Measures 167
Evaluation of Neural Refractoriness as a Tool for Setting
Implant Program Levels
Jace Wolfe, Ph.D.1
Jessica Ballard, Au.D.1
Andrew Quick, M.S.2
1Hearts for Hearing Foundation, Oklahoma City,
Oklahoma, USA
2Advanced Bionics, Valencia, California, USA
Electrically evoked compound action potentials
(ECAPs) often are used by clinicians to help set
stimulation levels for cochlear implant recipients.
However, only moderate correlations have been
shown to exist between ECAP thresholds and most
comfortable stimulation levels (M levels), especially
for recipients using sound processing strategies that
employ high rates of stimulation. Specifically, the
relationship between ECAP thresholds (which
are measured with low-rate stimuli) and M levels
(which are set with high-rate stimuli) is highly variable. Some users exhibit ECAP thresholds that are
similar to their perceptual thresholds of stimulation.
Other users have ECAP thresholds that exceed
their M levels.
One potential explanation for this considerable
intersubject variability is that neural refractoriness
differs among cochlear implant users. The primary
purpose of this study was to determine whether a
relationship exists between an objective measure of
168 neural refractoriness and program levels. Seventeen
adult HiRes 90K implant recipients participated in
the study. Three types of measurements were made
including (1) loudness growth functions for singlechannel and banded stimuli at low (29 pps) and high
(1933 pps) stimulation rates, (2) ECAP amplitude
growth functions for single-channel and banded
stimuli, and (3) a new objective measure that yielded
neural refractoriness (rate of recovery) curves.
ECAPs and neural refractoriness were measured
using a research version of the Neural Response
Imaging (NRI) software.
The measure of neural refractoriness was similar to
a modified version of the forward masking method
proposed by Miller and colleagues (2000). This
technique measures the ECAP elicited by a probe
stimulus that is preceded by a masker. When the
stimulus follows the masker closely in time, the
nerve should be in a refractory state and unable to
respond to the stimulus. As the delay between the
masker and probe is progressively increased, neurons
recover from the refractory state and respond to the
probe stimulus. Eventually, the delay between the
masker and probe is long enough so that no neurons
are in a refractory state and all are able to respond to
the stimulus (at which point the response amplitude
no longer grows). In this study, the masker was set at
the user’s maximum comfort level, and the stimulus
probe level was set to 85-95% of the masker level.
ECAP measurements were acquired with various
masker-probe delays to assess the rate of recovery of
the auditory nerve and to yield a refractory curve.
Advanced Bionics® Auditory Research Bulletin 2007
“These preliminary data do not follow a
clear pattern and suggest that
neural refractoriness cannot explain fully
the variability between ECAP responses
and program levels.”
Theoretically, one would expect a small value of
Tau to be associated with large differences between
ECAP thresholds elicited by low-rate stimulation
and behavioral thresholds for high-rate stimulation.
Conversely, greater values of Tau would be associated with small differences between ECAP thresholds and behavioral thresholds.
Figure 2 shows the relationship between Tau and
the difference (in Clinical Units) between low-rate
ECAP thresholds and high-rate behavioral thresholds. These preliminary data do not follow a clear
pattern and suggest that neural refractoriness cannot
explain fully the variability between ECAP responses
and program levels. Further work is exploring alternate stimulus parameters that may improve the efficacy of using Tau to estimate program levels from
ECAP thresholds.
S 14 (E 13)
Response Amplitude (µV)
400
S 04 (E 4)
350
300
τ = 254.8
250
200
150
100
τ = 932.5
50
0
0
500
1000
1500
2000
2500
Tau (µs)
Figure 1. Examples of neural refractory curves and their associated Tau values for two subjects. Small Tau values and steep
functions (Subject 14) indicate short recovery times whereas
larger Tau values and shallow functions (Subject 04) represent
long recovery periods.
tNRI - High-Rate T Level (Clinical Units)
The resulting refractory curves were fit by the exponential function 1 – e-t/Tau. Tau represents the
time constant and describes the rate of recovery of
the auditory nerve. Small values of Tau are associated with short recovery times and steep refractory
curves, while larger values of Tau are associated
with long recovery times and shallow refractory
curves. Figure 1 illustrates refractory curves and
their associated Tau values for two subjects.
180
160
140
120
100
80
60
40
20
0
0
500
1000
1500
2000
2500
Tau (µs)
References
Miller CA, Abbas PJ, Brown CJ. (2000) An improved method of reducing stimulus
artifact in the electrically evoked whole-nerve potential. Ear Hear 21:280–290.
Figure 2. Difference between tNRIs (low rate) and T levels (high
rate) as a function of Tau. No clear pattern is evident from these
preliminary data.
Objective Measures 169
Comparison of Intraoperative Stapedius Reflex Thresholds
and Postoperative M Levels
Rolf-Dieter Battmer, Prof., Ph.D.1
Yassaman Khajehnouri, Dipl. Inform.1, 2
Andreas Büchner, Ph.D.1
Martina Brendel, Dipl. Ing.1
Timo Stöver, Prof, M.D., Ph.D.1
Burkhard Schwab, Ph.D.1
Thomas Lenarz, Prof., M.D., Ph.D.1
1 Medizinische Hochschule Hannover, Hannover, Germany
2 Advanced Bionics, Europe
“ The ability to measure ESRTS
intraoperatively in all 40 subjects
indicates that speech burst stimulation
may be a much more effective method
for eliciting ESRTs
than single-electrode stimulation.”
The electrically evoked stapedius reflex threshold
(ESRT) represents an objective measure that may
be used to evaluate the most comfortable electrical
stimulation levels (M levels) required for a particular
cochlear implant recipient. With single-channel
stimulation, ESRTs typically fall between the level
of comfortable loudness and the level experienced
as uncomfortably loud (Hodges et al, 2003, Stephan
& Welz-Müller, 2000). An alternative stimulation
method is to stimulate a group of adjacent electrodes
(usually four) sequentially at the rate at which these
electrodes would be driven during normal cochlear
implant operation. ESRTs measured using this technique correlate highly with the M levels measured
using Speech Bursts in the SoundWave programming
software in adults (Buckler et al, 2003). Consequently, ESRTs may be a very promising objective
measure that can be used to set M levels, especially
for implant recipients who are not able to provide
behavioral judgments of loudness.
170 The objective of this prospective clinical study was
to measure ESRTs intraoperatively using speech
burst stimulation and to relate those levels to postoperative behavioral M levels. In the first phase of
the study, ESRTs were determined intraoperatively
for 20 adult subjects. The threshold was defined
as the smallest movement of the stapedius tendon
that could be observed via the operating microscope. ESRTs were measured by stimulating seven
electrodes spanning the implanted array. Subjects
subsequently were fit with a sequential 16-electrode program using SoundWave software and
the default parameter settings for pulse width and
T levels. Speech Bursts were used to determine M
levels. Postoperative programs were created without
the programmer knowing the ESRTs measured intraoperatively. Intraoperative ESRT values then were
correlated to the M levels in each subject´s program
three months after first fitting. In the second phase
of the study, intraoperative ESRTs were determined
for 20 children. The children were fit postoperatively
with paired 16-electrode programs. Again, Speech
Bursts were used to determine M levels. ESRT-M
level correlations were calculated in the same way as
in the adult group.
For the adults, ESRTs could be determined in all
subjects. On average, the M levels were about 75%
of the ESRTs (Figure 1). Excluding two outliers,
an average correlation of 0.53 (p < 0.05) was found
between the M levels at three months and the intraoperative ESRTs. For the children, ESRTs also
could be determined for all subjects. On average,
the M levels were at 54% of the ESRTs (Figure 2).
The overall average correlation was 0.33 (p < 0.05).
Advanced Bionics® Auditory Research Bulletin 2007
M Level
ESRT
250
Correlations
Clinical Units
200
0.23
0.21
0.44
0.62
0.74
3–6
5–8
7–10
9–12
0.77
0.63
150
100
50
0
1–4
11–14
13–16
Electrode Groups
Figure 1. Average three-month M levels and intraoperative
ESRTs for seven electrode groups in adults (n = 20). The average
correlation coefficient across the array was 0.53.
M Level
300
250
Clinical Units
The results of this study, as well as findings from
earlier studies, show that ESRTs can be a valuable tool for estimating postoperative M levels. The
ability to measure ESRTs intraoperatively in all
40 subjects indicates that speech burst stimulation
may be a much more effective method for eliciting
ESRTs than single-electrode stimulation. Moreover,
because HiRes can be programmed using Speech
Bursts, ESRTs measured using speech-burst stimulation bring even more value to the fitting process
than in conventional strategies, particularly for
children. For example, the clinician can begin
programming by stimulating a child with the ESRT
profile at very low levels and increase the levels based
on the behavioural reactions of the child. The fact
that the correlations between ESRTs and M levels
were lower for children than adults may be explained
partly by the early measurement time (three months
after initial fitting). At this point, the children’s
programs may not have stabilized as they do
typically in adults. Thus, the correlations might
increase if they were based upon M levels measured
at later times.
ESRT
Correlations
0.34
0.33
0.32
0.31
0.36
1–4
3–6
5–8
7–10
9–12
Electrode Groups
0.31
0.38
11–14
13–16
200
150
100
50
References
Buckler L, Dawson K, Overstreet E, Luetje CM, Thedinger BS. (2003) Relationship
between electrically evoked stapedius reflex measurements and HiResolution
sound processing program settings: potential tool for pediatric fittings. Paper
presented at the 9th Symposium on Cochlear Implants in Children, Washington, DC,
24-26 April, 2003.
Hodges AV, Butts SL, King JE. (2003). Electrically evoked stapedial reflexes: utility
in cochlear implant patients. In: Cullington HE, ed. Cochlear Implants: Objective
Measures. London: Whurr Publishers, 81-93.
0
Figure 2. Average three-month M levels and intraoperative
ESRTs for seven electrode groups in children (n = 20). The average correlation coefficient across the array was 0.33.
Stephan K, Welz-Müller K. (2000) Post-operative stapedius reflex tests with
simultaneous loudness scaling in patients supplied with cochlear implants. Audiol
39:13-18.
Objective Measures 171
Relationships among tNRIs, ESRTs, and Psychophysical
Measures of Loudness: Implications for Programming
the HiResolution Bionic Ear
Sarah F. Poissant, Ph.D.
Eva M. Bero, M.A.
Daniel J. Lee, M.D.
University of Massachusetts, Amherst, MA, USA
University of Massachusetts Memorial Medical Center,
Worcester, MA, USA
“The study results show that
a combination of tNRI and BESRT data
may be the most accurate predictor of
M levels. Moreover, clinicians may find
OMs useful in troubleshooting
underperforming adult CI patients.”
Electrically evoked responses of the auditory
system can play an important role in the fitting and
troubleshooting of cochlear implants—especially
when behavioral responses are difficult to quantify
(for example, in adults with prelinguistic hearing
loss or in very young children). The objectives of
this prospective multicenter clinical study were
(1) to develop clinical guidelines for using Neural
Response Imaging thresholds (tNRIs) and electrically evoked stapedial reflex thresholds (ESRTs) to
assist in setting and optimizing HiRes programs
and (2) to develop clinical guidelines for using
tNRIs and ESRTs as troubleshooting tools in situations where benefit is not consistent with predicted
performance.
