Neurophysiology of Cochlear Implant Users I

Neurophysiology of Cochlear Implant Users I
Neurophysiology of Cochlear Implant Users I: Effects
of Stimulus Current Level and Electrode Site on the
Electrical ABR, MLR, and N1-P2 Response
Jill B. Firszt, Ron D. Chambers, Nina Kraus, and Ruth M. Reeder
N1-P2 of the ELAR were characterized as functions
of current level and electrode site. Data from this
study may serve as a normative reference for expected latency, amplitude and threshold values for
the recording of electrically evoked auditory brainstem and cortical potentials. Responses recorded
from cochlear implant users show many similar
patterns, yet important distinctions, compared with
auditory potentials elicited with acoustic signals.
Objective: As the need for objective measures with
cochlear implant users increases, it is critical to
understand how electrical potentials behave when
stimulus parameters are systematically varied. The
purpose of this study was to record and evaluate the
effects of implanted electrode site and stimulus
current level on latency, amplitude, and threshold
measures of electrically evoked auditory potentials,
representing brainstem and cortical levels of the
auditory system.
(Ear & Hearing 2002;23;502–515)
Design: The electrical auditory brainstem response
(EABR), electrical auditory middle latency response (EAMLR), and the electrical late auditory
response (ELAR) were recorded from the same experimental subjects, 11 adult Clarion cochlear implant users. The Waves II, III, and V of the EABR,
the Na-Pa complex of the EAMLR and the N1-P2
complex of the ELAR were investigated relative to
electrode site (along the intra-cochlear electrode
array) and stimulus current level. Evoked potential
measures were examined for statistical significance
using analysis of variance (ANOVA) for repeated
measures.
Electrically evoked auditory potentials have been
recorded through a variety of implanted electrode
arrays. These potentials include the electrical auditory brainstem response (EABR), the electrical auditory middle latency response (EAMLR) and electrical late auditory responses (ELAR). The earliest
studies used experimental devices in deaf adult
subjects (Gardi, 1985; Starr & Brackmann, 1979)
and were fraught with technical recording difficulties due to increased stimulus artifact produced by
electrical stimulation. Later studies included FDA
approved and clinical-trial devices such as the 3M
single-channel device (Miyamoto, 1986), the Ineraid
system (Abbas & Brown, 1988, 1991), the Nucleus
device (Abbas & Brown, 1991; Brown, Abbas,
Fryauf-Bertschy, Kelsay, & Gantz, 1994; Hodges,
Ruth, Lambert, & Balkany, 1994; Mason, Sheppard,
Garnham, Lutman, O’Donoghue, & Gibbin, 1993;
Shallop, Beiter, Goin, & Mischke, 1990; Shallop,
Van Dyke, Goin, & Mischke, 1991), the Clarion
implant (Brown, Hughes, Lopez, & Abbas, 1999;
Firszt, Rotz, Chambers, & Novak, 1999) and the
Med-El system (Firszt, Gaggl, Wackym, & Reeder,
Reference Note 1).
Results of EABR measures have not been reported consistently in published studies. Thresholds, amplitudes and waveform morphologies have
differed across and within subjects for different
electrodes (Gardi, 1985; Hodges, et al., 1994; Shallop
et al., 1990). Some animal studies show a relation
between amplitude growth functions and surviving
ganglion cells (Simmons & Smith, 1983; Walsh &
Leake-Jones, 1982), whereas other investigations
Results: For the EABR, Wave V latency was significantly longer for the basal electrode (7) compared
with the mid (4) and apical (1) electrodes. For the
EAMLR and ELAR, there were no significant differences in latency by electrode site. For all subjects
and each of the evoked potentials, the apical electrodes tended to have the largest amplitude and the
basal electrodes the smallest amplitude, although
amplitude differences did not reach statistical significance. In general, decreases in stimulus current
level resulted in statistically significant decreases
in the amplitude of Wave V, Na-Pa and N1-P2. The
evoked potential thresholds for Wave V, Na-Pa, and
N1-P2 were significantly higher for the basal Electrode 7 than for Electrodes 4 and 1.
Conclusions: Electrophysiologic responses of Waves
II, III, and V of the EABR, Na-Pa of the EAMLR, and
Department of Otolaryngology and Communication Sciences
(J.B.F., R.M.R.), Medical College of Wisconsin, Milwaukee, Wisconsin; Department of Speech and Hearing Science (R.D.C.),
University of Illinois, Champaign-Urbana, Illinois; and Departments of Communication Sciences; Neurobiology, Physiology;
Otolaryngology (N.K.), Northwestern University, Evanston,
Illinois.
DOI: 10.1097/01.AUD.0000042153.40602.54
0196/0202/02/2306-0502/0 • Ear & Hearing • Copyright © 2002 by Lippincott Williams & Wilkins • Printed in the U.S.A.
502
EAR & HEARING, VOL. 23 NO. 6
have not confirmed these findings (van den Honert
& Stypulkowski, 1986).
The EAMLR is an attractive alternative to
shorter latency potentials. It has a longer latency
(positive peak at 27 to 37 msec), making it less likely
to be contaminated by the stimulus artifact that
occurs early in the response. Recordings in animals
and humans indicate that it is reliable (Burton,
Miller, & Kileny, 1989; Gardi, 1985; Groenen, Snik
& van den Broek, 1997; Kileny, Kemink, & Miller,
1989; Miyamoto, 1986) and may correlate with neural survival (Jyung, Miller, & Cannon, 1989). One
study of the electrical and acoustic MLR within the
same subject suggests that both responses are activated by the same neural generators of the central
system (Kileny et al., 1989). In general, there are
few published studies of the EAMLR in cochlear
implant users.
Reports of late auditory potentials evoked with
electrical stimulation have included recordings of
the N1, P2, P300, and mismatch negativity (MMN)
response elicited with pulsed tones (Oviatt & Kileny,
1991), stimulated electrode pairs (Makhdoum,
Groenen, Snik, & van den Broek, 1997; Ponton &
Don, 1995), and speech (Kaga, Kodera, Hirota, &
Tsuzuka, 1991; Kraus et al., 1993; Micco et al.,
1995). Results suggest that cortical responses provide a mechanism for understanding how electrical
stimuli are registered by the central auditory
system.
The need for objective measures with cochlear
implant users has increased due to unexplained
variation in patient performance, implantation of
children at younger ages, and an increase in programming options. With this in mind, it is critical to
understand how electrical potentials behave when
stimulus parameters are systematically varied. It is
often difficult to compare published findings of
early, middle, and late-latency potentials across
studies due to differences in electrode arrays (longitudinal, radial), stimulation mode (bipolar, monopolar), time and location of recordings (operating room,
clinic, intra-operative, postoperative), subjects (pediatric, adult) and stimulus and recording parameters. There are no published studies that characterize the latency, amplitude and threshold for EABR,
EAMLR, and ELAR recorded across the electrical
dynamic range from the same experimental cochlear
implant subjects.
The goals of this study were 1) to record Wave V
of the ABR, the Na-Pa complex of the MLR and the
cortical N1-P2 complex from intracochlear electrodes, and 2) to evaluate the effects of implanted
electrode site and stimulus level on electrically
evoked auditory potentials that represent the brainstem and cortical levels of the auditory system from
503
the same subject sample. Specifically, the EABR,
EAMLR, and ELAR were recorded for individual
subjects from three electrodes that represent basal,
mid, and apical positions along the Clarion electrode
array. Latency, amplitude, general morphology, and
threshold were analyzed across stimulus level and
electrode site within subjects for each evoked potential. In a companion paper in this issue, the analysis
of evoked potential measures in relation to speech
perception performance for this subject sample is
presented (Firszt, Chambers, & Kraus, 2002).
METHODS
AND
PROCEDURES
Subjects
Eleven adults (five women and six men) who
received the Clarion 1.2 radial electrode array participated in the study. All subjects had full electrode
insertions, used monopolar stimulation of the medial electrode contacts, and used the continuous
interleaved sampler speech-processing strategy. At
the time of study, the subjects ranged in age from 29
to 75 yr, with a mean of 56 yr. At the time of cochlear
implantation, the subjects ranged in age from 24 to
70 yr, with a mean of 53 yr. Subjects had used their
cochlear implants for at least 3 mo, with a maximum
length of use of 5 yr and a mean across subjects of
2.7 yr. Subjects were full-time users of their cochlear
implants with a range of daily wear between 10 to 15
hr. Table 1 provides background information for
each subject.
General Procedures
The EABR, EAMLR, and ELAR were recorded on
Electrode 1 (apical), Electrode 4 (mid), and Electrode
7 (basal) within the cochlea. Before recording these
potentials, behavioral data were collected that defined the subject’s electrical threshold and dynamic
range (number of electrical steps between threshold
and upper limit of comfortable loudness) on each
tested electrode. Because subjects were evaluated
while awake in the clinic, it was important that
stimuli were comfortably loud. Electrical stimulus
levels within the subject’s behavioral dynamic range
(BDR) were presented for each electrode and response type (e.g., EABR, EAMLR, ELAR).
Behavioral Measures
Stimuli, Equipment, and Procedures • Stimuli
for behavioral measures were biphasic current pulse
trains generated by a computer-controlled interface
unit and SCLIN for Windows software version 1.2.
Pulses were 75 ␮sec in duration and negative-leading in polarity. Stimulus amplitude was expressed
in clinical units (CU) relative to microamperes (␮A)
504
EAR & HEARING / DECEMBER 2002
TABLE 1. Demographic information on subjects.
Subject
Ear
AAO/HL-Degree
AAO/PHL
AAI
LOU-CI
AAT
1
R
L
R
L
R
L
R
L
R
L
R
L
R
L
R
L
R
L
R
L
R
L
32-Mild-Moderate
32-Mild-Moderate
38-Moderate
5-Severe/Profound
17-Mild
18-Mild
18-Moderate
18-Moderate
20-Mild
20-Severe
61-Severe
18-Severe
30-Moderate
30-Moderate
3-Severe
3-Mild
41 Moderate
48-Moderate
46-Mild
48-Mild
46-Moderate
7-Profound
50
45
47
5
44
38
40
40
64
64
61
61
31
31
10
12
58
64
66
66
47
7
53
2-2
55
Inherited
Progressive
52
3-5
55
RE possible Meniere’s; LE congenital unknown
Progressive
46
3-0
49
Unknown
Progressive
64
4-0
68
Possible cochlear otosclerosis
Progressive
64
1-0
65
Unknown
Progressive
67
3-0
70
RE unknown; LE chronic suppurative otitis media
Sudden
32
1-6
33
Autoimmune inner ear disease
Sudden
24
5-0
29
Congenital hereditary (twin sister w/ HL)
Progressive
70
5-0
75
Meniere’s disease
Progressive
66
0-7
67
Meniere’s disease
Progressive
47
0-3
47
RE hemangioma of internal auditory artery; LE mumps
Sudden
2
3
4
5
6
7
8
9
10
11
Etiology
Onset
Ear indicated in bold ⫽ implanted ear. AAO/HL-Degree ⫽ age at onset (years) and degree of hearing loss; AAO/PHL ⫽ age at onset (years) of profound hearing loss; AAI ⫽ age at implant;
LOU-CI ⫽ length of use (years-months) with cochlear implant; AAT ⫽ age at test.
and was varied between 3 and 3000 CU.* Procedures for the measures of behavioral threshold (BT),
most comfortable loudness level (MCL), and upper
limit of comfort level (ULCL) are described in detail
in the companion paper in this issue.
Electrophysiologic Measures
Stimuli, Equipment, and Procedures • Stimuli
used to elicit the electrical evoked potentials were
identical to those used to obtain BT, MCL, and
ULCL and were within the subject’s BDR for each
electrode. Electrically evoked potentials were recorded at levels that corresponded to a) 100%, 75%,
50%, and 25% of the BDR, b) the BT, and c) the
evoked potential threshold. As a final control,
EABR, EAMLR, and ELAR responses were recorded
at the lowest amplitude input (3 CU) generated by
*Electrical charge is the product of current amplitude and pulse
duration, and interacts with electrode impedance. For the Clarion
device used in this study, the charge is expressed in CUs. For this
subject sample and tested electrodes, the range of impedance
values was narrow (7–28 k⍀, mean ⫽ 18.6 k⍀). Actual electrical
charge was calculated for each current level using a function that
incorporated electrode number, measured impedance, and current for one internal cochlear stimulator. The estimated actual
current values (electrical charge) using this correction function
resulted in minimal current differences due to the low impedance
values for these subjects. The analyses described in our study do
not depend on the absolute current level, but rather proportions
within the subject’s dynamic range. It should be noted, however,
that CUs may not be directly comparable across subjects and
electrodes.
the stimulus software to evaluate response identification against the subject’s baseline EEG activity.
Responses were recorded with a Nicolet Compact
4 (C4) electrodiagnostic system externally triggered
by the stimulus output of the SCLIN for Windows
software and the interface unit. The interface unit
connected to a pulse stretcher that extended the
duration of the pulse (200 ␮sec) sent from the
interface unit to trigger the Nicolet C4 averaging
computer. The interface unit also connected to a
stock speech processor and the subject’s headpiece,
transmitting the stimulus signal across the skin to
the implanted device. The amplifier connected to
electrodes that were positioned through a radio
frequency filter with a low-pass cutoff of 32.125 kHz
preceding the C4 amplifier.
The electrical evoked potentials were recorded
using standard gold cup surface electrodes placed on
the forehead (⫹), nape of the neck (⫺), and contralateral earlobe (common). Recording of electrical
activity included two or three replications of 1000
sweeps (EABR), 500 sweeps (EAMLR), and 300
sweeps (ELAR) at each stimulus level with a time
window of 10 msec (EABR), 50 msec (EAMLR), and
300 msec (ELAR) for each stimulus condition. For
the ELAR, eye movements were monitored using
electrodes located on the superior and lateral canthus of one eye (Kraus et al., 1993). Artifact rejection
eliminated trials that included eye movement and
interfered with the recording of the response. Responses were amplified 100,000 times. Frequency
EAR & HEARING, VOL. 23 NO. 6
cutoffs of 100 and 3000 Hz (EABR), 5 and 500 Hz
(EAMLR), and 1 and 100 Hz (ELAR) were used
(Butterworth, 12 dB/octave). For each evoked potential type, subjects were tested in a quiet exam room
in a reclining chair. For EABR testing, subjects were
instructed to relax and encouraged to sleep during
the recording session. For the EAMLR and ELAR
testing, subjects were instructed to remain awake
and watch a captioned videotape.
Identification and Measures of Waveforms
The waveforms for early, middle and late responses were identified according to criteria that
were based on animal and human research with
acoustic and electrical stimulation. For the EABR, it
was expected that 1) a series of positive peaks would
occur between approximately 1 to 4 msec; 2) the
latency of Wave V would occur at approximately 3.5
to 4.0 msec at higher stimulus current levels and
increase with decreases in stimulus level; and 3)
Wave I would not be visible because it would have a
latency of approximately 0.75 msec (as seen in
recordings of the electrically evoked action potential, Abbas et al., 1998; Franck & Norton, 2001),
and would be embedded in stimulus artifact (Abbas & Brown, 1988, 1991; Picton, Hillyard,
Krausz, & Galambos, 1974; van den Honert &
Stypulkowski, 1986). For the EAMLR, a negative
trough, Na, at approximately 15 to 18 msec, would
be followed by a positive peak, Pa, at approximately 25 to 30 msec from the onset of the stimulus at suprathreshold levels (Jyung et al., 1989;
Kileny & Kemink, 1987; Özdamar & Kraus, 1983;
Picton et al., 1974). Electrical N1-P2 cortical responses would consist of a negative trough, N1, at
approximately 80 to 110 msec, followed by a positive peak, P2, at approximately 180 to 210 msec
(Kraus et al., 1993; Näätänen & Picton, 1987;
Picton et al., 1974; Ponton & Don, 1995). For all
evoked potentials, each waveform was compared
with that generated with a minimum current level
during the control run.