Forty-seven adult unilateral CII or HiRes 90K device
recipients were recruited from six implant centers.
Four patients had prelinguistic onset of deafness and
43 had postlinguistic onset of deafness. The mean
duration of severe-to-profound hearing loss was
11.22 years (SD = 15.19), mean age at implantation
was 56.53 years (SD = 15.66), and the mean duration
of implant use was 1.35 years (SD = 1.02). Subjects
provided demographic information (hearing loss
history, hearing aid history, M levels) and underwent speech perception and objective measures
(OMs) testing. The OM measures included tNRIs
and single-electrode ESRTs (SESRTs) obtained
from electrodes 3, 7, 11, and 15, as well as bandedelectrode ESRTs (BESRTs) obtained by stimulating four electrode groups comprised of electrodes
1-4, 5-8, 9-12, and 13-16. The numbers of subjects
contributing OM data varied by test because of time
constraints.
Overall response rates for OM measures were as
follows: tNRIs = 96% (184 attempted), BESRTs
= 84% (124 attempted), and SESRTs = 67%
172 Advanced Bionics® Auditory Research Bulletin 2007
Participating Centers and Clinics in North America
Carle Clinic Association, Urbana, IL
L’Hôtel-Dieu de Québec, Québec City, QU
Hearts for Hearing Foundation, Oklahoma City, OK
Sunnybrook & Women’s College Health Sciences Center, Toronto, ON
University of Massachusetts, Amherst, MA/University of Massachusetts Memorial Medical Center, Worcester, MA
University of Minnesota, Minneapolis, MN
Group data showed several trends when OM
measures were compared to behavioral M levels.
tNRIs were notably lower than M levels, BESRTs
were higher than or equal to M levels, and SESRTs
were notably higher than M levels. Figure 1 illustrates the relationship between tNRIs, BESRTs,
SESRTs, and M levels for the 10 study subjects
in whom all four OM measures were obtained.
Because of specific intercenter differences, regression
analyses were conducted using data from 19 subjects
at two study centers. Regression equations were used
to predict M levels. The statistic R2 represents the
percentage of the variance in either banded M levels
or average M levels across the array (the dependent
variables in the study) that is accounted for by the
OM results (the predictor variables). When tNRIs
were taken alone, R2 ranged from .676 to .807
depending upon the electrode location along the
array. R2 for BESRTs and SESRTs (from electrodes
3, 7, 11 only) ranged from 0.683 to 0.805 and 0.638
to 0.871, respectively. When tNRIs and SESRTs
were grouped (electrodes 3, 7, 11 only), R2 ranged
from 0.781 to 0.997. Furthermore, when these
two OMs were used to predict the average M level
across the entire array, the R2 value was 0.878. The
best predictions were obtained when BESRTs and
tNRIs were combined, resulting in R2 values ranging
between 0.788 to 0.983 for banded M levels, and
from 0.788 to 0.946 for average M levels.
NRI
350
BESRT
SESRT
Banded M
300
Clinical Units
(54 attempted). The ability to obtain a response
varied across the length of the electrode array. For
example, response rates for SESRT at electrodes 3,
7, 11, and 15 were 79%, 85%, 64%, and 38%, respectively. The mean CNC word score obtained at 70
dB SPL was 42% (SD = 29.06, range = 0-92%). No
significant correlations were found between CNC
scores and tNRIs, BESRTs, or SESRTs (p >.05).
250
200
150
100
50
0
Band 1
Band 2
Band 3
Band 4
Figure 1. Average tNRIs, BESRTs, SESRTS, and banded M levels
for 10 subjects for whom all four measures were available.
Band 1 includes data for tNRI 3, BESRT 1-4, SESRT 3, and
banded M 1-4; Band 2 includes data for tNRI 7, BESRT 5-8,
SESRT 7, and banded M 5-8; Band 3 includes data for tNRI 11,
BESRT 9-12, SESRT 11, and banded M 9-12; Band 4 includes
data for tNRI 15, BESRT 13-16, SESRT 15, and banded
M 13-16.
—continued on next page—
Objective Measures 173
—continued from previous page—
NRI
Banded M
350
300
250
200
150
100
Clinical Units
50
0
E3
E7
NRI
E 11
E 15
Banded M
350
300
250
200
150
100
50
0
Band 1–4
Band 5–8
Band 9–12
Band 13–16
Figure 2. tNRI (3, 7, 11, and 15) and banded M levels (1-4, 5-8,
9-12, and 13-16) for all subjects (top panel) and for four underperforming subjects (bottom panel). Note the inverse relationship between tNRIs and banded Ms despite nearly equivalent
average tNRIs for both groups of subjects.
174 Four of the study subjects had histories that would
have predicted good speech perception performance
(that is, postlinguistic onset of hearing loss, relatively
short duration of severe hearing loss), but they were
experiencing limited benefit from their devices.
Mean tNRIs and M levels from these four subjects
were compared with the group data (Figure 2). The
underperforming patients were found to have tNRIs
that were higher than their M levels. In contrast, the
remaining subjects with speech perception scores
appropriate to their histories had tNRIs that were
lower than their M levels. The mean tNRIs of this
group were only 11 clinical units off the average
tNRIs of all study subjects, yet their M levels were
greater than one standard deviation below the mean
of all subjects. These comparisons indicate that there
are no apparent physiological differences in neural
responses between the underperformers and the
group at large. Rather, the underperformers may
have been programmed with M levels that were too
low for optimal performance. Thus, the OM data
suggest that benefit might be improved by raising M
levels in these four subjects.
In summary, OMs are useful tools in the programming of adult cochlear implant recipients.
The study results show that a combination of
tNRI and BESRT data may be the most accurate
predictor of M levels. Moreover, clinicians may find
OMs useful in troubleshooting underperforming
adult CI patients. A careful analysis of the relationship between OM thresholds and M levels may
provide a starting point for modifications to implant
program parameters.
Advanced Bionics® Auditory Research Bulletin 2007
Amplitude Growth and Spatial Spread of
Neural Responses in HiRes 90K Users
Annelies Vermeiren, M. Aud.1
Andrzej Zarowski, M.D.1
Erwin Offeciers, Prof., M.D., Ph.D.1
Filiep Vanpoucke, Dr. Ir.2
Philippe Dykmans, M.Eng.2
1 Medical Center Sint Augustinus, Wilrijk, Belgium
2 Advanced Bionics European Research Center,
Antwerp, Belgium
The Research Studies Platform–Objective Measures
(RSPOM) software is an advanced Neural Response
Imaging (NRI) application that has been developed
by the Advanced Bionics European Research Center.
The NRI module in the clinical SoundWave software measures neural amplitude growth functions
using alternating polarity electrical artifact reduction. RSPOM extends this capability while at the
same time providing a similarly easy to use graphical
user interface and database.
The current version of the RSPOM software
provides three methods for measuring amplitude
growth functions, all of which reflect how the
auditory nerve responds to changes in stimulus
intensity. The first two methods use different electrical artifact suppression techniques— alternatingpolarity (AP) and masker-probe (MP). The third
method, termed SmartNRI™, is a NRI threshold
estimation algorithm that automatically detects
whether a trace contains a biological response.
RSPOM also provides two methods for studying
the spatial response of the auditory nerve, or spread
of excitation. The first method, SOE-VR, varies the
recording position across the electrode array while
keeping the probe position fixed. Electrical artifact
is reduced using the alternating polarity approach.
In the second method, SOE-VM, the probe and
the recording position are fixed, and the position of
a masker pulse is varied. A masker-probe approach
is used for electrical artifact reduction. Finally,
RSPOM provides a recovery function algorithm
for studying the temporal aspects of the auditory
nerve response.
“. . .spread of excitation measurements
provide limited added value to the
EFIM measurements. In contrast,
spatial selectivity profiles are a purer way
of assessing channel interaction.”
In an initial evaluation of RSPOM, all six NRI
methods were evaluated in 10 adult HiRes 90K
users. In addition, electrical field measurements were
made using the EFIM research software. A comparison of the amplitude growth functions showed
good overall correspondence between the alternating
polarity (AP) and masker-probe (MP) techniques.
MP responses exhibited a slightly lower noise floor.
Furthermore, in the MP measurements, the electrical artifact changed size and shape as stimulation intensity increased, making it more difficult to
separate the biological response from the electrical
stimulus artifact.
The spatial profiles were modeled using a two-sided
exponential decay curve. A comparison of the width
of the NRI spread of excitation curves and the EFIM
electrical fields revealed that the two measures were
broad and had similar slopes. We hypothesize that
these measurements reflect the electrical volume
conduction of a very small neural response localized
near the stimulation electrode. Therefore, the spread
of excitation measurements provide limited added
value to the EFIM measurements. In contrast,
spatial selectivity profiles are a purer way of assessing
channel interaction. Their width is restricted to plus
or minus four contacts, which is more in line with
psychoacoustic forward masking patterns.
Objective Measures 175
Channel Interaction in Cochlear Implant Users
Qing Tang, M.S.1
Raul Benitez, Ph.D.2
Fan-Gang Zeng, Ph.D.1
1 University of California, Irvine, CA, USA
2 Technical University of Catalonia, Barcelona, Spain
For many years, channel interaction has been thought
to underlie the limitations and variability in cochlear
implant performance (Hanekom and Shannon, 1998;
White et al, 1984; Wilson et al, 1991). However,
more recent studies have shown that channel interaction actually can increase the number of perceptual spectral channels within the limited number of
physical electrode contacts (Donaldson et al, 2005;
Koch et al, 2007; Kwon and van den Honert, 2006;
McDermott and McKay, 1994; McKay et al, 1996).
This channel interaction has been exploited using
current steering in the HiRes with Fidelity 120
sound processing algorithm implemented in the
Advanced Bionics cochlear implant. This algorithm
potentially can provide up to 120 spectral bands to
CII Bionic Ear and HiRes 90K recipients.
Channel interactions can arise at three levels within
the auditory system—physical, neural, and perceptual (Figure 1). At the physical level, electric fields
can sum together when electrode contacts are stimulated simultaneously. At the neural level, preceding
stimuli can prevent neurons from responding to
subsequent stimuli during the neural refractory
period. At the perceptual level, the brain can sum
Peripheral
Electrical
Stimuli
Central
Tissue
& Peripheral
Auditory
Nerve Fiber
Brain
Electrical
Field Interaction
Neural
Interaction
Perceptual
Interaction
Percepts
Electrical
Field Imaging
ECAP
Psychophysical
Experiments
Figure 1. Block diagram describing electrode interaction mechanisms at different levels.
176 the loudness of two simultaneous or non-simultaneous stimuli. Thus, the additional spectral channels
experienced by implant listeners may take advantage
of channel interactions at all three levels.
This study explored channel interactions at the
physical, neural, and perceptual levels in five CII and
HiRes 90K implant users (Figure 2). The Advanced
Bionics system incorporates reverse telemetry, which
allows measurement of electric field potentials and
the electrical compound action potential (ECAP).
At the physical level, electric field imaging (EFI)
spatial profiles were obtained by applying an electric
pulse at one electrode (probe) while recording the
voltages at the remaining electrodes. There was little
variability in the EFI profiles among the subjects,
indicating strong similarities in channel interactions
at the level of electrical stimulation. An important
implication of these physical data is that the potential of individuals to fully utilize virtual channels
must arise at the neural level or higher (for example,
in the patterns of surviving neurons).
At the neural level, ECAPs first were measured
as a function of current level on all electrodes. The
slopes of the ECAP growth functions then were
determined using linear regression. There was a
large degree of variability among subjects, which
agreed with the variability in psychophysical thresholds. Neural channel interaction was assessed with
various ECAP masking experiments. The asymmetries in the interaction patterns suggested that there
were asymmetries in the surviving populations of
neurons. The asymmetric neural interaction patterns
were different from the interaction patterns observed
with EFI, which were uniform across subjects. At the
perceptual level, channel interaction was explored
using threshold detection experiments. Consistent
with previous studies (e.g., Favre and Pelizzone,
1993; Boex et al, 2003), perceptual thresholds were
lowered by in-phase, subthreshold simultaneous
stimulation on two electrodes, but elevated by out-
Advanced Bionics® Auditory Research Bulletin 2007
“These results indicate that channel interaction measurements should be integrated
into the sound processor fitting process in
order to optimize implant benefit.”
of-phase stimulation. Generally, channel interaction decreased as the spatial separation between two
electrodes increased. Moreover, monopolar stimulation resulted in a broad electric field distribution
encompassing the whole electrode array even at the
threshold level. Channel interactions at the perceptual level were highly related to the EFI interaction
patterns (r2 = 0.85, sd = 0.14). This close relationship
indicated that perceptual interaction at the threshold
level was mainly a result of direct current summation at the physical electrodes. We defined complete
channel interaction as the point where electrical
field summation of two electrodes functioned effectively as a single neural channel. For two of the five
subjects, complete channel interaction occurred at as
many as six electrodes apart. These results indicate
that channel interaction measurements should be
integrated into the sound processor fitting process in
order to optimize implant benefit.
1
.8
.6
EL1 (0)
References
Boex C, de Balthasar C, Kos MI, Pelizzone M. (2003) Electrical field interactions in
different cochlear implant systems. J Acoust Soc Am 114:2049-2057.
Donaldson GS, Kreft HA, Litvak L. (2005) Place-pitch discrimination of single- versus
dual-electrode stimuli by cochlear implant users. J Acoust Soc Am 118:623-626.
Favre E, Pelizzone M. (1993) Channel interactions in patients using the Ineraid
multichannel cochlear implant. Hear Res 66:150-156.
Hanekom JJ, Shannon RV. (1998) Gap detection as a measure of electrode interaction
in cochlear implants. J Acoust Soc Am 104:2372-2384.
Koch DB, Downing M, Osberger MJ, Litvak L. (2007) Using current steering to
increase spectral resolution in CII and HiRes 90K users. Ear Hear 28:38S-41S.
Kwon BJ, van den Honert C. (2006) Dual-electrode pitch discrimination with
sequential interleaved stimulation by cochlear implant users. J Acoust Soc Am
120:EL1-6.
McDermott HJ, McKay CM. (1994) Pitch ranking with nonsimultaneous dualelectrode electrical stimulation of the cochlea. J Acoust Soc Am 96:155-162.
McKay CM, McDermott HJ, Clark GM. (1996) The perceptual dimensions of singleelectrode and nonsimultaneous dual-electrode stimuli in cochlear implantees.
J Acoust Soc Am 99:1079-1090.
White MW, Merzenich MM, Gardi JN. (1984) Multichannel cochlear implants.
Channel interactions and processor design. Arch Otolaryngol 110:493-501.
Wilson BS, Finley CC, Lawson DT, Wolford RD, Eddington DK, Rabinowitz WM.
(1991) Better speech recognition with cochlear implants. Nature 352:236-238.
EL9 (0)
ECAP
CII
EFI
EL15 (0)
Normalized ECAP, CII and EFI
.4
.2
0
1
.8
.6
S1
.4
.2
0
1
.8
.6
S2
.4
.2
0
1
.8
.6
S3
.4
.2
0
1
.8
.6
S4
.4
.2
0
S5
2
4
6
8
10
12
14
16
2
4
6
8
10
12
14
16
2
4
6
8
10
12
14
16
Masking (Recording) Electrode
Figure 2. Normalized neural (ECAP), psychophysical (CII) and physical (EFI) spatial profiles at three stimulating sites (columns) from
five subjects (rows).
Objective Measures 177
Clinical Tools
As cochlear implant technology continues to evolve, implant recipients are
reaching ceiling performance on traditional hearing assessments. Thus, a need
has arisen for new and more challenging measures of implant benefit.
The studies summarized in this section explore the development and clinical
application of assessment tools for cochlear implant recipients of all ages and
highlight the potential use of self-report questionnaires as tools for evaluating
implant success.
179
Everyday Listening Questionnaire: A New Evaluation Tool
for Adult Cochlear Implant Users
Jenny Goldring, M.Sc.
Alexandre Gault, M.Sc.
Advanced Bionics, Europe
“The ELQ2 now has been distributed
and can be used by clinicians to identify
hearing deficits that may not be apparent
during clinical testing.”
The Everyday Listening Questionnaire (ELQ) was
designed to help Advanced Bionics and clinics assess
how well cochlear implant recipients cope with the
challenges of common everyday listening situations.
The ELQ also serves to evaluate the practicality and
usefulness of various implant accessories and helps to
develop everyday listening tips for cochlear implant
users. The original version was piloted at six implant
centers and within a French user-contact network.
The feedback from that study was used to create an
improved version called ELQ2. The ELQ2 questionnaire requires approximately 30 minutes to complete,
and it has four major areas of inquiry: telephone use,
music, workplace, and social environment.
In this study, 73 adults completed the ELQ2. The
mean age of the subjects was 51 years and the mean
daily implant use was 15 hours per day. Improving
participation in social activities was rated as the most
significant motivation for seeking a cochlear implant.
Subjects reported using their cochlear implants in
all listening situations. The average self-assessment
of speech understanding (on a scale from 1 = very
difficult to 5 = very easy) was 4.4 in quiet and fell to
2.2 in the presence of noise. The majority of subjects
used telephones, went to the cinema, visited the
theatre, and listened to music. Ninety-two percent
used the telephone for an average of 5-10 times
per week. Ninety-three percent reported frequently
listening to music for enjoyment and most could
follow familiar melodies, recognize a familiar vocal
or instrumental track, and identify musical instrument categories.
Subjects also rated the frequency of use of
different earhooks and microphone configurations
(Figure 1). The most commonly used accessory was
180 Advanced Bionics® Auditory Research Bulletin 2007
Participating Centers and Clinics in Europe
Cochlear Implant Centrum Rhein-Main, Friedberg, Germany
Emmeline Cochlear Implant Programme, Cambridge, United Kingdom
Kurzel Rehabilitation Centre, Aachen, Germany
Medizinische Hochschule Hannover, Hannover, Germany
Nottingham Cochlear Implant Programme, Nottingham, United Kingdom
Medical Center Sint Augustinus, Antwerp, Belgium
Standard Earhook
Telephone Adapter
Auria T-Coil
Aux Microphone
Amplified Headset
T-Mic
4
3
2
1
s
Co
nf
er
e
Ca nce
lls
nt
ie
Cl
s
ue
ag
lle
Co
d
Go erin
od g
s
Or
Ap
po
in
tm
en
ts
ie
n
Fa ds
m &
ily
0
Fr
In a subgroup of subjects where speech test data
were available, the subjective questionnaire ratings
and objective speech results were compared. There
was no correlation (r = .115, df = 36, t = .69)
between the subjective ratings of speech understanding in background noise and the actual speech
test results in noise (HSM sentence test, +10 dB
SNR)—confirming the notion that speech testing
in controlled clinic settings does not always reflect
real-life situations. In quiet, 28 of 38 subjects tested
scored between 95-100% on the HSM sentence
test. This test therefore exhibited a ceiling effect,
thereby highlighting the need for more sensitive
clinical tests.
5
Mean Rating
the T-Mic, which generally achieved the highest
ratings for its effectiveness in difficult listening situations and was found particularly useful for listening
to music and hearing in challenging work environments. The only situation where the T-Mic was not
the most commonly used accessory was for listening
to music on an MP3 player or Walkman. In those
situations, the direct audio input adapter cable was
more common.
Figure 1. Earhook option ratings for different listening situations. The T-Mic was rated as the most frequently used
accessory option.
The ELQ2 now has been distributed and can be used
by clinicians to identify hearing deficits that may not
be apparent during clinical testing. These weaknesses
then can be addressed with respect to using the
implant and its accessories to their greatest benefit.
Feedback on accessories will guide Advanced Bionics
toward improved designs and will help professionals
to counsel users on how to use the accessories more
fully. The ELQ2 is available in English, French,
and German from Advanced Bionics. A pediatric
version for young children and teenagers also has
been developed and will be piloted in 2007.
Clinical Tools 181
Cochlear Implant Outcomes and Quality of Life in Adults
with Prelinguistic Deafness
W. Martin C. Klop, M.D.
Jeroen J. Briaire, M.Sc.
Anne M. Stiggelbout, Ph.D.
Johan H. M. Frijns, M.D., Ph.D.
Leiden University Medical Center, Leiden, The Netherlands
Cochlear implantation now is recognized as a
highly effective treatment option for congenitally
and prelinguistically deafened children as well as
for postlinguistically deafened adults. Until recently,
prelinguistically deafened adults were considered
poor implant candidates because improvement in
speech recognition was limited. However, recent
studies (e.g., Teoh et al, 2004) have suggested that
the latest implant technology may lead to improved
benefits in prelinguistically deafened adults. The first
aim of this study was to evaluate the effect that late
cochlear implantation has on speech perception in
prelinguistically deafened adults. The second aim was
to assess quality of life (QoL) changes produced by
implantation in the same subjects. The third aim was
to determine the demographic factors that influence
postimplant performance in prelinguistically deafened adults implanted at a later age.
Eight prelinguistically deafened subjects participated
in the study. They all met the Leiden University
Medical Center implant criteria for postlinguistiMean
100
Percent Correct
80
60
40
20
0
pre-op
1
2
4
8
12
26
52
104
Duration of Implant Use (weeks)
Figure 1. Mean and individual phoneme scores on a CVC word
test in quiet (65 dB SPL) measured preoperatively and after 1, 2,
4, 8, 12, 26, 52, and 104 weeks of implant use for 8 subjects.