The latencies were measures (in msec) at the
midpoint for the peak (Wave V, Pa, P2), or the
midpoint of the trough (Na, N1). The evoked potential threshold was defined as the lowest current
level that a repeatable response could be visually
detected. Amplitudes were calculated based on the
difference (in ␮V) between the positive peak and the
following trough for Waves II, III and V, between the
Na trough and Pa peak for the Na-Pa complex, and
between the N1 trough and P2 peak for the N1-P2
complex.
505
RESULTS
Effects of electrode location and stimulus level on
evoked potential amplitudes and latencies were
evaluated with separate repeated-measures analyses of variance (ANOVAs). The three stimulus levels
of 100% and 75% of the BDR and the evoked potential threshold were chosen because almost all subjects produced waveforms at these levels. The
threshold of Wave V, the Na-Pa complex, and the
N1-P2 complex were each examined with separate
1-way repeated-measure ANOVA to examine the
effect of the implanted electrode location. None of
the subjects had EABR, EAMLR, or ELAR present
at the BT, and therefore the recordings obtained at
BT are not discussed in the results.
EABR
EABR waveforms were recorded from 9 of 11
subjects. Two subjects had no measurable EABRs on
any of the three electrodes at any stimulus presentation level. For those subjects with EABRs, recordings contained one to three positive peaks, Waves II,
III, and V.
Morphology • Figure 1 displays EABR tracings recorded from Electrodes 1, 4, and 7 for two subjects at
the stimulus current level equal to 100% of the
subject’s BDR. Subject 1 had Waves II, III, and V
present on each tested electrode. Subject 5 had
Waves II, III and V present on Electrode 1, but only
Waves III and V on Electrodes 4 and 7. For all
subjects, Electrode 1 (apical location) tended to produce the best responses, i.e., those with the largest
amplitude, best morphology, and greatest number of
positive peaks. Recordings from the basal end of the
electrode array (Electrode 7) tended to have the
poorest morphology and smallest amplitudes.
Wave V was present at the upper portion of the
BDR (100% and 75%) for all subjects. Wave III was
present in the majority of subjects on Electrode 1 at
100% and 75% of the BDR. Wave II was present in
only 55% of subjects on Electrode 1 at 100% and 75%
of the BDR. Recordings on Electrode 7 showed fewer
EABR components overall, especially when the
stimulus level was at 50% or less of the BDR.
Absolute Latency and Interpeak Interval Measures • Table 2 displays the mean absolute latency
values and one standard deviation from the mean
for Waves II, III, and V for all subjects at each
stimulus level and each electrode, and the mean
latency across all electrodes. Wave V latency increased slightly as stimulus level was decreased.
Wave V latency is somewhat longer for the basal
Electrode 7 than either Electrode 1 or 4 at 100% or
75% of the BDR, and at Wave V threshold (WVT).
The mean latency values for all electrodes across
506
Figure 1. EABR tracings from two subjects at the stimulus
level that corresponded to 100% of the BDR. Recordings
from electrodes that represent apical (el 1), mid (el 4), and
basal (el 7) stimulation are displayed for Subject 1 (upper
panel) and Subject 5 (lower panel).
stimulus level increased on average 0.4 msec from
100% of the BDR to WVT.
For Wave V latency, ANOVAs with repeated
measures for electrode site and stimulus level were
completed. There were significant effects for both
electrode (F ⫽ 8.044, df ⫽ 2,16, p ⫽ 0.004) and
stimulus level (F ⫽ 36.719, df ⫽ 2,16, p ⬍ 0.001).
Post hoc comparisons for electrode site alone were
significant for Electrodes 1 versus 7 (F ⫽ 10.912, df
⫽ 1,8, p ⫽ 0.011) and Electrodes 4 versus 7 (F ⫽
10.630, df ⫽ 1,8, p ⫽ 0.012). Post hoc comparisons
for stimulus level were significant for all comparisons (p ⱕ 0.05). There were no significant interactions between electrode site and stimulus level for
Wave V latency. The mean interpeak latencies at
100% of the BDR averaged across electrodes for the
intervals II-III, III-V, and II-V were 0.79 msec, 1.62
msec, and 2.41 msec, respectively.
EAR & HEARING / DECEMBER 2002
Amplitude Measures • The mean Wave V amplitudes at 100% of the BDR for Electrode 1, 4, and 7
were 1.46 ␮V, 1.42 ␮V, and 1.16 ␮V, respectively.
The differences in mean Wave V amplitude across
stimulus levels and across electrodes are illustrated
in Figure 2. The mean amplitude of Wave V decreased with decrease in stimulus current for each
electrode. Electrode 1 tended to have the largest
Wave V amplitude, followed by Electrode 4 and then
7, although these differences were not statistically
significant.
There were significant effects for electrode stimulus level (F ⫽ 17.680, df ⫽ 2, 16, p ⬍ 0.001). Post
hoc tests for stimulus level were significant for
comparison of 100% and 75% of the BDR (F ⫽ 8.844,
df ⫽ 1,8, p ⫽ 0.018), 100% of the BDR and WVT (F
⫽ 19.687, df ⫽ 1,8, p ⫽ 0.002), and for 75% of the
BDR and WVT (F ⫽ 21.113, df ⫽ 1,8, p ⫽ 0.002).
There were no significant interactions between electrode site and stimulus level for Wave V amplitude.
The mean Wave II and III amplitude values for
all subjects at each tested electrode and stimulus
level, and the mean amplitude across all electrodes
is shown in Table 3. Amplitudes for Waves II and III
were smaller than those for Wave V, decreased with
decreasing stimulus level, and were substantially
smaller for Electrode 7 than for Electrodes 1 or 4.
Threshold Measures • Wave V was the most robust EABR component. EABR tracings from Subject
2 on Electrode 4 recorded at seven stimulus levels
are shown in Figure 3. The levels include 100%,
75%, 50%, and 25% of the BDR, as well as the level
producing WVT, and the stimulus level one step
below that needed for threshold. The waveform
elicited during a control run with a stimulus level of
3 CU is also shown.
Stimulation of Electrode 1 resulted in the lowest
mean WVT, followed by Electrode 4, and then Electrode 7 (highest mean threshold). The difference in
threshold across electrodes was statistically significant (F ⫽ 10.211, df ⫽ 2,16, p ⫽ 0.001) as were post
hoc tests for all WVT electrode comparisons (p ⱕ
0.05).
EAMLR
EAMLR waveforms were recorded on at least two
electrodes from 8 of 11 subjects. Three subjects had
no measurable EAMLRs on any electrodes at any
stimulus level. One subject had responses for Electrodes 1 and 4, but no response for Electrode 7.