182 cally deafened adults, and they frequently used
lipreading in their everyday communication. In
addition, all subjects experienced an onset of severe
hearing impairment before the age of four years,
were at least 16 years old at implantation, received
a “state of the art” device, were involved in an auditory-oral environment, and showed substantial motivation towards auditory communication. The eight
adults were implanted with a CII Bionic Ear (mean
age at implantation = 36 years, range: 21-55 years)
and were followed for at least two years following
implantation. Speech perception was evaluated in
free field using the Dutch Society of Audiology
standard CVC (monosyllabic) word list presented at
65 dB HL (Frijns et al, 2002). QoL was evaluated
preoperatively and postoperatively at 4-5 months
(shortly after completion of the initial training), 12
months, and 30 months. Three different instruments
were used to measure QoL—the Health Utility
Index (HUI-Mark II, Torrance et al, 1996), the
Nijmegen Cochlear Implant Questionnaire (NCIQ,
Hinderink et al, 2000), and a Visual Analogue
Scale (VAS) for hearing and health. The relationships between eight demographic variables (gender,
communication mode, hearing aid use, educational
background, preoperative hearing thresholds, preoperative CVC scores, duration of deafness, and age at
implantation) and the postoperative CVC phoneme
score were investigated. In addition, the quality of
the implant user’s own speech production (QoSP)
was assessed.
Results showed that postimplant speech perception
scores improved over time. This improvement was
more pronounced for phoneme scores than for word
scores (Figure 1). The mean phoneme score improved
significantly from 14% preimplant to 43% after 24
months of implant use (p = 0.001). The mean word
Advanced Bionics® Auditory Research Bulletin 2007
Different aspects of QoL improved significantly four
to five months after implantation, but no significant
changes occurred thereafter. Closer inspection of
individual attributes in the HUI-Mark II questionnaire revealed that the only significant improvement
occurred in the domain “sensation” (containing the
subdomains vision, hearing, and speech) (paired ttest, p = 0.002) (Figure 2). The hearing-related questionnaire (NCIQ) was more sensitive and measured
significant improvements in basic sound perception,
advanced sound perception, and social interaction.
The improvement in hearing acuity that subjects
rated on the VAS was significant (p = 0.009), while
general health did not change. A cost-utility result
of 18,872 euros per Quality Adjusted Life Year
(QALY) was calculated, showing that cochlear
implantation is a cost-effective intervention. There
were no significant correlations between postimplant
speech perception and preoperative factors. Nevertheless, an interesting positive correlation was found
between postoperative speech perception and QoSP.
This relationship is in line with the hypothesis that
postoperative hearing capabilities depend on the
viability of the central auditory system in long-term
deafness. QoSP might be considered an index representing the extent to which cochlear implant users
were able to hear speech prior to implantation.
In summary, speech perception and quality of life
improved in prelinguistically deafened adults when
they were implanted with state-of-the-art devices.
This study also suggests that the prognostic value of
QoSP should be investigated further.
Mean
Sensation
Cognition
Mobility
Self Care
Emotion
Pain
1
Mean Rating
score improved significantly from 2% to 15% for the
same time period (p = 0.009). However, individual
scores differed considerably for both phonemes and
words.
“. . .speech perception and quality of life
improved in prelinguistically deafened
adults when they were implanted
with state-of-the-art devices.”
0.8
0.6
0.4
0.2
0
pre-op
4
12
30
Duration of Implant Use (months)
Figure 2. Mean pre- and postoperative quality of life scores
measured preoperatively, and at 4, 12, and 30 months for the
HUI-II. A statistically significant change occurred between the
preoperative and four-month ratings for sensation and for the
overall mean.
References
Frijns JH, Briaire JJ, De Laat JA, Grote JJ. (2002) Initial evaluation of the Clarion CII
cochlear implant: speech perception and neural response imaging. Ear Hear 23:184197.
Hinderink JB, Krabbe PF, Van Den Broek BP. (2000) Development and application
of a health-related quality-of-life instrument for adults with cochlear implants: the
Nijmegen cochlear implant questionnaire. Otolaryngol Head Neck Surg 123:756765.
Teoh SW, Pisoni DB, Miyamoto RT. (2004) Cochlear implantation in adults with
prelingual deafness. Part I. Clinical results. Laryngoscope 114:1536-1540.
Torrance GW, Feeny DH, Furlong WJ, Barr RD, Zhang Y, Wang Q. (1996)
Multiattribute utility function for a comprehensive health status classification system.
Health Utilities Index Mark 2. Med Care 34:702-722.
Clinical Tools 183
More than the Words: Cochlear Implant Users’
Perception of the “Indexical” Properties of Speech
Rosalie M. Uchanski, Ph.D.
Laura K. Holden, M.A.
Washington University School of Medicine, St. Louis, MO, USA
“The individual scores for
the four implant users vary
but are generally poorer than
the scores of normal-hearing listeners.”
When we listen to the speech of others, we hear
more than the words that are spoken. We also hear
consciously, or subconsciously, substantial information about the talker. After listening to a talker for
a brief amount of time, we can usually answer questions such as: Is the talker a male or a female? Is the
talker a child, a young adult, or an elderly person?
Does the talker have a regional or foreign accent? Is
the talker tired or full of energy? What is the mood
of the talker? These properties of the speech message
associated with the talker are called “indexical” properties of speech (Abercrombie, 1967).
While it is vital to complete comprehension to
understand the words in a speech message, it is often
also important to perceive the indexical properties
of speech. For example, the words and the sarcastic
mood of a talker must be heard for a person to
receive the talker’s intended message. For people
with normal hearing, the ability to perceive indexical
properties is most often assumed. Studies have shown
that for normal hearers it is easy to distinguish male
talkers from female talkers, to distinguish one talker
from another talker of the same sex, and to tell (fairly
reliably) the approximate age of a talker (Lass et al,
1976; Ptacek & Sander, 1966; Schwartz & Rine,
1968; Weinberg & Bennett, 1971). However, people
with impaired hearing may have difficulty perceiving
the indexical properties of a speech message (e.g.,
Ross & Duffy, 1973). Recently, several researchers
have reported on some of these difficulties for
cochlear implant users (Cleary & Pisoni, 2002;
Fu et al, 2004; Luo et al, 2007; Spahr & Dorman,
2004). When discriminating between talkers,
these studies report consistently that implant users’
performance is much poorer than that of listeners
184 Advanced Bionics® Auditory Research Bulletin 2007
with normal hearing. And, for perceiving emotion in
speech, Luo et al (2007) report a similar result, that
is, cochlear implant users do not perceive emotions
nearly as well as listeners with normal hearing.
One acoustic cue that conveys talker characteristics is voice pitch, or fundamental frequency (F0).
However, in traditional implant speech processing
strategies, fundamental frequency may not be
preserved well in the signals that stimulate the electrodes. The HiRes with Fidelity 120 (HiRes 120)
strategy has been developed with the explicit intent
of improving frequency perception for cochlear
implant users. Since emotion identification and
talker discrimination may rely on good frequency
resolution, we hypothesize that the HiRes 120
strategy will improve implant users’ performance on
these types of tasks.
To test this hypothesis, we are examining emotion
identification and talker discrimination for listeners
using the standard HiRes and HiRes 120 strategies.
Subjects using the standard HiRes strategy were
recruited for this investigation that employed an
A-B-A-B trial design, as detailed in Table 1. Hence,
each test was performed twice (across two test
sessions) for each of the two strategies.
For the emotion identification test, the listener’s task
was to identify for each sentence presentation the
talker’s intended “mood” or “feeling” from a forcedchoice response set (angry, happy, sad, or scared). For
the talker discrimination tests, the listener’s task was
to indicate whether two sentences presented were
spoken by the same talker or by different talkers.
“Further study is required to determine
whether HiRes 120 improves perception
of these indexical properties compared to
other sound processing approaches.”
Table 1. Test sequence and test intervals
alternating between use of two processing strategies.*
Trial
A
B
A
B
Strategy
Own HiRes
HiRes 120
Own HiRes
HiRes 120
Interval
baseline
1 month
1 week
1 week
* Baseline performance established with each subject’s own standard
HiRes sound processing. Subsequent time intervals are approximate.
—continued on next page—
Clinical Tools 185
—continued from previous page—
Standard HiRes
100
HiRes 120
Percent Correct
80
60
40
20
0
S2
S4
S1
S3
Figure 1. Emotion identification performance for four cochlear
implant users using two processing strategies. Four-alternative
(angry, happy, sad, scared), forced-choice task. Chance performance is 25% correct.
Standard HiRes
100
HiRes 120
Percent Correct
80
60
Currently, four participants have completed the test
protocol. Figures 1 and 2 show the emotion identification and talker discrimination scores for these
subjects. For comparison, listeners with normal
hearing score 96% for emotion identification tasks,
100% for across-gender talker discrimination tasks,
and 97% for within-gender talker discrimination
tasks. The individual scores for the four implant
users vary but are generally poorer than the scores
of normal-hearing listeners. On average, the implant
subjects identified the four emotions with an accuracy of 61% correct. For the talker discrimination
tests, across-gender discriminations were easiest
(89% correct), within-male more difficult (71%), and
within-female most difficult (66%). For one subject
(S2), performance on within-gender tasks was much
better with HiRes 120 than with standard HiRes.
For the other listeners, there appeared to be little
difference in performance between the two strategies. Results from this small sample are preliminary.
Further study is required to determine whether
HiRes 120 improves perception of these indexical
properties compared to other sound processing
approaches.
40
20
References
0
Abercrombie D. (1967) Elements of general phonetics. Chicago: Aldine Publishing.
S2 S4 S1 S3
S2 S4 S1 S3
Across Gender
Within Female
S2 S4 S1 S3
Within Male
Figure 2. Talker discrimination performance for four cochlear
implant users using two processing strategies. Across-gender
discrimination consists of male versus female discrimination.
Task is two-alternative (same talker or different talker), forcedchoice. Chance performance is 50% correct.
Cleary M, Pisoni DB. (2002) Talker discrimination by prelingually deaf children with
cochlear implants: Preliminary results. Ann Otol Rhinol Laryngol 111(Suppl 189):
113-118.
Fu Q-J, Chinchilla S, Nogaki G, Galvin JJ. (2004) The role of spectral and temporal
cues in voice gender discrimination by normal-hearing listeners and cochlear
implant users. J Assoc Res Otolaryngol 5:253-260.
Lass NJ, Hughes KR, Bowyer MD, Waters LT, Bourne VT. (1976) Speaker sex
identification from voiced, whispered, and filtered isolated vowels. J Acoust Soc Am
59(3):675-678.
Luo X, Fu Q-J, et al. (2007) Vocal emotion recognition by normal-hearing listeners
and cochlear implant users. Trends Amplif 11:301-315.
Ptacek PH, Sander EK. (1966) Age recognition from voice. JSHR 9(2):273-277.
Ross M, Duffy RJ, et al. (1973) Contribution of the lower audible frequencies to the
recognition of emotions. Am Ann Deaf 118(1):37-42.
Schwartz MF, Rine HE. (1968) Identification of speaker sex from isolated, whispered
vowels. J Acoust Soc Am 44(6):1736-1737.
Spahr AJ, Dorman MF. (2004) Performance of subjects fit with the Advanced Bionics
CII and Nucleus 3G cochlear implant devices. Arch Otolaryngol Head Neck Surg
130:624-628.
Weinberg B, Bennett S. (1971) Speaker sex recognition of 5- and 6-year-old
children’s voices. J Acoust Soc Am 50(4):1210-1213.
186 Advanced Bionics® Auditory Research Bulletin 2007
New Study Init iat i ve
Optimization of Sound-Field Threshold Levels and Input Dynamic Range
for the Advanced Bionics Cochlear Implant System
Laura K. Holden, M.A.
Ruth M. Reeder, M.A.
Margaret W. Skinner, Ph.D.
Jill B. Firszt, Ph.D.
Washington University School of Medicine, St. Louis, MO
In order to best communicate in daily listening situations, it
the subjects’ processor and sound field threshold levels will
is critical for cochlear implant recipients to perceive speech
be measured for each program. The program that produces
at low levels. By doing so, recipients are able to understand
the best sound-field threshold levels will be chosen to
individuals who speak softly or from a distance, thereby
evaluate IDR settings.
expending less effort throughout the day to communicate
(Skinner et al, 2002). Perceiving soft speech is especially
At the second session, sound field threshold levels will be
important for overhearing indirect conversation and for
obtained (with the program that produced the best sound
incidental learning, both of which have specific implications
field thresholds) at three different IDR settings: 50, 65, and
for children who are learning to communicate (Flexer et al,
80 dB. Speech perception will be evaluated for monosyllabic
1999).
words at a soft level (50 dB SPL) and sentences in noise—
consisting of four-talker babble at individually set signal-
Previous research has shown that for cochlear implant
to-noise ratios—at a comfortably loud level (65 dB SPL). At
recipients to perceive soft speech, sound field thresholds
the third session, sound field threshold levels and speech
should fall below 30 dB HL across the frequency
perception testing will be repeated using the same three
range from 250-6000 Hz (Firszt et al, 2004; Holden
programs. The speech processor program from the two test
et al, 2005). The purpose of the study is to determine
sessions that results in the highest speech perception scores
(1) programming methods that will give Advanced Bionics
will be worn for two weeks and compared with the subject’s
cochlear implant recipients sound field threshold levels less
everyday program. At the conclusion of the study, subjects
than 30 dB HL from 250-6000 Hz and (2) whether a wider
will report through a questionnaire how the programs
input dynamic range (IDR) setting improves the perception of
compared in their daily life. Results from this study will assist
soft speech as well as speech in noise. Participants will be ten
in optimization of the Advanced Bionics cochlear implant
postlinguistically deafened adult recipients who use either
system for improved sound perception at soft levels and in
standard HiRes or HiRes 120 sound processing.
the presence of noise.
Subjects will participate in three study sessions. At the first
References
session, sound-field threshold levels will be obtained with the
Firszt JB, Holden LK, Skinner MW, Tobey EA, Peterson A, Gaggle W, RungeSamuelson CL, Wackym PA. (2004) Recognition of speech presented at soft
to loud levels by adult cochlear implant recipients of three cochlear implant
systems. Ear Hear 25:375-387.
subject’s everyday program. Ascending loudness judgments
(ALJ) and minimum (T) and maximum (M) stimulation levels
will be measured for all active electrodes. Three new speech
processor programs will be created using the M levels from
the subject’s everyday program. The T levels for each program
will be set (1) at 10% of the M levels, (2) at each subject’s
report of “barely audible” from the ALJ procedure, and
(3) at 10 Clinical Units below each subject’s report of “soft.”
All three speech processor programs will be downloaded to
Flexer CA. (1999) The importance of hearing and the impact of hearing
problems. In: Facilitating Hearing and Listening in Young Children. Second
Edition. San Diego: Singular Publishing Group, 1-22.
Holden LK, Vandali AE, Skinner MW, Fourakis MS, Holden TA. (2005) Speech
recognition with the Advanced Combination Encoder and the Transient Emphasis
Spectral Maxima strategies in Nucleus 24 recipients. J Speech Lang Hear Res
48:681-701.
Skinner MW, Binzer SB, Potts LG, Holden LK, Aaron RJ. (2002) Hearing
rehabilitation for individuals with severe and profound hearing impairment:
hearing aids, cochlear implants, and counseling. In: Valente M, ed. Strategies for
Selecting and Verifying Hearing Aid Fittings. New York: Theime, 311-344.
Clinical Tools 187
DVD Version of the UCL-CUNY Sentence Test
Wanda Aleksy, M.Sc.1
Francesca Pinto, M.Sc.1
Pippa Wilson, B.Sc.1
Laure Arnold, M.Sc.2
Patrick Boyle, M.Sc.2
1 Royal National Throat, Nose & Ear Hospital, London,
United Kingdom
2 Advanced Bionics, Europe
The CUNY (City University of New York) sentence
test was developed by Boothroyd and colleagues
(Boothroyd et al, 1985) as an audiovisual test of
sentence perception. The test consists of 26 lists of
12 topic-related sentences that have been translated
from American English to British English and rerecorded by the University College London (UCL)
cochlear implant program. Each sentence contains
between 3 and 15 words with length counterbalanced across topic. Table 1 shows a typical list from
the test. Testing may be conducted with or without
lipreading, either in quiet or in multitalker babble.
Table 1. Sample UCL-CUNY sentence list.
Topic
Sentence Item
1
Food
She made a fresh fruit salad.
2
Family
He shares a two bedroom flat in the city with
his two brothers.
3
Work
Where do you work?
4
Clothes
How much did they charge you for the
alterations?
5
Animals
Ask the neighbours to feed your fish when you
go on holiday.
6
Home
Tidy the house before the guests arrive.
7
Sports
It takes years of hard work to train to become a
professional football player.
8
Weather
9
Health
10
Hobbies
11
Money
You really should be able to save more of your
salary.
12
Music
Play that tune.
188 When the temperature goes down, it’s cold
outside.
What medicine are you taking?
Why don’t you find a hobby now that you’ve
retired?
Scoring and report generation is automated. The
UCL-CUNY test has become part of the standard
test battery in many clinics throughout the United
Kingdom. However, the original test required a laser
video disc, a technology that is no longer commercially available. Moreover, the original test setup was
cumbersome and required considerable clinic space.
Therefore, to overcome these obstacles, the UCLCUNY was remastered onto DVD (available from
Advanced Bionics, Europe). To preserve the test
materials and to update the test presentation mode,
the analog sound and video were recorded digitally
without compression onto DVD. The audiovisual
material was edited and individual files (.avi format)
were created for each sentence item using the “VirtualDubMod” software. Attention was paid to always
start and finish the video for each sentence item with
the speaker’s mouth closed. Next, using the Visual
Basic programming tool, software was developed to
meet the hardware system requirements, including
two monitors operating simultaneously—one “user”
monitor enabling lipreading and one “clinician”
monitor for control. Finally, the existing user interface was simplified and DVD control added.
This new system has been verified for sound and
video quality. As an alternative to the hardware
described, the new software may be installed on a
laptop computer connected to a flat (LCD) monitor,
making the whole system less bulky and potentially
portable. In the future, this interface also might be
used to create similar audiovisual speech tests in
other languages.
Reference
Aleksy W, Pinto F, Wilson P, Arnold L, Boyle P. Development of a DVD version of the
UCL-CUNY sentence test. Poster presented at the 9th International Conference on
Cochlear Implants, Vienna, Austria. 14-17 June, 2006.
Boothroyd A, Hanin L, Hnath T. (1985) A Sentence Test of Speech Perception:
Reliability, Set Equivalence, and Short Term Learning (Internal report, RCI 10). New
York, NY: City University of New York.
Advanced Bionics® Auditory Research Bulletin 2007
New Study Init iat i ves
Audiovisual Test for Pediatric Cochlear
Implant Users
ABQuest: Research Questionnaire
Management Software Pilot Study
David Canning, M.A., M.Ed.1
Patrick Boyle, M.Sc.2
Filiep Vanpoucke, Dr.Ir.1
Philippe Dykmans, M.Eng.1
Jenny Goldring, M.Sc.2
1 University College London, London, United Kingdom
2 Advanced Bionics, Europe
1 Advanced Bionics European Research Center,
Antwerp, Belgium
2 Advanced Bionics, Europe
Since cochlear implants have been available, the educational
opportunities for profoundly deaf children have increased
Questionnaire data are becoming more important in both
dramatically. Various estimates of educational placement
clinical studies and routine clinical practice. ABQuestTM is
suggest that 80% of young cochlear implant users are
an easy-to-use, flexible, and open software package that has
attending mainstream schools. However, learning in a
been developed to collect and analyze questionnaire data
mainstream classroom is not without difficulties. When
electronically. User profiles and all questionnaire results can
classrooms are busy and noisy, the ability to hear is
be stored in one program designed with a database structure
challenged for all children, especially for children who use
similar to that used in the SoundWave programming software.
cochlear implants.
Thus, the ABQuest software facilitates the comparison
of subjective data, objective measures, and program
Evidence from children who are failing to thrive in
parameters.
mainstream classes suggests that a major contributing factor
to successful mainstream education is the ability of a child to
Within Advanced Bionics, the ABQuest software has been
integrate visual and auditory information. Clinical evidence
evaluated with a range of questionnaire styles. The software
suggests that there are children who gain little, if any, benefit
makes it easy to create new questionnaires and incorporate
from visual information. It is important to identify this group
them into the software. The administration of questionnaires
as early as possible in order to inform habilitative options.
and storage of data works effectively and allows for
comparison of results with program levels and objective
An adaptive computerized test of audiovisual speech
measurements. Clear reports can be prepared and stored
understanding is being developed that will be suitable for
along with subject information.
children 3.5-12 years of age. The audio and video tokens
consist of words spoken by children in order to represent
In this pilot study, ABQuest will be beta tested in several
the communication situations likely to be encountered in
European clinics using the recently developed Everyday
everyday school environments. Each word is embedded
Listening Questionnaire. The questionnaire was designed to
in a carrier phrase and is presented in noise. Video signals
help Advanced Bionics and clinics assess how well cochlear
will be delivered from an auxiliary screen connected to a
implant recipients cope with the challenges of everyday
standard laptop computer. The clinician administering the
listening and to help clinicians counsel patients on the use
test will use the main laptop screen and laptop keyboard to
of their cochlear implants. Clinicians will have the option
present the test tokens and to score the child’s responses. The
of completing the questionnaire electronically together with
test’s adaptive algorithm will rapidly establish thresholds for
the cochlear implant users or having the cochlear implant
auditory and auditory-visual speech perception in classroom
patients complete the questionnaires themselves using a
noise. It is anticipated that the new test will be suitable for
laptop or personal computer.
use in the clinic and in the school environment. Pilot data
will be collected from normal-hearing children and children
with cochlear implants to validate the test procedures. In
addition, longitudinal data will be collected to map the
development of auditory and visual speech perception in
pediatric cochlear implant recipients.
Clinical Tools 189
Hearing with a Cochlear Implant:
The Everyday Listening Questionnaire I
Beate Krüger, M.Sc.
Martina Brendel, M.Sc.
Carolin Frohne-Büchner, Ph.D.*
Andreas Büchner, Ph.D.
Thomas Lenarz, Prof., M.D., Ph.D.
Medizinische Hochschule Hannover, Hannover, Germany
* also with Advanced Bionics, Europe
For more than 20 years, the performance of cochlear
implant users has improved continously as a result
of technical progress and with advances in clinical
practice (Büchner et al, 2005). Today, most adult
implant users experience open-set speech understanding, acceptable music appreciation, telephone
use, and speech understanding in noisy environments. To track the self-appraised development
of cochlear implant benefit, the first version of an
Everyday Listening Questionnaire I (ELQ-I) was
developed by Advanced Bionics to evaluate listening
benefit in CII Bionic Ear and HiRes 90K users.