Morphology • In Figure 4, EAMLR responses at
100% of the BDR from Subjects 5 and 6 for Electrodes 1, 4, and 7 are displayed. EAMLR waveforms
were similar in morphology across electrodes within
subjects. The Na-Pa complex tended to decrease in
EAR & HEARING, VOL. 23 NO. 6
507
TABLE 2. Means and standard deviations for EABR Waves II, III, and V absolute latencies across electrodes and stimulus levels
Wave II Absolute Latency
Electrode 1
Electrode 4
Electrode 7
All Electrodes
Percent of BDR
Mean
SD
n
Mean
SD
n
Mean
SD
n
Mean
SD
n*
100
75
50
25
1.29
1.31
1.37
1.38
0.11
0.11
0.06
0.08
5
5
3
2
1.31
1.33
1.34
0.08
0.07
0.08
3
3
3
1.29
1.29
0.07
0.07
2
2
1.30
1.31
1.35
1.38
0.09
0.08
0.07
0.08
10
10
6
2
Wave III Absolute Latency
Electrode 1
Electrode 4
Electrode 7
All Electrodes
Percent of BDR
Mean
SD
n
Mean
SD
n
Mean
SD
n
Mean
SD
n*
100
75
50
25
2.12
2.09
2.16
2.22
0.16
0.13
0.13
0.14
9
8
6
5
2.08
2.18
2.04
2.06
0.16
0.24
0.05
0.06
8
7
4
2
2.07
2.14
2.04
0.41
0.34
0.30
7
7
3
2.09
2.14
2.08
2.15
0.17
0.19
0.08
0.10
24
22
13
7
Wave V Absolute Latency
Electrode 1
Electrode 4
Electrode 7
All Electrodes
Percent of BDR
Mean
SD
n
Mean
SD
n
Mean
SD
n
Mean
SD
n*
100
75
50
25
3.69
3.70
3.83
3.86
0.14
0.16
0.21
0.27
8
9
9
6
3.68
3.70
3.80
3.84
0.16
0.18
0.17
0.12
8
9
9
5
3.76
3.87
3.87
3.86
0.20
0.20
0.19
0.12
9
9
5
3
3.71
3.76
3.83
3.85
0.17
0.18
0.19
0.17
25
27
23
14
WVT
4.07
0.12
9
4.08
0.17
9
4.17
0.19
9
4.11
0.16
27
Latency values are expressed in msec. BDR ⫽ behavioral dynamic range; WVT ⫽ wave V threshold; n* value represents total electrodes across subjects. The percentage of the behavioral
dynamic range at which wave V threshold occurred was not always less than 25% of the BDR for individual subjects.
amplitude with decrease in stimulus current level.
However, at lower stimulus levels, there was less
change in amplitude and the response disappears
rapidly as the visual detection threshold is crossed.
This characterization has been reported with acoustic recordings of the MLR in normal-hearing sub-
Figure 2. Amplitude of the EABR Wave V response averaged
across subjects and plotted for each tested electrode (1 open
bar, 4 filled bar, 7 hatched bar) at the stimulus levels that
represent 100%, 75%, 50%, and 25% of the BDR. WVT was
not always less than 25% of the BDR for individual subjects.
Error bars represent one standard deviation from the mean.
jects, where the growth of Pa amplitude occurs at
stimulus levels between 40 to 80 dB HL, with
relatively small change in amplitude for stimuli
below 40 dB HL (Özdamar & Kraus, 1983). Components Na and Pa were present in all subjects with
responses on Electrode 1 and 7 at 100% and 75% of
the BDR, and on Electrode 4 at 100%, 75%, and 50%
of the BDR. At least 50% of the subjects had Na-Pa
responses at 25% of the BDR for Electrodes 1 and 4.
Absolute Latency Measures • Mean latency data
for Na and Pa are shown in Table 4. Absolute
latencies for both Na and Pa increased slightly as
stimulus level decreased from 100% of the BDR to
threshold. For Na and Pa, respectively, the average
latency increase was 0.74 and 1.33 msec for Electrode 1, 1.19 and 0.60 msec for Electrode 4, and 1.41
and 0.26 msec for Electrode 7.
For the latency of Na, there were significant
effects for electrode (F ⫽ 5.178, df ⫽ 2,12, p ⫽ 0.024)
but not for stimulus level (F ⫽ 3.770, df ⫽ 2,12, p ⫽
0.054). Post hoc comparisons for electrode site were
significant for only Electrodes 4 versus 7 (F ⫽ 6.539,
df ⫽ 1,6, p ⫽ 0.043). There were no significant
interactions between electrode site and stimulus
level for Na latency. For Pa latency, there were no
508
EAR & HEARING / DECEMBER 2002
TABLE 3. Means and standard deviations for EABR Wave II and Wave III amplitudes across electrodes and stimulus levels.
Wave II Amplitude
Electrode 1
Electrode 4
Electrode 7
All Electrodes
Percent of BDR
Mean
SD
n
Mean
SD
n
Mean
SD
n
Mean
SD
n*
100
75
50
25
0.95
0.64
0.73
0.32
0.41
0.48
0.19
0.23
5
5
3
2
1.05
0.79
0.39
0.49
0.18
0.08
3
3
3
0.78
0.47
0.35
0.06
2
2
1.00
0.65
0.56
0.32
0.41
0.35
0.23
0.23
10
10
6
2
Wave III Amplitude
Electrode 1
Electrode 4
Electrode 7
All Electrodes
Percent of BDR
Mean
SD
n
Mean
SD
n
Mean
SD
n
Mean
SD
n*
100
75
50
25
1.06
0.88
0.72
0.42
0.74
0.47
0.47
0.23
9
8
6
5
0.86
0.99
0.77
0.43
0.60
0.60
0.24
0.22
8
7
4
2
0.52
0.40
0.46
0.39
0.33
0.37
7
7
3
0.82
0.76
0.73
0.42
0.65
0.52
0.35
0.21
24
22
13
7
Amplitude values are expressed in ␮V. BDR ⫽ behavioral dynamic range; n* value represents total electrodes across subjects.
significant effects for either electrode site or stimulus level.
Amplitude Measures • The Na-Pa amplitude was
determined by measuring the trough-to-peak amplitude between the negative Na trough and the positive peak of Pa. In Figure 5, the mean Na-Pa
amplitude data by electrode site and stimulus level
are illustrated. At 100% of the BDR, Electrode 1
showed the largest amplitude, followed by Electrodes 4 and 7, but these were not statistically
significant differences.
The Na-Pa amplitude decreased with a decrease
in stimulus level. There were significant effects for
stimulus level (F ⫽ 10.991, df ⫽ 2,12, p ⫽ 0.002).
Post hoc comparisons for stimulus level were signif-
icant for 100% of the BDR versus the Na-PaT (F ⫽
19.189, df ⫽ 1,6, p ⫽ 0.005), and 75% of the BDR and
the Na-PaT (F ⫽ 15.388, df ⫽ 1,6, p ⫽ 0.008). There
were no significant interactions between electrode
Figure 3. EABR waveforms recorded from Subject 2 on
Electrode 4 are displayed. Two recordings are shown for each
stimulus current level ranging from 100% of the BDR to a
control run at 3 CU. WVT was identified at 218 CU.
Figure 4. EAMLR tracings (Na-Pa) obtained from two subjects
at the stimulus level that corresponded to 100% of the BDR.
Recordings from Electrodes 1, 4, and 7 are displayed for
Subject 5 (upper panel) and Subject 6 (lower panel).
EAR & HEARING, VOL. 23 NO. 6
509
TABLE 4. Means and standard deviations for EAMLR Na and Pa absolute latencies across electrodes and stimulus levels
Na Absolute Latency
Electrode 1
Percent of BDR
100
75
50
25
NaT
Electrode 4
Electrode 7
All Electrodes
Mean
SD
n
Mean
SD
n
Mean
SD
n
Mean
SD
n*
15.78
15.23
15.68
16.79
1.47
1.32
1.24
1.61
8
8
6
5
15.24
15.56
16.09
16.28
1.56
1.26
1.04
2.33
8
8
8
4
15.20
15.81
16.70
16.52
1.38
0.91
0.85
1.11
7
7
5
3
15.41
15.53
16.16
16.53
1.47
1.16
1.04
1.68
23
23
19
12
16.52
1.35
8
16.43
1.25
8
16.61
1.07
7
16.52
1.22
23
Pa Absolute Latency
Electrode 1
Percent of BDR
100
75
50
25
PaT
Electrode 4
Electrode 7
All Electrodes
Mean
SD
n
Mean
SD
n
Mean
SD
n
Mean
SD
n*
25.76
26.66
27.54
28.17
0.74
0.47
0.47
0.23
8
8
6
5
26.86
27.44
27.28
27.25
1.91
1.88
2.26
2.25
8
8
8
4
26.92
27.17
27.53
28.45
1.68
1.68
2.10
1.65
7
7
5
3
26.37
27.09
27.45
27.96
1.77
1.78
2.05
2.00
23
23
19
12
27.09
1.77
8
27.04
2.25
8
27.18
2.34
7
27.10
2.12
23
Latency values are expressed in msec. BDR ⫽ behavioral dynamic range; Values at NaT ⫽ latency of Na at the threshold of the Na-Pa complex; Values at PaT ⫽ latency of Pa at the threshold
of the Na-Pa complex; n* value represents total electrodes across subjects. The percentage of the behavioral dynamic range at which threshold occurred for the Na-Pa complex was not
always less than 25% of the BDR for individual subjects.
site and stimulus level for the amplitude of the
Na-Pa complex.