Table 1. Percent of respondents indicating their ability to
understand the speech of familiar and
unfamiliar speakers over the telephone.*
Understanding Rating
Familiar
Speakers
Unfamiliar
Speakers
0
5.4
1
Very Poor—Impossible
2
Poor—Very Difficult
2.7
16.2
3
Adequate
8.1
48.7
4
Good
32.4
21.6
5
Very Good
56.8
5.4
NR Not Responding
0
2.7
* n = 37
190 Based on a questionnaire designed by the House
Ear Institute, the ELQ-I addresses medical history,
social background, motivation for implantation,
telephone use, music perception, the workplace,
and social environments. Subjects rate their performance and difficulties in listening on a five-point
rating scale (1 = no understanding possible, 5 = very
good understanding). The ELQ-I also evaluates the
use and benefit of implant accessories. Each accessory is rated on a five-point scale (A = always used,
E = never used).
A group of 37 adult subjects with a minimum of six
months of implant experience were enrolled in the
initial evaluation of the questionnaire. All had been
implanted either with the CII Bionic Ear or HiRes
90K device. Nine used the body-worn PSP processor
and 28 used a behind-the-ear (BTE) processor
(Auria or CII-BTE). The mean age was 50.0 years
(range: 24.6 to 75.8 years) and the mean duration of
deafness was 3.7 years (range: 1.4 to 5.9 years).
The ELQ-I required about one hour to complete.
Most subjects reported that improvement of speech
understanding for social activities was their major
motivation for seeking an implant. This response
was closely followed by the desire to communicate
on the telephone. The third important motivation
was hope for better communication in work-related
environments. Improvement in music appreciation
was the least important motivation for most of the
subjects.
As an example of some of the data acquired with the
ELQ-I, subjects’ evaluation of their ability to use the
telephone is summarized in Table 1. With familiar
speakers, 89.2% of the subjects rated conversation on
the telephone as “good” or “very good.” In contrast,
21.6% of subjects indicated that understanding
unfamiliar speakers was “poor” or even “impossible.”
Advanced Bionics® Auditory Research Bulletin 2007
Table 2 summarizes telephone accessory use for PSP
and BTE users. Independent of familiarity with
the person speaking on the telephone, 44.4% of the
PSP wearers used their standard earhooks whereas
55.6% used the Platinum telephone adapter. BTE
users were more varied in their choice of accessories,
but the majority preferred the T-Mic (now standard
with the Auria and Harmony sound processors).
Only 3.6% of BTE wearers occasionally used a telephone adapter. No respondent used the auxiliary
microphone with the telephone.
Social environments elicited a range of responses.
While the majority of cochlear implant users subjectively rated their understanding of news on TV
as “possible” or “very possible,” they rated understanding the plot of a movie as “difficult” or “very
difficult.” The majority of survey respondents rated
group conversations at parties as their most difficult
listening situations (35.1% “very difficult” and 29.7%
“impossible”)—presumably because of background
noise and multiple talkers. Notably, respondents
indicated that they rarely used accessories, even
though they reported difficulty hearing in most
everyday listening situations.
“. . .the ELQ-I proved valuable
in assessing everyday implant benefit
and in revealing the resistance of subjects
to using accessories.”
Table 2. Percent of respondents reporting use
of various assistive telephone accessories.*
PSP Users
BTE Users
Familiar
Speakers
Unfamiliar
Speakers
Familiar
Speakers
Unfamiliar
Speakers
Standard
Earhook
44.4
44.4
25.0
21.4
Telephone
Adapter
55.6
55.6
21.4
3.6
Telecoil
0
0
7.1
10.7
Amplified
Handset
0
0
25.0
17.9
T-Mic
0
0
67.9
53.6
* PSP n = 9. BTE n = 28. Note that T-Mic is only available to BTE users.
In summary, the ELQ-I proved valuable in assessing
everyday implant benefit and in revealing the resistance of subjects to using accessories. However, the
questionnaire required too much time to complete,
thereby precluding its clinical use. Based upon
this experience, a revised, shorter format Everyday
Listening Questionnaire (ELQ-II) was designed to
focus specifically on accessory use.(See also Krüger
et al, page 192 of this bulletin.)
Reference
Büchner A, Joseph G, Rost U, Strauss-Schier A, Lenarz T. Evaluation of speech
perception results in adult cochlear implant patients from 1984 till today:
Classification in performance groups. Poster presented at the 5th Asia-Pacific
Symposium on Cochlear Implants and Related Science, Hong Kong, China.
26-28 November, 2005.
Clinical Tools 191
Impact of Cochlear Implant Technology on Everyday
Listening: The Everyday Listening Questionnaire II
Beate Krüger, M.Sc.
Martina Brendel, M.Sc.
Carolin Frohne-Büchner, Ph.D.*
Andreas Büchner, Ph.D.
Thomas Lenarz, Prof., M.D., Ph.D.
Medizinische Hochschule Hannover, Hannover, Germany
* also with Advanced Bionics, Europe
“. . .the ELQ-II proved easier
for subjects to complete
than the previous questionnaire.”
One-on-One Situations
5
In this study, 60 adult implant users have completed
the new questionnaire—24 using body-worn processors (Clarion 1.2, S-Series, or PSP) and 36 using
behind-the-ear processors (Platinum BTE, CIIBTE, Auria, or Harmony). The mean age was 55.0
years (range: 25.4 to 78.2 years) and the mean duration of implant use was 4.3 years (range: 0 to 37.9
years).
Group Situations
Mean Rating
4
3
2
P
E
P
BW
BT
an
da
rd
on
on
St
St
an
da
rd
n
n
Re
so
Hi
Re
so
Hi
BW
BT
E
P
E
on
BW
BT
St
an
da
rd
on
BW
St
an
da
rd
so
n
Re
Hi
Hi
Re
so
n
BT
E
P
1
Figure 1. Average subjective ratings for understanding speech
at a party in one-on-one situations and in group situations.
(BTE = behind-the-ear processor; BWP = body-worn processor.
HiRes processing includes either standard HiRes or HiRes 120;
conventional processing includes CIS, PPS, or SAS.)
192 The Everyday Listening Questionnaire I (ELQ-I)
provided information about how cochlear implants
impact speech understanding, telephone use, music
perception, and communication in the workplace and
at social events. (Büchner et al, 2005; see also Krüger
et al, page 190 of this bulletin.) However, the ELQI did not probe the resistance of implant recipients
to using assistive listening accessories. Therefore, a
revised questionnaire (the ELQ-II) was developed
to address this issue and to include subjects with
older implant systems who may depend on assistive
listening devices. In addition, the number of items
was reduced and the rating scales were standardized
to make the questions easier and faster to answer.
Subjects rated the ability to understand speech in a
variety of communication situations on a five-point
scale (1 = impossible; 5 = very easy). All subjects
rated understanding in a one-on-one conversation
significantly better than in a group. While 35% of
the subjects had no problem understanding speech in
one-to-one conversations, only 2% reported that they
could understand speech easily in a group situation.
A majority of subjects (63%) were unable to understand conversations in a group at a party. However,
at a party, there was a trend for better understanding
in a group situation for users of standard HiRes or
HiRes with Fidelity 120 sound processing compared
to users of conventional strategies (Figure 1).
Advanced Bionics® Auditory Research Bulletin 2007
In summary, the ELQ-II proved easier for subjects
to complete than the previous questionnaire. Moreover, the administration time was reduced from one
hour to 20 minutes, thereby making the ELQ-II
practical for everyday clinical use. In general, the
study data indicate that assistive listening accessories
are used rarely, primarily because they are perceived
as impractical or as having no benefit.
HiRes 120 on BTE (n=33)
HiRes 120 on Body-Worn (n=6)
Standard Mode on BTE (n=4)
Standard Mode on Body-Worn (n=17)
100
Percent of Users
80
60
40
20
er
sw
An
No
W
i
Sy rele
st ss
em
A
M ux
icr ili
op ary
ho
ne
T-
M
ic
0
Ac Wit
ce ho
ss ut
or
ie
s
Users of body-worn processors rarely took advantage
of assistive accessories in group situations, while the
majority of BTE users were able to use the T-Mic to
hear better in groups (Figure 2). Neither the strategy
nor the processor type had any significant influence
on the use of most accessories with the exception
of the T-Mic, which was used commonly by BTE
users fit with HiRes or HiRes 120. Thirty-three
percent of the body-worn processor users and 44%
of the BTE users regarded most of the accessories
as impractical. Interestingly, most BTE users did
not perceive the T-Mic as an accessory, but rather
as a hassle-free option for everyday use. Twentynine percent of body-worn processor users and 28%
of the BTE users reported that they received little
benefit from accessories. Of particular concern was
that 25% of the body-worn processor users and 8%
of the BTE users reported that they were unaware
that the accessories even existed.
“. . .the study data indicate that
assistive listening accessories
are used rarely, primarily because
they are perceived as impractical
or as having no benefit.”
Figure 2. Percent of subjects using accessories for understanding speech in a group conversation. Subjects are sorted by
body-worn processor or BTE. (Note that the T-Mic is available
for BTE users only).
Reference
Büchner A, Joseph G, Rost U, Strauss-Schier A, Lenarz T. Evaluation of speech
perception results in adult cochlear implant patients from 1984 till today:
Classification in performance groups. Poster presented at the 5th Asia-Pacific
Symposium on Cochlear Implants and Related Science, Hong Kong, China.
26-28 November, 2005.
Clinical Tools 193
Using Tele-Medicine to Program
Cochlear Implants Remotely
Angel Ramos, M.D.1
Carina Rodriguez, B.Sc.1
Paz Martinez-Beneyto, M.D.2
Daniel Perez, M.D.1
Alexandre Gault, M.Sc.2
Patrick Boyle, M.Sc.2
1Hospital Insular Gran Canaria, Las Palmas, Spain
2Advanced Bionics, Europe
Local Equipment
Remote Equipment
Voice
Video
Screen
Control
Patient
Local Programmer
Remote Programmer
Figure 1. Hardware setup for remote programming. Connections between both stations provide voice, video, display
monitor, and remote-programmer controls as illustrated.
Certain geographic situations, such as long distances
between a cochlear implant center and its implanted
patients, call for innovative approaches for efficient delivery of clinical services and support. One
potential approach is to use “tele-medicine” whereby
implant center expert personnel communicate
with off-site clinical professionals via networked
computer communications through the Internet.
Such a communication system can improve both
the frequency and quality of patient care by allowing
the implant center to view and control an off-site
clinic’s computer during fitting and troubleshooting
sessions, thereby reducing the number of times a
patient must travel to the cochlear implant center.
This study explored the feasibility of remote
programming as a clinical service. Five adults
who had used HiRes 90K implants for one to
three months were recruited for this study. All
subjects were programmed with the SoundWave
fitting software and were assigned randomly to
receive follow-up clinical services either at an offsite clinic closer to home or at the implant center
directly. For subjects programmed at an off-site
clinic, the implant center’s programming computer
was connected to the off-site computer via a DSL
line. Web conferencing software provided video
and audio communications between the center’s
clinician and the off-site clinician and patient
(Figure 1). Performance and programming variables
were compared across subjects to assess the equivalence of remote versus one-to-one programming. In
addition, questionnaires were administered to the
subjects as well as to all participating clinicians.
Thus far, no significant differences have been found
in performance or programming variables among
patients receiving either one-to-one or remote
194 Advanced Bionics® Auditory Research Bulletin 2007
clinical services. Session times across patients were
similar in duration. Clinicians have reported that
hardware connections were stable and reliable
and that all remote programming sessions were
conducted easily and without incident. Specifically,
no critical system errors occurred such as uncontrolled increases in stimulus intensity or sudden
overstimulations. Moreover, responses to questionnaires indicated that remote programming was well
accepted by the patients.
“. . .remote programming
of cochlear implants
is a technically feasible approach
as well as a clinically safe and
cost-effective means of providing
clinical services to patients. . .”
With remote programming, a delay of approximately one-half second occurred between the time
an action was run at the implant center and the
time the action occurred at the off-site computer. A
similar delay occurred between the off-site computer
and the center computer. However, clinicians quickly
adapted to the delay and incorporated it into the
fitting process. Nevertheless, this delay might be a
potential drawback for measuring psychophysical
thresholds, although those measures are not usually
required to program the CII or HiRes 90K devices
through SoundWave. Additional tests showed that
higher speed Internet connections shorten the
time delay between sites, which may minimize the
required conditioning time for proper fitting.
In summary, this pilot study demonstrates that
remote programming of cochlear implants is a
technically feasible approach as well as a clinically
safe and cost-effective means of providing clinical
services to patients who live distant to their implant
centers. This study will be expanded to include additional European cochlear implant sites.
Clinical Tools 195
Contributing Authors
Advanced Bionics gratefully acknowledges the following authors whose contributions to this research
bulletin appear on the pages listed.
A
C
Akin, Istemihan 162
Aleksy, Wanda 188
Almario, Jorge E. 58
Ambrosch, Petra 159
Archer, Karen 91
Armstrong, Shelly 108
Arnold, Laure 53, 156, 160, 162, 164, 188
Aygener, Nurdan 146
Cadieux, Jamie H. 137
Caner, Gül 164
Canning, David 189
Carlyon, Robert P. 81
Carroll, Jeff 32
Chang, Yu-Tuan 51
Chen, Xueqing 64
Chénier, Josée 108
Chinnici, Jill 135
Chun, Young-Myoung 48
Costa, Orozimbo 60
Cowdrey, Lisa 102, 104
Crozat-Teissier, Natacha 53
B
Baker, Christa 22
Ballantyne, Deborah 144
Ballard, Jessica 168
Basta, Dietmar 25
Battmer, Rolf-Dieter 30, 170
Belgin, Erol 150
Benitez, Raul 176
Berenstein, Carlo 88
Bero, Eva M. 172
Bestel, Julie 46
Bevan, Jane 118
Bevilacqua, Maria Cecilia 60
Bierer, Julie Arenberg 75
Black, Jeanne 132
Blevins, Nikolas H. 32
Borden, Lamar 47
Bosco, Ersilia 56, 62
Bowditch, Stephen 135
Boyle, Patrick 25, 53, 91, 122, 146, 156,
160, 188, 189, 194
Bracke, Peter 82
Brendel, Martina 84, 86, 96, 98, 110, 170,
190, 192
Briaire, Jeroen J. 72, 152, 182
Brown, Carolyn J. 135
Brunsden, Barry 19
Buchman, Craig 135
Büchner, Andreas 30, 84, 86, 96, 98, 100,
110, 122, 170, 190, 192
Budak, Bilgehan 150
196 D
D’Agosta, Luciana 56, 62
D’Elia, Chiara 29, 56, 62, 144
Daft, Mark 54
Dawson, Patricia 47
Deeks, John M. 81
Dimitrijevic, Andrew 32
Dincer, Hilal 146, 149, 150, 162
Dingemanse, J. Gertjan 72
Donaldson, Gail 47
Döring, Wolfgang H. 159
Dorman, Michael F. 103
Downing, Mark 83
Drennan, Ward R. 145
Dujardin, Herbert 159
Dykmans, Philippe 175, 189
E
Edler, Bernd 84, 86
Edmonson, Danielle 130
Eisen, Marc D. 78
Elsholz, Alexander 159
Emadi, Gulam 81
Erfurt, Peter 18
Ernst, Arne 25
Eskridge, Hannah 135
Advanced Bionics® Auditory Research Bulletin 2007
F
I
Faulkner, Andrew 76
Fernandez-Vega, Susana 29
Filipo, Roberto 29, 56, 144
Finley, Charles C. 19
Firszt, Jill B. 83, 137, 187
Fiscus, Amy 130
Fitzpatrick, Douglas C. 26
Fitzpatrick, Elizabeth 108
Frachet, Bruno 46, 160
Franck, Kevin H. 78
Frederigue, Natália Barreto 60
Fridman, Gene Y. 91
Frijns, Johan H. M. 72, 106, 152, 182
Frohne-Büchner, Carolin 84, 86, 96, 98, 110,
190, 192
Fuller, Shawnda 130
Incesulu, Armagan 146, 149
G
Gault, Alexandre 110, 118, 180, 194
Goetze, Romy 25
Goldberg, Donald M. 138
Goldring, Jenny 180, 189
Green, Tim 76
Grieco, Tina 133
Groeschel, Moritz 25
Gulpen, Patricia M.H. 152
Gültekin, Gürol 164
H
Hallam, Angie 130
Hammes, Dianne 130
Han, Demin 64
Haumann, Sabine 100
Hawker, Kelvin 54
Hedley-Williams, Andrea 116
Holden, Laura K. 184, 187
Holden, Timothy A. 19
Horler, Wendy 118
Hughes, Michelle L. 70
Hullar, Timothy E. 19
J
Janke, Oliver 25
Jenkins, Marsha 118
Jensen, Jamie 74
Johnson, Susan 54
K
Kasturi, Kalyan 79
Kasulis, Heather 166
Kecik, Cem 149
Kettel, Jerrica L. 137
Khajehnouri, Yassaman 18, 30, 33, 100, 122, 170
Kimura, Márcia Yuri Tsmura 60
Kirkim, Gunay 146
Klop, W. Martin C. 182
Koch, Dawn Burton 83
Kohl, Elizabeth 118
Kohler, Matthew 78
Kong, Ying 64
Kopelovich, Jonathan 78
Krüger, Beate 96, 98, 190, 192
Kuran, Gokhan 162
L
Larky, Jannine 32
Lee, Daniel J. 172
Lee, Jee-Yeon 48
Lee, Kathy Y. S. 50, 51
Lee, Seung-Chul 48
Lenarz, Thomas 18, 30, 33, 84, 86, 96, 98, 100,
110, 122, 170, 190, 192
Lesinski-Schiedat, Anke 30
Lewis, Kristen 102, 104
Limb, Charles J. vii
Lisbona, Kristin 102, 104
Litovsky, Ruth 133
Litvak, Leonid 32, 78, 81, 82, 83, 91
Liu, Bo 64
—continued on next page—
Contributing Authors 197
List of Contributing Authors—continued from previous page
L—continued
P
Liu, Haihong 64
Loiselle, Louise H. 103
Loizou, Philipos 79
Lorenzi, Christian 46
Luk, Betty P. K. 50, 51
Park, Hong-Joon 48
Peeters, Stefaan 20, 80
Perez, Daniel 194
Peters, B. Robert 132
Pinto, Francesca 188
Poissant, Sarah F. 172
Postnov, Andrei 20
Prieto, Jose Alberto 58
M
Machery, Olivier 81
Mancini, Patrizia 56, 62, 144
Martinez-Beneyto, Paz 194
Mens, Lucas 88
Meyer, Bernard 46
Moore, Brian C. J. 122
Moret, Adriane Lima Mortari 60
Morse, Peter F. 91
Morse, Robert P. 91
Morsnowski, Andre 159
Mourik, Ruud 152
Mulder, Jef 88
Müller-Deile, Joachim 159
Murotonen, Salomo 82
Mutlu, Murad 162
N
Nascimento, Leandra Tabanez 60
Neely, J. Gail 19
Niparko, John 135
Noël-Petroff, Nathalie 53
Nogueira, Waldo 84, 86
Novak, Michael A. 130
Nuñez, Maria Piedad 58
Nunn, Terry B. 91
O
O’Neil, Jahn N. 22
Offeciers, Erwin 20, 80, 175
Olgun, Levent 146, 164
Onder, Sevginar 146
Ormezzano, Yves 160
Osberger, Mary Joe 83
198 Q
Quaranta, Antonio 29
Quaranta, Nicola 29
Quick, Andrew 168
R
Rafter, Diane 118
Ramos, Angel 194
Reeder, Ruth M. 137, 187
Rivas, Adriana 58
Rivas, Jose Antonio vi, 58
Roberto, Filipo 62
Rodriguez, Carina 194
Rosen, Stuart 76
Roth, Elizabeth L. 52
Rotz, Lee Ann 130
Rubinstein, Jay 145
Runge-Samuelson, Christina 74
Ryugo, David K. 22
S
Saffran, Jenny 133
Saoji, Aniket 91
Saxon, Eugene A. 19
Schafer, Erin C. 112, 114, 120
Schramm, David 108
Schwab, Burkhard 170
Séguin, Christiane 108
Sennaroglu, Gonca 150
Skinner, Margaret W. 19, 187
Smoorenburg, Guido ix
Soede, Wim 106
Spahr, Anthony J. 81, 91, 103
Advanced Bionics® Auditory Research Bulletin 2007
S—continued
X
Starr, Arnold 32
Sterkers, Olivier 46
Stiggelbout, Anne M. 182
Stille, Lisa J. 70
Stone, Michael A. 122
Stöver, Timo 18, 170
Summerfield, A. Quentin 136
Xu, Li 64
T
Tang, Qing 32, 176
Teagle, Holly 135
Tharpe, Anne Marie 116
Thomas, Jean 130
Tong, Michael C. F. viii, 50, 51
Traisci, Gabriella 56, 62
Truy, Eric 46
Twomey, Tracey 54
Y
Yang, Hui-Mei 51
Yeagle, Jennifer 135
Yucel, Esra 146, 150
Z
Zarowski, Andrzej 20, 80, 175
Zdanski, Carlton 135
Zeng, Fan-Gang 32, 176
Zheng, Yan 64
Zhou, Ning 64
U
Uchanski, Rosalie M. 184
V
Vanpoucke, Filiep 20, 80, 82, 88, 175, 189
Van Den Abbeele, Thierry 53
Van der Beek, F. B. 106
van Hasselt, Charles Andrew 50
van Wieringen, Astrid 81
Veekmans, Kim 54
Vermeiren, Annelies 80, 175
Vickers, Deborah 44, 118, 146, 150
Voie, Arne H. 19
W
Wackym, P. Ashley 74
Westhofen, Martin 159
Whiting, Bruce R. 19
Willis, Mary 130
Wilson, Pippa 188
Wirth, Julie 130
Wolfe, Jace 112, 114, 120, 166, 168
Wong, Terence K. C. 50, 51
Woulters, Jan 81
Wu, Jiunn-Liang 51
Contributing Authors 199
Participating Research Centers
Advanced Bionics gratefully acknowledges the contributions of the following institutions to this Research
Bulletin. These research centers and clinics provided a project summary, participated in a multicenter
study, are involved in a new study initiative, or participate in ongoing projects with Advanced Bionics.