Threshold Measures • Stimulation of Electrode 1
resulted in the lowest Na-PaT, followed by Electrode
4, and then Electrode 7. These differences in threshold across electrodes were statistically significant (F
⫽ 8.623, df ⫽ 2,12, p ⫽ 0.005). Post hoc comparisons
were significant for Electrodes 1 and 4 (F ⫽ 11.195,
df ⫽ 1,6, p ⫽ 0.015) and Electrodes 1 and 7 (F ⫽
12.650, df ⫽ 1,6, p ⫽ 0.012), but not for Electrodes 4
and 7.
Figure 5. Amplitude of the Na-Pa complex averaged across
subjects and plotted for Electrodes 1 (open bar), 4 (filled bar),
and 7 (hatched bar) at the stimulus levels that represent
100%, 75%, 50%, and 25% of the BDR. Error bars represent
one standard deviation from the mean.
ELAR
Electrical N1-P2 cortical responses of the ELAR
were recorded for at least two electrodes from 9 of 11
subjects. As with the EABR and EAMLR, two subjects had no measurable responses. One subject, who
had no EAMLRs, had measurable N1-P2 responses
for Electrode 1 only. The same subject who had no
EAMLR response present for Electrode 7 did not
produce an ELAR on that electrode.
Morphology • Figure 6 displays N1-P2 responses
at 100% of the BDR for two subjects for Electrodes 1,
4, and 7. The consistency across subjects and across
electrodes for the N1-P2 waveform was similar to
that observed for the Na-Pa complex. The effects of
decreasing stimulus level were less orderly than
those noted for the EABR, and more similar to those
observed with the EAMLR. Components N1 and P2
were present in all subjects on Electrode 1 at 100%
and 75% of the BDR, with fewer components recorded at 50% and 25% on all three electrodes. The
fewest responses were present at 25% of the BDR for
Electrode 7.
Absolute Latency Measures • The mean N1 and
P2 absolute latency values and one standard deviation from the mean for all subjects, and the average
values across electrodes are shown in Table 5. The
mean latency across electrodes at 100% of the BDR
was 86.50 msec for N1 and 181.02 msec for P2. The
latencies for N1 and P2 increased with decreased
stimulus level from 100% of the BDR to the thresh-
510
EAR & HEARING / DECEMBER 2002
old of the N1-P2 complex. N1 increased across the
BDR by 3.90 msec for Electrode 1, 5.41 msec for
Electrode 4, and 3.73 msec for Electrode 7, while P2
increased 7.82 msec, 8.78 msec, and 9.07 msec for
Electrodes 1, 4, and 7, respectively.
For both the N1 and P2 latencies, there were no
significant main effects for electrode but there were
significant effects for stimulus level (for N1, F ⫽
7.701, df ⫽ 2,12, p ⫽ 0.007; for P2, F ⫽ 62.764, df ⫽
2,12, p ⬍ 0.001). For N1 latency, post hoc comparisons for stimulus level were significant for 100%
versus 75% of the BDR (F ⫽ 16.240, df ⫽ 1,6, p ⫽
0.007) and 100% of the BDR versus Na-PaT (F ⫽
10.325, df ⫽ 1,6, p ⫽ 0.018). For P2 latency, post hoc
tests for stimulus level were significant for all comparisons (p ⬍ 0.01). There were no significant interactions between electrode site and stimulus level for
either N1 or P2 latency.
Amplitude Measures • The N1-P2 amplitude was
determined by measuring the trough-to-peak amplitude between the negative N1 trough and the
positive P2 peak. The mean amplitude at 100% of
the BDR for Electrodes 1, 4 and 7 was 5.65 ␮V,
4.78 ␮V, and 3.85 ␮V, respectively. Figure 7 illustrates the mean N1-P2 amplitude data for subjects
by electrode site and stimulus level. The N1-P2
amplitude on each electrode decreased with decrease in stimulus level. Consistent with the
EABR and EAMLR results, the N1-P2 amplitude
tended to be smallest for Electrode 7 compared
with Electrode 1 or 4, but differences were not
significant statistically.
Figure 6. ELAR tracings (N1-P2) obtained from two subjects at
the stimulus level that corresponded to 100% of the BDR.
Recordings are displayed for stimulation of Electrodes 1, 4,
and 7 for Subject 2 (upper panel) and Subject 11 (lower
panel).
TABLE 5. Means and standard deviations for ELAR N1 and P2 absolute latencies across electrodes and stimulus levels
N1 Absolute Latency
Electrode 1
Electrode 4
Electrode 7
All Electrodes
Percent of BDR
Mean
SD
n
Mean
SD
n
Mean
SD
n
Mean
SD
n*
100
75
50
25
87.31
87.54
87.30
88.75
11.51
12.31
10.68
12.12
9
9
7
6
86.85
87.85
87.74
88.66
15.14
15.20
12.54
13.16
8
8
7
7
85.04
86.69
86.93
74.70
13.10
16.48
13.45
11.46
7
7
5
2
86.50
87.40
87.35
86.83
12.71
13.06
11.55
12.66
24
24
19
15
N1T
91.21
15.24
9
92.26
15.19
8
88.77
11.75
7
90.85
13.75
24
P2 Absolute Latency
Electrode 1
Electrode 4
Electrode 7
All Electrodes
Percent of BDR
Mean
SD
n
Mean
SD
n
Mean
SD
n
Mean
SD
n*
100
75
50
25
182.49
184.34
183.37
185.90
12.97
12.22
13.92
14.61
9
9
7
6
182.76
184.15
185.16
187.52
12.49
12.92
14.11
15.95
8
8
7
7
177.14
181.36
184.52
172.50
12.52
13.35
11.88
14.85
7
7
5
2
181.02
183.41
185.16
184.89
12.38
12.29
12.96
15.07
24
24
19
15
P2T
190.31
11.14
9
191.54
14.80
8
186.21
11.59
7
189.52
12.33
24
Latency values are expressed in msec. BDR ⫽ behavioral dynamic range; Values at N1T ⫽ latency of N1 at the threshold of the N1-P2 complex; Values at P2T ⫽ latency of P2 at the threshold
of the N1-P2 complex; n* value represents total electrodes across subjects. The percentage of the behavioral dynamic range at which threshold occurred for the N1-P2 complex was not
always less than 25% of the BDR for individual subjects.
EAR & HEARING, VOL. 23 NO. 6
511
those for N1 and P2 were significantly correlated (r
⫽ 0.87 to 0.96). A similar analysis was performed for
the measures of amplitude and threshold for all
evoked potentials, and significant relations for the
waveform components were not indicated for this
subject sample.
DISCUSSION
Figure 7. Amplitude of the N1-P2 complex averaged across
subjects and plotted for Electrodes 1 (open bar), 4 (filled bar),
and 7 (hatched bar) at the stimulus levels that represent
100%, 75%, 50%, and 25% of the BDR. Error bars represent
one standard deviation from the mean.
There were significant effects for stimulus level
(F ⫽ 28.687, df ⫽ 2,12, p ⬍ 0.001). Post hoc
comparisons for stimulus level were significant for
100% of the BDR versus N1-P2T (F ⫽ 26.379, df ⫽
1,6, p ⫽ 0.002), and 75% of the BDR versus
N1-P2T (F ⫽ 51.296, df ⫽ 1,6, p ⬍ 0.001). There
were no significant interactions between electrode
site and stimulus level for the amplitude of the
N1-P2 complex.
The amplitude of N1-P2 was plotted as a function
of stimulus current level for individual subjects and
electrodes. For each subject, an increase in current
level generally resulted in an increase in amplitude,
as observed for the other evoked potentials under
study. There were five subjects and eight tested
electrodes that showed evidence of amplitude saturation. One subject, S8, had little amplitude growth
for N1-P2, which also was observed for the EABR.
Threshold Measures • Stimulation of Electrode 1
had the lowest mean threshold, and Electrode 7 had
the highest. The differences in threshold across
electrodes were statistically significant (F ⫽ 10.871,
df ⫽ 2,14, p ⫽ 0.001). Post hoc testing indicates
thresholds for Electrodes 1 and 7 (F ⫽ 22.219, df ⫽
1,7, p ⫽ 0.002) and Electrodes 4 and 7 (F ⫽ 8.806, df
⫽ 1,7, p ⫽ 0.025) were significantly different.
Comparison of Wave V, Na-Pa, and N1-P2
Across Subjects
The latencies of Wave V, Na, Pa, N1 and P2 were
included in analyses for each electrode at 100% of
the BDR to determine whether effects of latency
were consistent across the evoked potential responses under study. The analyses were repeated
for 75% of the BDR and the evoked potential threshold stimulus level. Of the latency comparisons, only
A primary goal of this study was to record the
EABR, the EAMLR, and the ELAR and to characterize the responses with regard to the effects of
electrode location and stimulus current level. To
achieve this goal, evoked potentials were elicited on
three intra-cochlear electrodes along the implanted
array, and responses were recorded at five stimulus
levels that corresponded to proportions of the BDR.
The evoked potential thresholds for the EABR,
EAMLR, and ELAR were also obtained for each
subject on each electrode.
Presence of the EABR, EAMLR, and ELAR
For the EABR, Wave V was the most robust, was
present on more electrodes, and was identified at
more stimulus presentation levels than Waves II
and III. Wave I could not be identified due to
stimulus artifact. Past studies (Brown, Abbas, &
Gantz, 1990) and present studies with newer implant technology (Abbas et al., 1998; Franck &
Norton, 2001) have demonstrated that the electrical
whole nerve action potential (or wave I of the EABR)
can be recorded with innovative recording paradigms that overcome the effects of stimulus artifact.
For the EAMLR, the Pa component was the most
robust, and Na and Pa were present for all subjects
at high proportions of the BDR. For the ELAR, wave
P2 was the most robust, and N1 and P2 were present
at 100% and 75% of the BDR for most subjects and
most electrodes.
Effects of Electrode Site and Stimulus Level
on the EABR, EAMLR, and ELAR
For the main effect of electrode site, Wave V
latency was significantly longer for Electrode 7 (basal) compared with either Electrode 1 (apical) or
Electrode 4 (mid). Some studies have reported significant latency differences across electrodes for
subjects where basal electrodes have longer latencies than apical electrodes (Abbas & Brown, 1991;
Shallop et al., 1990). Other studies have reported no
significant differences (Abbas & Brown, 1988; van
den Honert & Stypulkowski, 1986). Wave V latency
increased as stimulus current level decreased. This
also occurs with acoustic stimulation, although with
electrical stimulation, the latency changes are
512
smaller with corresponding changes in electrical
stimulus current. There are no published reports for
comparison of latency characteristics across electrodes with subjects who have received the Clarion
implant.
Wave V amplitudes were largest for Electrode 1,
second largest for Electrode 4, and smallest for
Electrode 7, but these differences were not statistically significant. The amplitude values suggest that
direct electrical stimulation tends to produce a Wave
V that is larger in amplitude than that produced
with acoustic stimuli, a finding noted in other reports of electrically evoked potentials (Abbas, 1993;
van den Honert & Stypulkowski, 1986) and attributed to increased synchrony of auditory nerve fiber
responses for electrical stimulation.
The WVT for each electrode was significantly
different. The lowest WVT was for Electrode 1 and
the highest for Electrode 7. Threshold measures are
influenced by the proximity of the implanted electrode array to spiral ganglion cells. It may be that
higher thresholds for the most basal electrode are a
result of greater distance from neural elements
compared with more apical electrodes, which are
located further along the scala tympani.
For the EAMLR, there were fewer significant
effects related to latency compared with the EABR.
The Na component latency was significantly different for Electrodes 4 and 7 only, and for Pa, there
were no significant latency differences for electrode
site. Although Na and Pa latencies increased
slightly with decrease in stimulus current level, the
differences were not statistically significant. This
finding regarding EAMLR and EABR latencies with
respect to stimulus intensity parallels another report that used acoustic stimuli for recordings from
normal-hearing subjects, in which the ABR was
found to be highly stimulus intensity dependent but
the AMLR was not (Özdamar & Kraus, 1983).
There are few published studies that discuss the
EAMLR parameters, including latency, when recorded from intracochlear electrodes. A number of
studies have reported findings of the EAMLR with
transtympanic stimulation. Acoustic and electrical
MLRs were compared for six subjects (Kileny et
al.,1989) where electrical Pa latencies were shorter
than acoustic Pa latencies (average electrical Pa
latency of 31 msec, acoustic Pa latency 35 msec). In
the present study, the average Pa latency for intracochlear stimulation across electrodes (26.37 msec
at 100% of the BDR) was also earlier than typical Pa
latencies elicited with acoustic stimuli.
The amplitude of the Na-Pa complex was largest
for Electrode 1, and smallest for Electrode 7, but
these differences were not statistically significant.
The main effect for stimulus level on the amplitude
EAR & HEARING / DECEMBER 2002
of the response was significant for the EAMLR, but
not at all stimulus levels, as it was for the EABR.
This finding is similar to comparisons reported by
Özdamar and Kraus (1983) between the acoustic
ABR and AMLR recorded from the same subjects in
which the Wave V response continued to increase in
amplitude with increase in stimulus intensity while
the Pa response showed no further increase in
amplitude beyond 50 to 60 dB HL. Electrical Pa
amplitudes have been shown to be larger (mean of
1.01 ␮V) than acoustic Pa amplitudes (mean of 0.70
␮V) in the same subjects (Kileny et al.,1989). This
contrast may result from enhanced neural synchrony with electrical stimulation.
Similar to the EABR in the present study, the
electrical threshold for the Na-Pa complex (Na-PaT)
was lower for Electrode 1, highest for Electrode 7,
and differences by electrode site were generally
significant. There are no other studies of the
EAMLR components with electrical stimulation that
report the evoked potential thresholds by electrode
site.
For the ELAR, neither the latency of N1 nor the
latency of P2 showed statistically significant differences between Electrodes 1, 4, and 7. The average
latency across electrodes at 100% of the BDR was 86
msec for N1 and 181 msec for P2, which may be
slightly earlier than the acoustic counterpart for
each component, depending on the comparison
study’s subject sample and stimulus characteristics.
In a study by Ponton and Don (1995), peak latencies
(N1 and the MMN difference wave) for Nucleus 22
cochlear implant subjects were earlier (10 to 20
msec) than the responses for normal-hearing subjects. For computer-generated speech stimuli delivered through a loudspeaker in the sound field, a
study of the P300 response indicated that N1 and P2
latencies were not significantly different between
nine Nucleus implant subjects and nine normalhearing control subjects (Micco et al., 1995). Click
stimuli delivered directly to the implanted electrodes, as in the present study and that of Ponton
and Don (1995), result in earlier ELAR latencies
than those elicited with speech stimuli delivered
through a loudspeaker in the sound field, as in the
Micco et al. (1995) study.
There were significant differences found for the
main effect of stimulus level for the latency of both
N1 and P2. Previous studies tended to elicit the
ELAR at suprathreshold stimulus levels; therefore,
there are no published reports of the changes in
latency for N1 or P2 in cochlear implant users with
stimulus levels in the lower portion of the BDR, or at
the threshold of the evoked response. With acoustic
studies, latency changes with intensity for N1 and
EAR & HEARING, VOL. 23 NO. 6
P2 vary, depending on whether tonal, click or speech
stimuli have been used to evoke the response.
As seen with the EAMLR, the amplitude of the
N1-P2 complex of the ELAR was largest for Electrode 1 and smallest for Electrode 7, but statistical
tests indicate no significant effects for electrode site.
In the Ponton and Don (1995) study, the MMN was
larger and more frequently identified for the apical
electrode pair than for the basal pair. In the current
study, there were also significant main effects for
stimulus level at 100% or 75% of the BDR versus the
threshold. Until now, there are no published reports
of the changes in amplitude for N1 or P2 with
differing stimulus levels in cochlear implant users.
Acoustical studies indicate that amplitude increases
with increase in stimulus intensity, but that the
largest increase occurs within the first 20 to 30 dB
above the auditory threshold, followed by increases
in amplitude that are more gradual with increase in
stimulus intensity, and for some subjects at high
intensities, amplitude measures can reach a point of
saturation (Buchsbaum, 1976; Davis & Zerlin, 1966;
Näätänen & Picton, 1987; Picton, Woods, BaribeauBraun, & Healey, 1977; Rapin, Schimmel, Tourk,
Krasnegor, & Pollak, 1966). In the present study,
there were 10 instances of apparent amplitude saturation for the ELAR (average amplitude decrease
compared with the maximum amplitude of 0.92 ␮V,
and a range of 0.14 ␮V to 2.51 ␮V) compared with
five occurrences for the EAMLR (average amplitude
decrease of 0.54 ␮V, and a range of 0.15 ␮V to 1.01
␮V), and four occurrences for the EABR that were of
very small magnitude (average amplitude decrease
of 0.05 ␮V, and a range of 0.03 ␮V to 0.09 ␮V).
The pattern of lowest evoked potential threshold
for Electrode 1, second lowest for Electrode 4, and
highest for Electrode 7 held true for the ELAR as
seen for the EABR and the EAMLR. For the ELAR,
there were statistically significant differences for
the threshold comparisons of Electrodes 1 and 7, and
4 and 7. There are no other published reports for late
auditory response components with electrical stimulation and implant users that report threshold
differences across electrode site.
Interpeak Intervals
In this sample of subjects, interpeak intervals for
the EABR wave components occurred at shorter
intervals than those that are typical for the acoustic
ABR (i.e., 1 msec between each positive peak from
Wave I to Wave V). As early as 1985, Gardi reported
reduced interpeak latency values (0.75 msec to 0.80
msec). EABR interpeak latencies in the present
study were 0.79 msec (II-III), 1.62 msec (III-V) and
2.41 msec (II-V). In other publications that have
513
compared acoustic and electrical evoked potentials
at higher levels of the auditory system, such as the
middle and late latency responses, the electrical
components often occur earlier than the acoustic
components resulting in shorter latencies, as described previously. Enhanced neural synchrony with
electrical stimulation may be related to compressed
interpeak intervals throughout the auditory pathway. The site of initial electrical stimulation in the
periphery bypasses the cochlear transduction processes and thus accounts for some of the small
decreases in the latency of peaks. This small reduction, however, can not account for the larger differences observed in peak latencies for more centrally
evoked potentials, which may be related to faster
conduction time throughout the auditory system
with electrical stimulation.
Recording Artifact
Several studies have indicated that signal contamination can occur when recording electrically
evoked potentials that are sources of stimulus or
muscle artifact (Abbas & Brown, 1991; van den
Honert & Stypulkowski, 1986). Stimulus artifact
interferes with the electrical recording of Wave I
of the EABR. The spread of electrical stimulation
can result in activation of nonauditory tissue or
nerve fibers, such as muscle or vestibular endings,
particularly at high stimulus current levels, and
may result in large amplitude responses that are
atypical. In this study, unusually large amplitude
responses were observed at 100% of the BDR for
one subject for EABR Wave V (Subject 6) and one
subject for the Na-Pa complex (Subject 10) that
were well outside the amplitude range for the
respective response based on the other subjects,
and were probably associated with nonauditory
stimulation. Subject 1 had a large potential that
occurred at 7 to 8 msec during the EAMLR recording at 100% of the BDR on Electrodes 1 (stimulus
level 679 CU), 4 (stimulus level 751 CU), and 7
(stimulus level 635 CU), which then disappeared
at stimulus levels that represented 75% of the
BDR for each electrode. This potential was several
milliseconds earlier than the latencies of Na and
Pa, and did not interfere with the measures of the
later Na and Pa components. Subject 2 also demonstrated a large triphasic potential between 5
msec and 9 msec at 100% of the BDR for Electrodes 1 (stimulus level 702 CU), 4 (stimulus level
679 CU), and 7 (stimulus level 679 CU), that was
most likely stimulation of nonauditory fibers.
Electrically evoked auditory potentials that are
not similar to acoustically evoked auditory potentials in waveform morphology or that have unusu-
514
EAR & HEARING / DECEMBER 2002
ally large amplitudes should be interpreted
cautiously.
In conclusion, this study summarizes the response parameters of the EABR (Waves II, III, V),
EAMLR (Na-Pa complex), and ELAR (N1-P2 complex) obtained from 11 Clarion cochlear implant
users. Results of electrophysiologic recordings from
three intra-cochlear electrodes and five stimulus
current levels that corresponded to proportions of
the BDR are presented. It is possible that other
functional relationships would be evident if additional subjects, electrode sites, and stimulus levels
were included. There are no published studies that
discuss the latency, amplitude, and threshold values
for EABR, EAMLR, and ELAR across the electrical
dynamic range that have been elicited from the
same experimental subjects. Electrically evoked auditory responses can provide a reliable and objective
assessment of auditory function in cochlear implant
recipients. As the need for objective measures with
cochlear implant users increases, it is critical to
understand how these electrical potentials behave
when stimulus parameters are systematically varied. Data from this study may serve as a normative
reference for expected latency, amplitude and
threshold values for the recording of brainstem and
cortical electrically evoked auditory potentials. Responses recorded from cochlear implant users show
many similar patterns, yet important distinctions,
compared with auditory potentials elicited with
acoustic signals.
ACKNOWLEDGMENTS:
This manuscript is based on the Ph.D. dissertation of the first
author, submitted to the University of Illinois, ChampaignUrbana, Department of Speech and Hearing Science, with chair
Ron D. Chambers (U of Illinois) and committee member/mentor
Nina Kraus (Northwestern University). We acknowledge the
contributions of dissertation committee members Dr. Robert C.
Bilger and Dr. Charissa Lansing, University of Illinois; Dr. Mary
Joe Osberger, Advanced Bionics Corporation, Sylmar, CA; and
Dr. Carolyn Brown, University of Iowa. We express appreciation
to Dr. Dawn Koch for feedback and suggestions during the study
period. Finally, we thank the subjects who participated in this
study, and acknowledge sources of funding, equipment, and
support that included Advanced Bionics Corporation, Carle Clinic
Association, Carle Foundation, and the University of Illinois.
Address for correspondence: Jill B. Firszt, Ph.D., Department of
Otolaryngology and Communication Sciences, Medical College of
Wisconsin, 9200 West Wisconsin Avenue, Milwaukee, WI 53226.
E-mail: [email protected]
Received May 21, 2001; accepted September 3, 2002
REFERENCES
Abbas, P. J. (1993). Electrophysiology. In R. S. Tyler (Ed.),
Cochlear Implants: Audiological Foundations. San Diego: Singular Publishing Group.
Abbas, P. J., & Brown, C. J. (1988). Electrically evoked brainstem
potentials in cochlear implant patients with multi-electrode
stimulation. Hearing Research, 36, 153–162.
Abbas, P. J., & Brown, C. J. (1991). Electrically evoked auditory
brainstem response: Growth of response with current level.
Hearing Research, 51, 123–138.
Abbas, P. J., Brown, C. J., Shallop, J. K., Firszt, J. B., Hughes,
M. L., Hong, S. H., & Staller, S. J. (1998). Summary of results
using the Nucleus CI24M implant to record the electrically
evoked compound action potential (EAP). Ear and Hearing, 20,
45–59.
Brown, C. J., Abbas, P. J., Fryauf-Bertschy, H., Kelsay, D., &
Gantz, B. J. (1994). Intraoperative and postoperative electrically evoked auditory brainstem responses in Nucleus cochlear
implant users: Implications for the fitting process. Ear and
Hearing, 15, 177–183.
Brown, C. J., Abbas, P. J., & Gantz, B. J. (1990). Electrically
evoked whole-nerve action potentials I. Data from Symbion
cochlear implant users. Journal of the Acoustical Society of
America, 88, 1385–1391.
Brown, C. J., Hughes, M. L., Lopez, S. M., & Abbas, P. J. (1999).
Relationship between EABR thresholds and levels used to
program the Clarion speech processor. Annals of Otology,
Rhinology, and Laryngology (Suppl. 177), 108, 4 –2, 50 –57.
Burton, N. J., Miller, J. M., & Kileny, P. R. (1989). Middle latency
responses: I. Electrical and acoustic stimulation. Archives of
Otolaryngology, 115, 59 – 62.
Buchsbaum, M. (1976). Self-regulation of stimulus intensity:
Augmenting/reducing and the average evoked response. In
G. E. Schwartz & D. Shapiro (Eds.), Consciousness and SelfRegulation (pp. 101–135). New York: Plenum Publishing Corp.
Davis, H., & Zerlin, S. (1966). Acoustic relations of the human
vertex potential. Journal of the Acoustical Society of America,
39, 109 –116.
Firszt, J. B., Chambers, R. D., & Kraus, N. (2002). Neurophysiology of cochlear implant users II: Comparison among speech
perception, dynamic range and physiological measures. Ear
and Hearing, 23, 516 –531.
Firszt, J. B., Rotz, L. A., Chambers, R. D., & Novak, M. A. (1999).
Electrically evoked potentials recorded in adult and pediatric
Clarion implant users. Annals of Otology, Rhinology, and
Laryngology (Suppl. 177), 108, 4 –2, 58 – 63.
Franck, K. H., & Norton, S. J. (2001). Estimation of psychophysical levels using the electrically evoked compound action potential measured with the neural response telemetry capabilities of Cochlear Corporation’s CI24M device. Ear and Hearing,
22, 289 –299.
Gardi, J. N. (1985). Human brainstem and middle latency responses to electrical stimulation: Preliminary observations. In
R. A. Schindler & M. M. Merzenich (Eds.), Cochlear Implants
(pp. 351–363). New York: Raven Press.
Groenen, P., Snik, A., & van den Broek, P. (1997). Electrically
evoked middle latency responses versus perception abilities in
cochlear implant users. Audiology, 36, 83–97.
Hodges, A. V., Ruth, R. A., Lambert, P. R., & Balkany, T. J.
(1994). Electric auditory brain stem responses in Nucleus
multichannel cochlear implant users. Archives of Otolaryngology-Head and Neck Surgery, 120, 1093–1099.
Jyung, R. W., Miller, J. M., & Cannon, S. C. (1989). Evaluation of
eighth nerve integrity by the electrically evoked middle latency
response. Otolaryngology Head and Neck Surgery, 101, 670 –
682.
Kaga, K., Kodera, K., Hirota, E., & Tsuzuka, T. (1991). P300
response to tones and speech sounds after cochlear implant: A
case report. Laryngoscope, 101, 905–907.
EAR & HEARING, VOL. 23 NO. 6
Kileny, P. R., & Kemink, J. L. (1987). Electrically evoked middle
latency potentials in cochlear implant candidates. Archives of
Otolaryngology Head and Neck Surgery, 113, 1072–1077.
Kileny, P. R., Kemink, J. L., & Miller, J. M. (1989). An intrasubject comparison of electrical and acoustic middle latency
responses. American Journal of Otology, 10, 23–27.
Kraus, N., Micco, A. G., Koch, D. B., McGee, T., Carrell, T.,
Sharma, A., Wiet, R. J., & Weingarten, C. Z. (1993). The
mismatch negativity cortical evoked potential elicited by
speech in cochlear-implant users. Hearing Research, 65, 118 –
124.
Makhdoum, M. J., Groenen, P., Snik, A., & van den Broek, P.
(1997). Intra-and interindividual correlations between auditory evoked potentials and speech perception in cochlear implant users. Scandinavian Audiology, 22, 1– 8.
Mason, S. M., Sheppard, S., Garnham, C. W., Lutman, M. E.,
O’Donoghue, G. M., & Gibbin, K. P. (1993). Application of
intraoperative recordings of electrically evoked ABRs in a
paediatric cochlear implant programme. In F. B. Deguine (Ed.),
Cochlear Implants: New Perspectives (pp. 136 –141). Basel:
Karger.
Micco, A. G., Kraus, N., Koch, D. B., McGee, T. J., Carrell, T. D.,
Sharma, A., Nicol, T., & Wiet, R. J. (1995). Speech-evoked
cognitive P300 potentials in cochlear implant recipients. American Journal of Otology, 16, 514 –520.
Miyamoto, R. T. (1986). Electrically evoked potentials in cochlear
implant subjects. Laryngoscope, 96, 178 –185.
Näätänen, R., & Picton, T. (1987). The N1 wave of the human
electric and magnetic response to sound: A review and an
analysis of the component structure. Psychophysiology, 24,
375– 425.
Oviatt, D. L., & Kileny, P. R. (1991). Auditory event-related
potentials elicited from cochlear implant recipients and hearing subjects. American Journal of Audiology, 1, 48 –55.
Özdamar, O., & Kraus, N. (1983). Auditory middle latency responses in humans. Audiology, 22, 34 – 49.
Picton, T. W., Hillyard, S. A., Krausz, H. I., & Galambos, R.
(1974). Human auditory evoked potentials. I. Evaluation of
components. Electroencephalography and Clinical Neurophysiology, 36, 179 –190.
515
Picton, T. W., Woods, D. L., Baribeau-Braun, J., & Healey,
T. M. G. (1977). Evoked potential audiometry. The Journal Of
Otolaryngology, 6, 90 –119.
Ponton, C. W., & Don, M. (1995). The mismatch negativity in
cochlear implant users. Ear and Hearing, 16, 130 –146.
Rapin, I., Schimmel, H., Tourk, L. M., Krasnegor, N. A., & Pollak,
C. (1966). Evoked responses to clicks and tones of varying
intensity in waking adults. Electroencephalography and Clinical Neurophysiology, 21, 335– 45234.4.
Shallop, J. K., Beiter, A. L., Goin, D. W., & Mischke, R. E. (1990).
Electrically evoked auditory brainstem responses (EABR) and
middle latency responses (EMLR) obtained from patients with
the Nucleus Multichannel cochlear implant. Ear and Hearing,
11, 5–15.
Shallop, J. K., Van Dyke, L., Goin, D. W., & Mischke, R. E. (1991).
Prediction of behavioral threshold and comfort values for
Nucleus 22-channel implant patients from electrical auditory
brainstem response test results. Annals of Otology, Rhinology,
and Laryngology, 100, 896 – 898.
Simmons, F. B., & Smith, L. (1983). Estimating nerve survival by
electrical ABR. Annals of New York Academy of Science, 405,
422– 423.
Starr, A., & Brackmann, D. E. (1979). Brainstem potentials
evoked by electrical stimulation of the cochlea in human
subjects. Annals of Otology, Rhinology and Laryngology, 88,
550 –556.
van den Honert, C., & Stypulkowski, P. H. (1986). Characterization of the electrically evoked auditory brainstem response
(ABR) in cats and humans. Hearing Research, 21, 109 –126.
Walsh, S. N., & Leake-Jones, P. (1982). Chronic electrical stimulation of auditory nerve in cat: physiological and histological
results. Hearing Research, 7, 281–304.
REFERENCE NOTE
1 Firszt, J. B., Gaggl, W., Wackym, P. A., & Reeder, R. M. (2000)
Electrical evoked potentials recorded with the Med El cochlear
implant device. Paper presented at the US Investigator Meeting, Stans, Austria, November.
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