AFRICA
EUROPE
MOROCCO
BELGIUM
Clinique Rachidi
Catholic University of Leuven
Casablanca
Leuven
Cliniques Universitaires Saint-Luc
Brussels
ASIA
Medical Center Sint Augustinus
Wilrijk
JAPAN
The Eargroup
Osaka University
Antwerp
Osaka
University of Ghent
Toranomon Hospital
Ghent
Tokyo
University of Antwerp
Antwerp
PEOPLE’S REPUBLIC OF CHINA
The Chinese University of Hong Kong
FRANCE
Hong Kong Special Administrative District
Hôpital Avicenne
Tongren Hospital
Bobigny
Beijing
Hôpital Beaujon
Clichy
TAIWAN
Cheng Hsin Rehabilitation Medical Center
Taipei
National Cheng Kung University Hospital
Tainan
Hôpital Charles Nicolle
Rouen
Hôpital Edouard Herriot
Lyon
Hôpital Gui de Chauliac
Montpellier
KOREA
Hôpital Pellegrin
Hallyn University
Hôpital Purpan
Bordeaux
Chuncheon
Toulouse
Samsung Medical Center
Seoul
Paris
Soree Ear Clinic/Soree Hearing Center
Seoul
Hôpital Robert Debré
Hôpital Roger Salengro
Lille
Hôpital Saint-Antoine
Paris
Institut Francilien d’Implantation Cochléaire
Paris
Institut Saint-Pierre
Palavas-les-Flots
200 Advanced Bionics® Auditory Research Bulletin 2007
FRANCE—continued
NETHERLANDS
L’Ecole Normale Supérieure
Academisch Ziekenhuis
Paris
Groningen
Université Claude Bernard
Erasmus MC
Lyon
Rotterdam
Leiden University Medical Center
GERMANY
Cochlear Implant Centrum Rhein-Main
Friedberg
Humboldt University
Leiden
Radboud University Nijmegen Medical Center
Nijmegen
Berlin
SPAIN
Klinikum d. J W Goethe Universität
Clinica Universitaria de Navarra
Frankfurt
Pamplona
Klinikum d. Albert-Ludwigs-Universität
Hospital Insular de Gran Canaria
Freiburg
Las Palmas
Kurzel Rehabilitation Centre
Technical University of Catalonia
Aachen
University of Hannover
Hannover
Medizinische Hochschule Hannover
Hannover
Barcelona
SWEDEN
Karolinska University Hospital Huddinge
Stockholm
Unfallkrankenhaus Charité Medical School
Berlin
Universitätsklinikum Aachen
Aachen
Universitätsklinikum Schleswig Holstein
Kiel
TURKEY
Dokuz Eylul University
Izmir
Erisci Elektroniks
Ankara
ITALY
Hacettepe University Hospital
Azienda Ospedaleria di Padova
Osman Gazi University
Padova
Cattedra di Audiologia
Ferrara
University of Bari
Bari
University of Rome La Sapienza
Rome
Ankara
Eskisehir
SB Ankara Diskapi Hospital
Ankara
SB Ankara Numune Hospital
Ankara
SB Izmir Bozyaka Hospital
Izmir
SB Izmir Tepecik Hospital
Izmir
—continued on next page—
Participating Research Centers 201
Participating Research Centers—continued from previous page
EUROPE—continued
MIDDLE EAST
UNITED KINGDOM
INDIA
Aston University
Jaslok Hospital and DeSa’s Hospital
Cochlear Implant Center
Birmingham
Birmingham Adult Cochlear Implant Programme
Mumbai
Birmingham
Birmingham Children’s Hospital
Birmingham
Emmeline Cochlear Implant Programme
Cambridge
Guys’ and St Thomas’ NHS Foundation Trust
London
Keele University
Keele
Kilmarnock Cochlear Implant Programme
Kilmarnock
Manchester Royal Infirmary
Manchester
ISRAEL
Bnai Zion Medical Center
Haifa
Chaim Sheba Medical Center
Tel Hashomer
Hadassah University Hospital
Jerusalem
Schneider Children’s Medical Centre
Petah Tikva
Tel Aviv University
Tel Hashomer
Mary Hare School
Newbury, Berkshire
Nottingham Cochlear Implant Programme
Nottingham
Royal National Throat, Nose, & Ear Hospital
London
The Open University in the South
Oxford
South of England Cochlear Implant Center
Southampton
University College London
London
University of Cambridge
Cambridge
University of York
York
Warwick University
Warwick
202 Advanced Bionics® Auditory Research Bulletin 2007
NORTH AMERICA
CANADA
UNITED STATES—continued
Alberta
California
Glenrose Rehabilitation Hospital
California Ear Institute/Let Them Hear
Foundation
Edmonton
Palo Alto
Ontario
House Ear Clinic/House Ear Institute
Children’s Hospital of Eastern Ontario
Los Angeles
Ottawa
Kaiser Permanente
Ottawa Hospital (Civic Campus)
Sacramento
Ottawa
Stanford University
Sunnybrook & Women’s College
Health Sciences Center
Palo Alto
Toronto
Irvine
University of Ottawa
Ottawa
Québec
L’Hôtel-Dieu de Québec
Québec City
UNITED STATES
Alabama
Children’s Hospital of Alabama
Birmingham
Arizona
Arizona State University
Tempe
Mayo Clinic
Scottsdale
Phoenix Children’s Hospital
Phoenix
Arkansas
Arkansas Children’s Hospital
Little Rock
University of California
University of California
Los Angeles
University of California
San Francisco
Connecticut
Yale University
New Haven
Florida
Tampa Bay Hearing & Balance Center
Tampa
University of South Florida
Tampa
University of Miami
Miami
Georgia
Children’s Healthcare of Atlanta
Atlanta
Illinois
Carle Clinic Association/Carle Foundation
Hospital
Urbana
The Children’s Memorial Hospital
Chicago
Northwestern University
Chicago
—continued on next page—
Participating Research Centers 203
Participating Research Centers—continued from previous page
NORTH AMERICA—continued
UNITED STATES—continued
Indiana
Nebraska
Indiana University
Boys Town National Research Hospital
Indianapolis
Omaha
Iowa
New York
University of Iowa
Beth Israel-New York Eye and Ear
Cochlear Implant Center
Iowa City
Maryland
New York
Columbia University Medical Center
Johns Hopkins University
New York
Baltimore
New York University Medical Center
Massachusetts
New York
Massachusetts Eye and Ear Infirmary
North Carolina
Boston
University of North Carolina
University of Massachusetts
Chapel Hill
Amherst
University of Massachusetts Memorial
Medical Center
Ohio
Cleveland Clinic Foundation
Worcester
Cleveland
Michigan
Oklahoma
Spectrum Health
Hearts for Hearing Foundation
Grand Rapids
Oklahoma City
University of Michigan
Ann Arbor
Minnesota
Mayo Clinic
Oregon
Oregon Health Sciences University
Portland
Rochester
Pennsylvania
University of Minnesota
Children’s Hospital of Philadelphia
Minneapolis
Philadelphia
Missouri
South Carolina
Midwest Ear Institute
Medical University of South Carolina
Kansas City
St. Louis Children’s Hospital
Charleston
St. Louis
Tennessee
Washington University
Vanderbilt University
St. Louis
204 Nashville
Advanced Bionics® Auditory Research Bulletin 2007
SOUTH AMERICA
BRAZIL
Texas
University of Sao Paulo
Bauro
Dallas Otolaryngology
Dallas
Ear Medical Group
San Antonio
Houston Ear Research Foundation
Houston
University of North Texas
Denton
University of Texas at Southwestern
Medical Center
COLOMBIA
Clínica Rivas
Bogotá
Clínica San Rafael
Bogotá
Hospital San Ignacio
Bogotá
Dallas
Virginia
University of Virginia
Charlottesville
Washington
University of Washington
Seattle
Wisconsin
Medical College of Wisconsin
Milwaukee
University of Wisconsin
Madison
Participating Research Centers 205
Research Staff Worldwide
NORTH AMERICA
Clinical Research
Research and Development
Mary Joe Osberger, Ph.D.
Mike Faltys
Director, Global Clinical Research
Vice President, Research and Development
Drina Aldana
Gulam Emadi, Ph.D.
Clinical Regulatory & Contracts Associate
Senior Research Scientist
Ann Kalberer, M.S.
Tracey Kruger, M.S.
Clinical Research Specialist
Director, Clinical Engineering and Development
Dawn Burton Koch, Ph.D.
Abhijit Kulkarni, Ph.D.
Senior Clinical Research Scientist
Director, Systems Development
Elizabeth MacDonald, M.Ed.
Leonid Litvak, Ph.D.
Clinical Research Scientist
Science Fellow
Ann Masuda, M.S.
Salomo Murotonen, B.S.
Clinical Project Manager
Senior Engineer
Lauri Nelson, Ph.D.
Frank Nigro
Clinical Research Scientist
Research Associate
Andrew Quick, M.S.
Edward Overstreet, Ph.D.
Manager, Auditory Clinical Research
Director, Investigational Technologies
Lisa Tracey, M.A.
Aniket Saoji, Ph.D.
Clinical Research Associate
Research Audiologist
Sue Zimmerman-Phillips, M.S.
Field Manager, Auditory Clinical Studies
206 Advanced Bionics® Auditory Research Bulletin 2007
EUROPE
Clinical Research
ASIA & LATIN AMERICA
Patrick Boyle, M.Sc.
Margaret Cheng, M.Sc.
Director, Clinical Research
Laure Arnold, M.Sc.
Manager, Clinical Studies.
Carolin Frohne-Büchner, Ph.D.
Clinical Research Scientist
Manager, Clinical Research and Education, Asia-Pacific
Jae Lee, Ph.D.
Principal Clinical Engineer, Asia-Pacific
Franco Portillo, Ph.D.
Manager, Latin American Affairs
Jenny Goldring, M.Sc.
Studies Project Manager
Lu-Ming Joffo, M.Sc.
Clinical Research Scientist
Yassaman Khajehnouri, M.Sc.
Clinical Research Scientist
Barbara Kienast, Ing.
Clinical Research Scientist
Fiona Robinson
Clinical Research Logistics Manager
European Research Center
Antwerp, Belgium
Filiep Vanpoucke, Ph.D.
Head, European Research Center
Philippe Dykmans, M. Eng.
Senior Software Engineer
Tim Nauwelaers, M. Eng.
Electrode Engineer
Stefaan Peeters, Prof., Dr. Ir.
Consultant
Research Staff Worldwide 207
Cover/qxp
1/22/08
10:28 AM
Page 1
Auditory Research Bulletin
DEC07 2-091520
© 2007 by Advanced Bionics Corporation or its affiliates. All rights reserved. Printed in U.S.A.
2007 BIENNIAL EDITION
www.BionicEar.com
Auditory
Research Bulletin
2007 BIENNIAL EDITION
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement