HEARLab: bringing hearing to infants

1
HEARLab: bringing hearing to infants
First Edition
By Robert Martin, Melissa Villaseñor, Harvey Dillon, Lyndal Carter, Suzanne Purdy
1
First Edition (Dec 2008)
This paper introduces “cortical” measurements and employs case studies to suggest ways to
use “cortical” measurements to fit hearing aids and improve hearing aid outcomes for people
who are difficult to test. The short DVD “The NAL HEARLab,” by Frye Electronics, Inc., shows
the HEARLab system being used. The HEARLab NAL-ACA reference guide is a step-by-step
set of instructions on how to operate the HEARLab machine.
Introduction
We are very excited about this topic! It is great fun!
The clinical use of cortical measurements will revolutionize how hearing tests and hearing aid fittings
are done on infants and people who cannot or will
not cooperate with the hearing tests. We want this
paper to be a joy to read. So rather than talking
about the “machine” we want to talk about infants
and the experience their families have with hearing
loss.
This paper begins with the story of Dr. Judy Smith,
an audiologist, and two of her tiny patients: sweet
Suzy and precious Danny. These infants and this
doctor are not “real” people; they are inventions
of the authors. Numerous case studies have been
pulled together into this story, a story of bringing
hearing to infants. This paper ends with the stories of
John Doe and Tiffany; two complicated patients.
This is not a scientific paper. These are fun stories
about a new clinical tool.
The profession of Audiology has changed each time
a new diagnostic tool has been integrated into clinical practice. Audiologists have assimilated ABR,
real ear, and tympanomentry into their clinical armamentarium. We believe Audiology is about to
take another step forward as cortical measurements
are rediscovered and become another standard,
very useful diagnostic test.
Our professional language is jam packed with
hundreds, perhaps thousands of technical terms
on evoked responses. It is hard to keep track of
all of the terms. We will take big liberties with the
technical language and talk about “cortical” curves
or “cortical” responses—please remember this is
a fun story, not a technical publication. When we
say “cortical” curves we are referring to the acronym “CAEP” which stands for Cortical Auditory
Evoked Potentials which we discuss later.
“Cortical” responses allow us to “see” if the patient
is “detecting” a sound. Many patients, old and
young, cannot respond to a hearing test by raising their hand or repeating a word. Infants cannot
talk to you as you adjust the amplification. Their
baby brains, however, produce cortical responses
that give us much insight into their perception of
speech sounds.
The use of “cortical” measurements will not replace ABR, OAEs, or any other standard test. “Cortical” tests give us a new way, a better way to see if
amplified speech sounds are “audible” (detected).
This concept will become clearer as we continue
our story of Suzy and Danny.
Now back to the kids.
Hours after Suzy and Danny were born, they both
failed their newborn hearing screenings. They were
referred to Dr. Judy Smith, a local audiologist. Dr.
Judy obtained extensive histories. She did ABRs,
otoacoustic emission tests (OAEs), tympanomentry and other tests on these little babies. It was obvious they both had noteworthy hearing loss and
needed amplification, so she made impressions for
earmolds and scheduled them for a Pediatric Hearing Aid Evaluation.
But now what?
If you are Dr. Judy and you determine Suzy has
a substantial hearing loss and Danny has a more
profound hearing loss, how do you determine the
exact amount of gain each infant needs? Do you
put all of your “faith” in the target recommendations of NAL or DSL software? Do you use the
guidelines you learned in graduate school? The
most important question is, “How do you determine
if the amplification levels you select for Suzy and Danny
are correct?” Dr. Judy needs a machine that can tell
her whether or not Suzy and Danny are perceiving
speech—their neural systems are detecting speech
sounds—at appropriate levels after they are fitted
with hearing aids.
2
We have great news!!
Speech-Like Test Sounds
With great excitement and pride, Frye Electronics in Oregon and The National Acoustic Laboratories (NAL) in Australia are pleased to introduce
HEARLab, a test instrument used to measure cortical responses—with and without hearing aids.
Currently, there are two parts to HEARLab: ACA
(Aided Cortical Assessment) and CTE (Cortical
Threshold Estimate).
NAL did numerous studies using many different
types of sounds to evoke cortical responses. It was
decided to limit the speech-like test sounds to three
test stimuli: /m/, /g/, and /t/. These consonants
were extracted from continuous discourse spoken
by a female and filtered to match the International
Long-term Average Speech Spectrum (ILTASS). A
high-pass filter was also applied at 250 Hz to /t/
and /g/ to remove additional unwanted low frequency noise.
ACA is used to evaluate hearing aid fittings using
real speech sounds presented through a speaker.
You view the evoked cortical responses of a tiny
infant who is wearing hearing aids. In this story
we call ACA the “aided” tests. It is more properly
called NAL-ACA (National Acoustic Labs – Aided
Cortical Assessment) a study of evoked cortical responses while hearing aids are worn.
CTE is used to estimate unaided hearing thresholds. Tones are presented via insert earphones or
a bone vibrator. The evoked cortical responses are
studied. In this paper we call CTE the “threshold
estimates” or “threshold tests.” We will discuss
both ACA and CTE.
ACA: Aided Cortical Assessment
Dr. Judy has many diagnostic tools at her disposal
that she uses to evaluate the hearing ability of infants. But, until now, none of these tests could directly confirm whether or not the amplification was
stimulating the auditory cortex. Suzy and Danny
are only a few weeks old and we cannot ask them,
“Can you hear the sound of my voice?” In the past
it was almost impossible to evaluate the appropriateness of the amplification.
With HEARLab, we now have a new tool that directly measures the infant’s cortical responses so
we know if amplified speech sounds are detected
by the infant wearing hearing aids.
When you measure the level of these speech sounds
(with a sound level meter set to an “impulse” time
setting) they are within a few decibels of the long
term level of speech from which they were extracted. For example, if the cortical testing is done
with these consonants set to 65 dB, they will be presented very close to the level they would have in
speech whose long-term level was 65 dB SPL.
The inter-stimulus duration is 1125 msec. The stimulus duration is 30 msec. A hundred epochs can be
presented in about two minutes.
These brief consonant sounds have very little of
the vowel transition. They are essentially vowelfree stimuli that have spectral emphasis in the low,
mid, and high frequency regions. Figure 1 shows
the one-third octave spectra for these three speechlike sounds. They have the potential to give diagnostic information about the perception of speech
sounds in different frequency regions.
1/3 o c ta v e s p e c tr a o f s p e e c h s tim u li @ 65d B S P L
70
60
50
40
/m/
/g/
After a little training, you learn to see whether or
not the cortex is responding when speech sounds
are presented. A hearing aid is fitted in the usual
fashion. The speech consonants /m/, /g/, and
/t/ are presented to the infant using sound field
speakers. Insert earphones cannot be used; you
need room for the hearing aids and ear molds to be
placed over and in the ears. The cortical response
is then measured and viewed on the monitor along
with information about testing sufficiency and
whether or not a cortical response is present.
/t/
30
20
10
0
100
1000
10000
Centre frequency (Hz)
Figure 1: The one-third octave spectra for three speech sounds:
/m/, /g/, and /t/.
3
Cortical ≠ ABR
The “cortical” response (CAEP) is the electrical signal produced by the neurons in the auditory cortex
of the brain. We are NOT talking about the brain
stem response measured in an ABR test that happens in the first 10 milliseconds after a stimulus has
been presented. Cortical measurements are done
while the child is awake. Brain stem tests are often
done with the child asleep. The “Cortical” response
happens much later, roughly 50-500 milliseconds
after the onset of the signal. We will return to this
topic again later in this paper.
The Objective for the Aided (ACA) Test
We want to be sure the infant/child gets the information across the frequency spectrum of normal
conversational speech. Low, mid, and high pitched
speech-like sounds are presented to the infant at
the normal speech level (65 dB SPL) while the infant is wearing a hearing aid. The cortical responses are studied.
If needed, the amplification can be changed and the
cortical curves re-evaluated to study the fitting.
If no response is observed with a high level of amplification, and if the averaged waveform has a
sufficiently low level of residual noise (more on
this later), supplements to hearing aids (cochlear
implant, education using sign language etc.) can be
considered.
Two questions haunt the professional fitting hearing aids. First, “Is the patient getting enough sound
to develop speech and language? And second, “Is
the patient getting too much sound—am I causing discomfort or endangering the patient’s hearing with excessive amplification?” We will look
at these two questions as we discuss the stories of
Suzy and Danny. Cortical testing helps us directly
answer the first question, and sometimes gives us
insight into the second question.
Suzy
Suzy is a beautiful, three month old, infant girl.
She is very little! She has a quiet mother, an oldfashioned unsupportive father who is in denial of
her hearing loss. As Suzy grows up, her family will
say she is a graceful, shy, quiet, but a sophisticated
young lady. The story of Suzy has a happy ending.
With well-fitted hearing aids: She hears well. She
gets a top notch education and a job she loves. She
becomes very successful.
But, before this wonderful story can happen, Dr.
Judy has to fine tune this tiny infant’s amplification
and deal with a father who does not believe Suzy
should be wearing hearing aids.
Suzy is held in her mom’s lap, electrodes are placed
on the top of her head, her forehead and mastoid.
She is restless but calms down after she nurses.
Heavy nursing is avoided as infants tend to go to
sleep if they nurse a lot.
Suzy wants to go to sleep but is kept awake by a
video, some toys, and baby-friendly books while
the tests are done. Testing cannot be done if the
child is asleep. Sleep has various stages and most
of them result in diminished or absent cortical responses (relative to those that occur while we are
awake)— just as we are generally less sensitive to
sound when we are asleep. Infants like to sleep and
they go to sleep quickly. Without EEG equipment
and training, it is difficult, perhaps impossible, to
judge if they are lightly asleep or deeply asleep. It
is strongly recommended that the child be awake
for the tests.
Let’s take a look at how Suzy and Danny were fitted with hearing aids. Suzy was born with moderate-severe hearing loss. Danny has mixed profound
hearing loss: complicated because he has fluid in
his middle ears.
Suzy is encouraged to watch a silent cartoon on TV.
The video entertains her and helps to keep her quiet while the tests are done. At times, she becomes
bored watching the TV, so her father entertains her
with her beloved teddy bear. When that does not
work dad uses a light-up toy and a mirror. Babies
4
love to look at themselves in the mirror.
Dr. Judy has many “distracters” she uses to keep
the babies entertained. The child must be calm
while the test is being done. These distracters include:
•
•
•
•
•
•
Small bright colored soft toys and puppets.
Light-up toys.
A TV showing a silent colorful cartoon.
Hand held mirrors.
A shiny, sparkly ball on a string.
Some kids’ books with colorful pictures—
little babies love black, white, and red contrasting colors and pictures of other babies.
• Squeeze toys that changes shape when you
squeeze them. Avoid the ones with squeakers inside.
• Blowing party soap bubbles—silent and
fascinating to most babies.
Stores like Party City have a large supply of fun-tolook-at and fun-to-play-with toys.
“Whirly” light up toys that go round-and-round
are great distracters during electrode placement.
They can be a little too noisy during testing.
When the sparkly-ball-on-a-string is used, dad is
told to not to move it around too much. You don’t
want the child’s eyes and head moving around excessively. You want to keep “muscle noise” to a
minimum.
Suzy is fitted with moderate gain hearing aids; the
left hearing aid fitting is arbitrarily studied first.
Speech sounds are presented at a normal speech
level (65 dB SPL) from a nearby speaker. The
CAEPs, Cortical Auditory Evoked Potentials, are
viewed on the monitor.
The real-ear equipment Dr. Judy uses recommends
35 dB of ave. gain (the gain averaged for the frequencies of 500, 1000, and 2000 Hz) for medium
intensity input signals. Note: for simplicity we use
term “ave.” (averaged) gain.
Dr. Judy is worried about over amplification in Suzy’s case; she has a marked startle response when
the hearing aids are switched on. Dr. Judy’s testing
strategy is to do the tests at the recommended level,
i.e. 65 dB SPL. Then, if Suzy is not too restless, she
plans to reduce the gain and continue the study.
The cortical test is done with the stimulus set to
65 dB SPL, normal speech level. As the data start
to accumulate, the cortical response is quickly and
clearly seen on the monitor, and the statistical calculator indicates that it is a “real” response—i.e.
one related to the stimulus. Dr. Judy is delighted
that the fitting is in the right ball park. These gain
settings are then recorded as Suzy’s recommended
amplification level. The right hearing aid is then
fitted and the process repeated.
After the initial ACA tests Dr. Judy markedly lowers the gain of one of the instruments and she restarts the study. After about 100 epochs she stops
the test; the cortical response was obscure, not as
clear as the initial tests. She concludes the initial
results provided better amplification. She is now
less concerned about the possibility of over amplification.
Note: In Suzy’s case, comprehensive diagnostics
tests (OAEs, ABRs) were done before the cortical measurements (ACA) so no “unaided” testing
was done. If OAE and ABR information had not
been available, then the first cortical measurement
would have been an “unaided” study using tonal
stimuli.
If you suspect the child has auditory neuropathy/
dys-synchrony it is recommended you do “unaided” tests (CTE) before the “aided” tests (ACA). It
is difficult to predict the audiogram from the ABR
for these babies.
Danny
Danny is a robust, beautiful, high-energy infant.
His mother had a difficult, long-duration labor that
ended in an emergency cesarean section. Danny
was large at birth, over nine-and-a-half pounds; a
well developed infant. His Apgar and other birth
scores were all normal.
Danny is held by his father and intently watches
TV while the cortical tests are done. Danny is tested while wearing his high power hearing aids (one
at a time). The test signal is set to 65 dB SPL. Dr.
Judy sets the “ave.” gain of the hearing aids to 55
dB, a flat frequency response.
Danny has fluid in his middle-ears that has probably been there since birth. His tympanograms are
flat. A note about middle-ear fluid was placed in
his chart. He will be rescheduled sooner-than-nor-
5
mal for follow up visits to monitor the changes in
his hearing due to this conductive component. Less
gain will be needed as soon as the fluid clears out
of his ears.
As Danny’s testing with the hearing aid progressed, a waveform that looked a bit like a cortical
response was observed, but the statistical calculator indicated that it was not significantly different
from random noise. Dr. Judy stopped the test and
increased the hearing aid gain by 10 dB. Danny
was hungry so his mother fed him a little while Dr.
Judy was adjusting the hearing aid. The test was restarted, and everyone cheered when clear cortical
responses with correspondingly high significance
levels were observed.
When hearing aids are fitted to adults, we spend a
lot of time asking the patient, “How does it sound?
Is my voice a little too loud or a little too soft?”
HEARLab give us the ability to present speech signals to infants and observe cortical responses that
are highly correlated with perceived speech. The
tests are done with the hearing aid set to the audiologist’s recommended level.
If the desired cortical responses are observed, the
fitting is “verified”—however the word “verified”
should not be taken literally. It is more accurate to
say the fitting is “in the right ballgame” or “a level has been found where the input signal actively
stimulates the auditory cortex’. When cortical responses are observed, we should not assume the
hearing aid fitting needs no further review. Hearing aid fittings on infants are “works in progress”.
Continuous monitoring and repetitive re-evaluations are needed.
If the desired cortical responses are not observed,
and the residual noise is low enough that a response
should be visible --- if one was actually there, --then everything about the fitting must be checked:
• Are the electrodes still properly attached?
• Do they still have sufficiently low impedance?
• Is the child still awake?
• Were the thresholds on which the prescription was based, correct?
• Were the thresholds properly transferred to
the prescription software?
• Is the response of the hearing aid (measured in a coupler) the one prescribed for
this degree of hearing loss?
• Is there anything unusual, such as a blockage or a very loose fit, about the earmold?
Typically, the hearing aids are programmed by the
audiologist prior to the child/infant coming to the
office. The child and the supporting family then
come to the office and the hearing aids are fitted.
Soon after, the child is scheduled for “cortical” tests.
On a rare, very calm peaceful child, the cortical tests
can be done while the hearing aid is attached to a
separate computer used to program the hearing
aid. On the more typically active child the addition
of extra wires and computer make everything too
complicated. The gain/output controls can be adjusted to the desired level at which CAEP responses
are seen on the HEARLab monitor.
The example of Danny who has middle ear fluid
helps to reveal one of the major advantages of the
HEARLab tests. The aided cortical assessment includes all parts of the “pathway of hearing.” A test
signal is emitted from a speaker and the response is
observed in the auditory cortex. Many factors markedly impact the amount and quality of sound as it
passes from the outside world, through the hearing
aid and ear and into the brain. These include but are
not limited to:
• The filter in the hook of the BTE.
• The space (residual volume) between the
earmold and eardrum.
• The presence or absence of middle ear fluid.
• The integrity of the cochlear hair cells.
• The integrity of the ascending auditory
nerves.
We tend to think of the “pathway of hearing” as the
steps or stages in hearing from the outside world to
the brain. We need to remember that the auditory
cortex is part of the “pathway.” Brains mature rapidly after birth. The auditory cortex of infants with
normal hearing “organize” and “mature” rapidly in
response to sound stimulation. HEARLab tests incorporate many variables. HEARLab tests take us
directly to the cortical response that correlates well
with the perception of speech.
In Danny’s case, “poorer” hearing (an absent ABR)
and the presence of middle ear fluid caused Dr. Judy
to worry about “under” amplification.
In Suzy’s case Dr. Judy was worried about “over”
amplification. Suzy’s initial “aided” (ACA) cortical
tests looked good. Dr. Judy extended the tests using a much lower gain setting: Poor quality cortical
6
responses were seen. Dr. Judy interprets this data
to mean the auditory neural cortex system was not
adequately stimulated at the lower gain setting. It
seems the amplified sound was below Suzy’s auditory threshold levels, i.e., Suzy’s neural system was
not “detecting” the speech-sound used for the test.
So she rejected the possibility of using significantly
less gain.
When Dr. Judy started Danny’s test, poor quality
cortical responses were seen. Dr. Judy interpreted
this data to mean the initial fitting was “too weak,”
so she increased the gain of the hearing aids. Her
judgment was confirmed when she saw good cortical curves at the higher gain/output settings.
All of us understand that high power hearing aids
can be very harmful if fitted improperly. HEARLab
is an exciting new tool that helps you evaluate a
hearing aid fitting taking into account the extra efficiency of a tiny baby’s ear: it is more efficient in
the higher frequencies due to the smaller volume in
the ear canal. HEARLab also help us see whether
or not the presence of middle-ear fluid has attenuated the amplified sound to a level below threshold.
Gain/output can be indirectly managed using cortical tests. However, please read the following note
carefully:
Important Note!
If the gain/output of the hearing aid is much too
high, you see “typical” cortical responses. Nothing
in the data tells you that you are over amplifying
this person.
If the initial settings produce “good” quality cortical
responses, the patient’s auditory neural system is
responding to the amplification. The test results do
not tell you whether or not you are “over” amplifying the patient. Mindful of this fact, Dr. Judy Smith
was cautious with the gain level for Suzy.
If, you adjust the hearing aid to the prescribed gain
level and you run a cortical (ACA) test with the hearing aid on: If, the test conditions are good (quiet, the
baby is awake) and the cortical response is “poor or
absent:” You may need to review the hearing aid
prescription, the estimate of the hearing loss may
be too low (poor quality ABR results, etc). When the
initial test did not produce a “good” cortical curve,
Dr. Judy was prepared to increase the gain level for
Danny because all of his “hearing test” data pointed
to a profound mixed hearing loss.
Let’s turn our attention to the family of the patient.
Dealing with the Family
Suzy’s father was especially problematic. He is
an example on how the HEARLab test results can
help in counseling situations. Suzy’s father refuses
to accept the fact that Suzy has a hearing loss and
needs to wear hearing aids.
When Dr. Judy realized the family was having
problems, she scheduled them for a counseling session. She used the cortical responses to show the
father that Suzy had a hearing problem. The conversation went something like this:
“Suzy’s brain can show us whether or not she is
hearing.” Pointing to the cortical responses saved
in Suzy’s file Dr. Judy said, “Look, here is the response Suzy’s brain makes when she wears the
hearing aids. The neurons in her brain are telling
us, ‘I can hear! I can hear!’ When we remove the
hearing aids, these responses disappear! She cannot hear! When she wears the hearing aids she will
hear well. If we remove the hearing aids, she will
not hear, and she will not learn to talk. Her ability
to learn depends on her ability to hear. You need
to help make sure she wears the hearing aids and
they are working well.”
The visual response produced by HEARLab helps
parents see the result of the amplification. It is very
comforting to “see” that the amplification is working well. If families are having trouble accepting
that their child needs to wear amplification, it can
be very helpful to show them a graph that shows
no brain responses to speech sounds without hearing aids, and good brain responses with hearing
aids.
Let’s return to our story about Danny.
HEARLab measurements can be done easily on a
child that cooperates by being quiet. Danny is a
precious little baby who seems to like the TV when
he is in the office. Luckily today the TV has an almost hypnotic effect on him. Tympanograms are
obtained quickly. The amplification is lowered as
soon as the middle-ear fluid is gone from Danny’s
tiny ears and the aided cortical tests are repeated
to make sure he is still able to detect speech sounds
with the lower gain settings.
7
Some children like Danny will need additional
gain from the hearing aids when they go through
periods of middle ear fluid. HEARLab simplifies
the decision making process. Cortical tests confirm
the adequacy of the levels provided by a hearing
aid fitting. HEARLab gives the audiologist a tool
that can be used to study the needed gain.
predictable, and responses could be observed on
sleeping babies. Unfortunately, the very brief stimuli used for ABRs, and the requirement that the
baby be asleep, make them less suitable than cortical responses for evaluating aided performance.
Cortical responses also more completely measure
the auditory system.
We know through years of seeing children with
middle-ear fluid that the amount of a conductive
hearing loss can vary anywhere from a small (5-10
dB) loss to a large (35-45 dB) loss. Pneumo-otoscopy and tympanometric studies are used to diagnose middle ear fluid. However, it is impossible to
accurately estimate the size of the conductive component created by middle ear effusion without air
and bone conduction measurements.
NAL has built on the significant research done
internationally on cortical responses, and has in
particular focused on how to make the test as automated as possible, with as much interpretation
as possible done automatically by the equipment.
NAL has published some of their research on this
topic, with more in the publication process.
We have seen older children with middle-ear fluid
that have a 35 dB conductive hearing loss in the
low frequencies on “Monday” and have normal
hearing in two weeks when the fluid dissipates.
Hearing levels can change quickly when middleear fluid is eliminated. The tests done with HEARLab help to remove this complication. The effect of
middle-ear function is factored into the tests.
In Danny’s case, Dr. Judy’s initial gain recommendation was increased as a result of the initial cortical test. Dr. Judy’s fear of over amplifying Danny
was lessened because she understood that part of
his loss was conductive. When the middle-ear fluid
resolves and Danny has normal tympanograms,
his hearing aid gain is reduced and his cortical responses re-evaluated.
Who is behind this system?
NAL, The National Acoustic Laboratories, in Sydney Australia is a large sophisticated research facility. Audiologists and engineers at NAL set out with
the explicit aim of coming up with some method
for evaluating hearing aid fittings on babies diagnosed through universal newborn screening. Cortical responses quickly emerged as the most practical method.
If you have not followed the development of “cortical” tests you might believe these tests are only a
few years old. In fact, cortical responses have been
measured by researchers internationally since the
1960s. Their use for measuring hearing thresholds
was overtaken by ABR in the 1980s. ABRs had
the advantage that the response shape was more
Research, particularly that by North American
researchers Curtis Ponton and Anu Sharma, has
shown that cortical responses can tell us more than
just whether or not the sound is perceived. The
latency of the positive peak in infants and young
children is highly correlated with the amount of
exposure to sound that the child has had. For babies with normal hearing, the latency decreases
from over 200 ms at birth to around 120 ms by age
2 years.
Infants who do not receive adequate stimulation
until receiving a hearing aid or cochlear implant at,
say 12 months of age, initially have latency similar
to that of a newborn baby. Over the following year
of exposure to sound, the latency will decrease
by the same amount that normally occurs during
the first 12 months of life. The latency of the observed response thus tells us about the maturity of
the processing system within the cortex—maturity
that can develop only in the presence of auditory
stimulation. HEARLab includes a graphical representation of normal latency (given the age of the
child) so that the maturity of the response of the
individual child being measured can be assessed.
NAL has had a prototype unit in operation for a
few years. The current production unit, HEARLab
H 1000, incorporates many enhancements made
over this time.
NAL licensed the HEARLab unit to Frye Electronics in Oregon who manufactures the system.
Where is “cortical” testing done?
This paper discusses two modules: the “aided”
(ACA) test and the “threshold prediction” test
8
(CTE). When you measure cortical responses with
hearing aids, the usual input level is 65 dB SPL,
not very soft levels. No special consideration for
ambient noise needs to be made for the “aided”
(ACA) tests. These tests can be done in a wide variety of reasonably quiet listening situations. Family
members just need to be as quiet as possible when
sounds are being presented. Of course we cannot
stop infants from ‘talking’ during the testing but
we can try to engage them in activities that keep
noise to a minimum. Drinking from a baby training cup or eating soft foods (that don’t involve lots
of chewing) can be a good way to keep a little one
quiet and busy.
The primary consideration is the comfort of the patient. The infant must be relaxed and reasonably
still. Most “room noise” like people talking at soft
levels will be factored out by the averaging of the
responses that occurs within the testing system. It
is helpful but not necessary to do “aided” (ACA)
tests in a “sound booth.”
Neuromuscular noise produced when the child
moves or cries creates a disturbance in the responses, and when this disturbance is large enough to be
recognized by HEARLab as an unusual occurrence;
each epoch containing these high noise levels is
omitted from the average. Either way, neuromuscular noise lengthens the measurement process: In
the first case, a greater number of epochs must be
averaged to reduce the noise down to low levels, in
the second case; a greater number of stimuli must
be presented so that the required number of epochs
provide data that are accepted for processing. Consequently, it is important that the child be comfortable and quiet when the tests are done.
When “threshold prediction” tests (CTE) are done,
the ambient floor noise in the room and the natural muscular noise of the child all become critical
factors. As you attempt to test at lower and lower
levels, all noise (acoustic and neuromuscular) becomes critically important.
Some infants, like children with cerebral palsy, are
difficult or impossible to test because the “noise”
generated by muscular contractions cannot be factored out from the test data. Some medications prescribed to special needs babies (e.g. anti-epileptic
medicines) may also interfere with the response.
Here are some practical suggestions on testing.
Note: Lyndal Carter contributed many pages of
practical suggestions to this document. They are
included in the appendix. This following list was
abstracted from Suzanne Purdy et. al.(Chapter
Eight).
• The timing is very important. The infant
needs to be settled and awake.
• If the infant is too active, the recordings will
take too long and contain too much muscle
activity.
• Pause the tests when the infant gets too vocal.
• Have everyone monitor the infant’s “state.”
Infants fall asleep unexpectedly or can be
close to sleep with their eyes open.
• Electrode applications need to be fast, painless, and secure since the infant will be sitting up, awake on the caregiver’s lap.
• Tell the family not to use cream or hair conditioner on the infant’s head prior to coming to the appointment as the electrodes
can slide off.
• Drape electrodes away from the face and
hands so that they are less likely to be noticed and pulled at by the infant.
• Use a lanolin-based cream or an adhesive
remover wipe to assist in electrode removal
as it is much more pleasant for the family
if you don’t upset baby by just pulling off
the electrode sensor pads when the test is
finished.
• Make sure the hearing aids have new batteries for the test. You don’t want a battery
going dead during the test.
When is cortical testing done?
The overall process can be summarized as: (1)
screening, (2) diagnosis, (3) fitting, (4) cortical evaluation, (5) then long term follow-up and fine tuning.
1. Many infants receive a mandatory hearing screening at birth. Children who fail
the screening are referred to an audiologist
who specializes in pediatric evaluations.
2. ABR or ASSR, otoacoustic emission (OAEs),
and immittance audiometry (tympanometry and acoustic reflex testing): are often
used in these evaluations. After the initial
pediatric hearing evaluation, children who
9
have hearing problems are scheduled for a
pediatric hearing aid evaluation: impressions are made, and hearing aids are ordered.
3. The aids are programmed when they are
received from the factory: The audiologist
uses current guidelines (NAL or DSL i/o)
to prescribe the gain and output. The child
and the family are scheduled and the hearing aids are fitted. Considerable counseling
is done and the family is taught how to care
for the hearing aids.
4. The child is then scheduled for cortical testing. It is at this point that the HEARLab
becomes invaluable. The HEARLab testing sequence is ideal for infants who have
documented hearing loss and have been fitted with hearing aids. HEARLab helps to
confirm that the hearing aids have been adjusted to the prescription and that these settings are indeed meeting the child’s needs.
Cortical tests are also ideal for any child
that has developmental delay and cannot
respond reliably to behavioral testing, or
for that matter, people of any age who cannot reliably respond.
5. Follow up visits should be done regularly
to ensure that the child is receiving proper
amplification. Depending on the clinic, visits should be done every 2-4 weeks,
According to Harvey Dillon, “…cortical tests are
being positioned as a very informative evaluation
tool, rather than a fitting tool. It is anticipated that
further research and evaluation might show these
measurements to be so useful that this approach
will become a fitting tool rather than an evaluation
tool. For now clinicians should make their own
decisions on adjusting amplification, and what to
do if HEARLab says there is no response, even
though residual EEG noise has been reduced to a
level where a response ought to be visible, if one is
indeed present.”
Early intervention is critical.
Early detection and treatment of hearing is critical.
Here is a quote from the NAL annual report. “Newborn hearing screening programs aim to diagnose
hearing-impaired infants within the first few weeks
of life so that early intervention, including the fit-
ting of hearing aids, can be provided before the age
of six months. This is important because it has been
demonstrated that children diagnosed before the
age of 6 months develop significantly better speech
and language than those diagnosed after this period. Clearly, appropriate amplification is a crucial
factor…” Yoshinaga-Itano et al. (1998).
What about adults?
The tests discussed above (ACA) can be used to
evaluate hearing aid fittings on people of any age,
not just infants and babies. Later we discuss how
the cortical response changes as infants mature.
HEARLab also comes with CTE (Cortical Threshold Estimate) which is used to estimate hearing
thresholds. Tonal test signals can be presented
using insert earphones, or a bone vibrator. Insert
earphones reduce the ambient noise and allows
us to do “hearing tests” at lower intensity levels
without hearing test chambers. They are also more
accepted by small children than are supra-aural
headphones, and provide greater inter-aural attenuation, thus reducing the need for masking.
Clinical audiologists see a wide variety of difficultto-test patients. Some patients have physical or
emotional disabilities that prohibit them from cooperating fully with a typical hearing test. Other
patients are highly motivated to have “hearing
loss” for medical or legal reasons. Cortical tests are
ideal for difficult-to-test patients of any age where
objective “biologically based” responses are needed. These tests are especially helpful for patients
with exaggerated hearing loss or a patient involved
in a medical/legal disputes where “objective” data
are needed for legal reasons. For people with sensorineural hearing loss, cortical responses are usually visible within 10 dB of behavioral thresholds.
A Little Information about Brains.
If you have recent training in ABR you might want
to skip this section.
Nerve cells (neurons) produce electrical signals
throughout the body. There are huge numbers of
neurons in the brain. When activated, the neurons
in the brain make a lot of electricity. When a large
group of neurons fire in synchrony, they produce
enough electricity that we can detect and record
these responses using “surface” electrodes on the
scalp.
10
According to Teplan: The human brain starts maturing long before the baby is born and electrical
activity is seen around 17-23 weeks of prenatal development. At birth the number of neurons is huge
(10 to the 11th power). Adults have about 500 trillion synapses… “…the brain can be divided into
three sections: cerebrum, cerebellum, and brain
stem. The cerebrum consists of the left and right
hemisphere with a highly convoluted surface layer called cerebral cortex. The cortex is a dominant
part of the central nervous system. The cerebrum
contains centers for movement initiation, conscious
awareness of sensation (our italics), complex analysis, and expression of emotions and behavior. The
cerebellum coordinates voluntary movement of
muscles and balance maintaining. The brain stem
controls respiration, heart regulation, biorhythms,
neuro-hormone and hormone secretion. “The highest influence to EEG comes from the electric activity of
the cerebral cortex due to its surface position.” (Our
italics) (From M. Teplan, Fundamentals of EEG
Measurement, Measurement Science Review, Vol
2, Section 2, 2002).
“Evoked potentials, also called event-related potentials, are made by external stimuli (our test signals). Evoked potentials are extracted from sets of
recording by the digital averaging of epochs. An
epoch is a recoding period, time-locked to the repeated presentation of the test signal. The spontaneous background EEG fluctuations, which
are random relatively to the point in time when
the stimuli occurred, are averaged out, leaving
the event-related (evoked) brain potential. These
evoked electrical signals reflect only that activity
which is consistently associated with the stimulus
processing in a time-locked way. “The evoked potential thus reflects, with high temporal resolution, the patterns of neuronal activity evoked by a stimulus.” (Our
italics) (from Teplan, same reference as preceding
paragraph).
It is important to realize that while the averaging
process reduces the level of noise arising from all
sources other than neural circuits responding to
the stimuli, it never totally eliminates this noise.
HEARLab uses the variation between the individual epochs, combined with the number of epochs
included in the average, to estimate the residual
noise in the averaged response. It displays this as
a number in microVolts, and also uses this value
to control a set of traffic lights. The lights go green
when the residual noise is less than 3.4 microVolts.
NAL’s research has shown that if the residual noise
is less than 3.4 microVolts, cortical responses from
young children can almost always be seen when
the stimulus is 10 dB above behavioral threshold.
What Is a HEARLab system?
The patient’s point of view
There are two very different perspectives of HEARLab: the family and patient point of view, and the
Audiologist’s.
The HEARLab unit allows audiologists to create a picture of the electrical activity of the brain
when speech sounds are detected in the auditory
cortex. We see the evoked responses on the computer monitor and the test results are saved in the
computer’s data base. This is important information to professionals, but exciting emotionally-loaded
information to the parents of these children. Suzy and
Danny’s parents are delighted with the “graphs”
that show their babies can “hear” speech with their
hearing aids on.
We should give colored copies of these evoked responses to the patient’s family. Parents treasure items like
the “ultra sound” recording of the “baby” three
months after conception. We need “patient friendly” words to put on our graphs, something like,
“The first sounds Suzy heard!” Or, “Danny’s brain
saying, Yes, I can hear!”
Most “labs” have photo boards that are full of
“thank you” letters and photographs of the patient.
We suggest you make space for this board and ask
patients for thank you letters and photos to display
on the board. This board will become the “heart” of
HEARLab. Parents will be drawn to this “success”
board and the many photographs of beautiful infants.
There are three “parts” to HEARLab: the computer, the stimulus controller box, and the Electrode
Processor box.
The computer
Thirty years ago an ABR unit was expensive and
had limited processing power. Today’s computers are a thousand times better. Today’s computers have more memory and a higher processing
speed than main frame computers thirty years ago.
HEARLab presents sounds and records the tiny
electrical signals produced in the brain, while re-
11
ducing, by averaging, the background noise made
by the brain. Years ago, you needed to spend a lot
of money on a computer to do this. Today it can
be done with a normal computer. A “gig” of RAM
combined with the modern processing speed and
huge storage abilities of computers gives us all the
signal processing capabilities we need.
The HEARLab computer is a typical off-the-shelf
computer. It comes with the software installed.
One word of warning! Do not be tempted to install
any other software on this computer. HEARLab
is a “dedicated” computer designed for a single
purpose. The integrity of the system has to be protected so the processing capabilities can be maintained. The software is designed to “reject” the installation of additional programs. If you “bypass”
these built-in protections, the system will crash and
Frye will not furnish software support! So please,
make sure that all people who use the HEARLab
system understand it is a “dedicated” system and
must only run the HEARLab software.
The HEARLab software has a database used to
save all the patient information and the test results.
It also has “work” screens used during the testing
and reporting procedures. For more information
see the HEARLab NAL-ACA reference guide.
The Stimulus Controller (SC)
The computer sends commands to the Stimulus
Controller through the USB port connection. Signal
generation is also done here. The system does not
use a sound card found in many PCs. This SC unit
contains signal generation circuits, microphone
amplifiers and stimulus amplifiers.
SC has a number of connectors on its back panel
that are used to drive the sound field speaker(s),
the insert earphones and the bone oscillator. Other connectors that have other uses are seen on the
back panel. This unit has its own power supply
and must be plugged into a wall outlet. It supplies
power to the electrode processor box via a standard
multi-wire cable.
The Electrode Processor box
This box receives direct signals from the patient
and it contains a meticulously crafted system of
isolators that separates the main power circuitry
from the patient. The separation prevents the possibility of accidental electrical shock. The primary
amplifier is found in the SC unit which is electrically isolated from the patient.
The electrode processor box has amplifiers and connectors for electrodes. There are three electrodes
that will be used initially in the first software module: one active (non-inverting) electrode, one reference (inverting) electrode, and one ground electrode. However there are five spaces for future use
that can hold: three active electrodes, one reference
electrode and one ground electrode.
This box sends a tiny electrical signal to the electrodes and measures the resistance between the active and ground electrodes and also between the
reference and the ground electrodes. The cortical
signals are picked up by the difference between
the signals present at the three electrodes. The active electrode is placed at the top of the head, the
reference electrode at the mastoid, and the ground
electrode at the forehead.
This system uses the new “active” type of electrodes that produces much stronger signals right
at the head. The active and reference electrodes
have the active circuit element on a miniature circuit board which is encased in a rubber molded
housing. The main result of this is that the leads
between the head and the rest of the electronics are
much less prone to external interference than is the
case with normal passive electrodes.
The electrode is connected to a snap connector
that takes standard blue EKG sensor pads, preferably kid size. Electrodes are color coded: Black
is ground (forehead). Blue is reference (mastoid).
Yellow is active (vertex). Colors are marked on the
panel and on the electrodes themselves.
Before the start of each test, tiny electrical signals
are automatically delivered to the head by the
electrodes to measure the skin impedance interface. A color bar is shown for the reference and
test electrode: green indicates good impedance
(0-5k ohms), yellow is satisfactory (5-10k ohms),
marginal (10-15k ohms), and red (greater than 20
k ohms) indicates poor impedance. It usually helps
to reduce the impedance seen by the electrodes; the
green range is better than the other colors. Be sure
the skin is cleaned well before applying the electrodes.
Clinical Judgment & Expertise
12
Older brain stem units had dozens of controls; most
of them critically important to each measurement.
In contrast the new HEARLab unit looks very simple to use. All of the decisions that are made for filtering, stimulus timing, and data viewing are built
into the software. Much of the complexity has been
removed.
The software provides easy-to-read information
about the reliability of the response, and whether
there has been sufficient averaging to reach a conclusion that no response is present. This critical
information in presented with a green traffic light
that illuminates when adequate sample have been
obtained. Prior to this point in the test it is wrong to
conclude that the child is not “hearing” the sound.
The lack of a cortical response is meaningless until
adequate samples have been made. The response
may just be buried in the noise, i.e. more sample
may be needed to average the noise to a lower level.
Do not get the idea, however, that the need for
clinical judgment and professional expertise has
disappeared. It has not. The value of the data will
completely depend on the skill and experience of
the clinician. The clinician also has a vital role in
ensuring that testing conditions are optimized and
explaining the testing and results to families.
Cortical testing can currently be done without
HEARLab, however, the complexity of the system
requires the clinician to be an expert electrophysiologist and the cost of the system is substantial.
There are many variables that can alter the results
of a test; a few of these include: restlessness, sedation, age of the patient, sleepiness, attending or not
attending to the task, sweating from crying or hyper activity, electrical interference, developmental
delay, and nervous parents.
All patients present different challenges. The protocol must be tailor-made for each patient. The clinician must always be on their toes and be prepared
for anything that comes their way. The machine
appears to be very simple to use:
• It has few buttons to work with.
• It automatically indicates when a response
is present.
• It automatically indicates when residual
noise is low enough to take seriously the
absence of a response.
In spite of these technological improvements, the
responsibility for the audiologist is still considerable. The clinician must be experienced and be able
to read results and understand them and ensure
they tie in with other audiological findings. They
should also expect to get different results for each
patient and be prepared to handle them accordingly.
Suzy’s Follow-Up Visit
After Suzy had worn her hearing aids for about
a couple of months she was evaluated again with
HEARLab as part of her periodic review. Suzy had
acclimated to wearing the hearing aids and the initial responses were good. Now, Dr. Judy was concerned about the gain in the lower frequencies. Dr.
Judy had initially set the “ave.” gain to 35 dB and
the frequency response (the slope) to a 6 dB roll off.
This was done because Suzy’s diagnostic hearing
tests had indicated a moderate loss not a profound
impairment. There is a history of genetic hearing
loss in Suzy’s family, so Dr. Judy had wondered if
Suzy had a “cookie bite” audiogram or a more traditional “flat” loss. The slope of the frequency response was initially selected to be rather cautious.
Dr. Judy did not want to over amplify the lower
frequencies in case Suzy happened to have good
hearing in that zone.
During the follow-up test Dr. Judy noted that the
low frequency cortical response to the /m/ stimuli
was unclear. She increased the gain in the lower
frequencies by 10 dB and the tests were repeated.
This adjustment produced a marked improvement
in the cortical response for the low frequency emphasis speech stimulus /m/. Judy was delighted.
Danny’s Follow-Up Visits
Precious Danny was seen several times in an attempt to deal with three problems:
• middle ear fluid,
• feedback, and
• poorly defined cortical response for the
high-frequency emphasis speech sound
/t/.
A month after Danny was fitted with hearing aids
the middle ear fluid disappeared; then returned
later. Dr. Judy attempted to give Danny enough
gain so cortical responses could be visualized however increases in gain precipitated feedback. Sev-
13
eral sets of earmolds were tried. Finally, a different
set of hearing aids with improved feedback cancellation capability were used and the feedback problem was temporarily solved.
Whenever the middle-ear fluid returned, Dr. Judy
increased the gain in the low and mid frequencies,
not the high frequencies, as increases in high frequency gain often produced feedback (the effects
of middle-ear fluid are usually observed in the
lower frequencies, not higher).
While Danny’s case is challenging and frustrating,
HEARLab gives you a powerful tool that helps
you decide how aggressively to amplify various
frequency zones. Without cortical measurements,
a clinician may be tempted to make a large permanent reduction in high frequency gain to avoid
feedback. The HEARLab test helps to show us the
effects of changes we make to the hearing aid response. We can check to insure we have not compromised the response to the extent that we no
longer see cortical responses for all three speech
sounds.
In Danny’s case the response for the high-frequency
speech sound /t/ is often weak: the inbuilt statistical test indicates that a response is not reliably present. This may be because the high-frequency gain
in the hearing aid is turned down to avoid feedback. However, as an experiment, Dr. Judy Smith
covered Danny’s earmold with a small amount of
glycerin and she carefully inserted it: snugly. The
glycerin formed an acoustic seal and eliminated
feedback during the duration of the cortical tests.
Under this condition, the gain in the instrument
could be set to the “ideal” level, and cortical curves
could then be clearly seen for all of the three speech
sounds. This helps Dr. Judy decide that the change
to the hearing aid response will be OK as long as
well fitting molds are maintained for Danny.
Because Danny’s middle-ear fluid problem kept
returning, once the feedback problem had been resolved, Dr. Judy enabled the second memory in his
hearing aids so that he had one gain setting that
resulted in good cortical responses when his ears
were clear and another gain setting that gave cortical responses when he had flat tympanograms.
Tests like this help us clearly see our goal: adequate
amplification in all frequency zones that produces the
desired cortical responses. It is difficult to achieve this
level of amplification on a day-to-day basis. But at
least we have a clear view of our goal. Until proven different by additional tests when Danny is old
enough for reliable behavioral testing, adequate
amplification in all zones is important.
The three input levels: 55, 65, 75 dB SPL
The HEARLab ACA unit has three input level settings: 55, 65, and 75 dB SPL. Most “aided” (ACA)
testing will be done at 65 dB SPL to simulate the
presentation of speech at the normal level. Most
hearing aid fittings will be in the right “ballpark”
and the aided tests (ACA) will confirm the adequacy of the fitting.
If “good” responses are observed at 65 dB SPL, it
helps to reduce the input level to 55 dB SPL and
gather additional data. If good responses are seen
at 55 dB SPL you have no fear that the fitting has inadequate amplification: you may or may not have
excessive amplification. On the other hand if there
are no responses marked as significant at 65 dB
SPL increase the input level to 75 dB and continue
the tests. If “good” responses are seen at 75 dB you
may decide that the hearing aid fitting is possibly
“a little weak,” i.e., you need to increase the gain
a bit. If no responses are seen using the 75 dB SPL
input, despite the ‘noise’ traffic light being green,
you know that something is significantly wrong.
You need to check the hearing aid and be sure the
gain/output is at your recommended level. If so,
you probably need to increase the gain/output and
repeat the tests.
It is much easier to adjust the test level on the HEARLab system than to adjust the volume or gain of the
hearing aid. Most volume controls are deactivated
for these fittings. If you want to “drop” the speech
intensity level 10 dB and re-evaluate or “increase”
the intensity level 10 dB and re-evaluate, this can
be done easily with the software.
Testing Restless Children
Under ideal situations the patient is awake, calm
and quiet. You use all test signals to evoke cortical responses at several intensity levels. However,
some children and adults with special needs are
difficult to test. For example, they may be hyperactive or be prone to involuntary movements, making it difficult to get them to sit still and allow you
to properly place the electrodes and do the tests. In
these instances you may want to consider chang-
14
ing your techniques or consider alternate testing
procedures.
The speech-like signals, /m/, /g/, and /t/, used
by the HEARLab ACA test have a much longer duration than the “click” used for ABR tests. It takes
about a second for the brain’s response to each
sound to die down and for the brain to be strongly
responsive to the next sound. Consequently the
sounds are presented a little over one second apart.
This period from one sound to the next is called an
epoch. Roughly 100 error-free epochs are needed
before you see a “good” response—one that has a
high probability of detection. The test time for 100
epochs is about two minutes. If the patient is excessively active, this time will increase, either because
epochs are rejected, or because accepted epochs
contain more noise than when the child is in a quiet
relaxed state.
If the patient you are testing will not stay still for
the entire test, try to get as much data as you can
initially; then you can try again to get the remainder of the data later. For example, if you are checking a hearing aid fitting on a difficult-to-test patient
you might use one speech sound—and the corresponding cortical response—for your “primary”
testing goal. You gather data for this “primary”
target quickly: you can gather the rest of the data at
a later time. Later does not mean one month later,
it is possible to let the child nurse or take a short
nap and continue testing when they wake or after
they are fed.
If you get one highly reliable “bit” of excellent data,
you can often use this to insure you have a hearing
aid fitting that is providing some benefit.
The sound field speaker is usually located near the
patient (1 meter away) directly in front of the infant (at 0 degrees). A person who is distracting the
child can sit on a little chair positioned so that the
sound path from the speaker to the child’s hearing
aid microphone is not obstructed. The opposite ear
is occluded with the child’s own earmold and hearing aid switched off.
Some infants have tiny ear canals that become occluded easily. External ear tissue in infants is very
soft and pliable and can collapse easily. Before
cortical tests are given, it is a good idea to do an
otoscopic exam and check the ear canals for debris
and the middle-ears for fluid. Obviously, we do
not want the ear canal to be obstructed during the
test.
Earmolds and insert earphones hold the infant’s
ear canal open. If you attempt to measure an “unaided” response on an infant with the ACA test
using a sound field speaker; you need to be careful and insure the ear canal stays open when the
child is held by the parent. If ear canal collapse is
in question, switch to the CTE test and use insert
earphones.
Also, little heads need to be supported. Have the
parent hold the infant in their arms with the babies’
head held safely in their hand. Make sure that the
hearing aid microphone ports are not obstructed
when baby nestles back on Mom or Dad’s lap.
Cortical Responses
CAEP: cortical auditory evoked potentials
James W. Hall III, (New Handbook of Auditory
Evoked Responses) discusses the Auditory Late
Response (ALR):
• “The maximum response is typically obtained for moderate- (50-60 dB) verses high
intensity stimuli.
• The response is highly susceptible to alterations in state of arousal (sleep stages) and to
the effects of drugs, such as sedatives.
• Considerable intra- and inter- subject variability is common.
• There has been a recent resurgence of interest in clinical application of the response
with computed evoked potential topography techniques and sophistical stimulation
(e.g. speech stimuli).”
Large evoked cortical responses are seen in infants at about 200 milliseconds. All speech sounds
generate broadly similar CAEP responses. Close
15
inspection, however, indicates that the peak amplitudes and/or latencies vary systematically for
some speech sounds relative to others. HEARLab
indicates the probability that the response to each
of the sounds /m/, /g/ and /t/ is different from
each other. If any two of these responses are indicated as different from each other for an individual
patient, then you can take this as an extra reassurance that complex processes are underway in the
child’s brain in response to the speech sounds.
However, there is no need for concern if the responses, though present, are not marked as different from each other. Normal-hearing children can
differentiate all the sounds of speech, even though
the cortical responses are often indistinguishable
from each other.
Figure 2 and 3 (from Suzanne Purdy) show the
average curves for adults and infants. Notice the
differences between the adult and infant curves. It
takes considerable time for an infant’s brain to mature. The early peak and trough seen in adult data
become more apparent after the child is five to ten
years old.
+
1.25µV
---
-
-100
0
100
200
300
400
500
Time (ms)
Figure 2: Grand average adult (N=14) CAEP waveform for
the eight tonal and speech stimuli, recorded at Cz.
+
5µV
--
-
-100
0
100
200
Time (ms)
300
400
500
Figure 3: Grand average (N=20) infant cortical waveforms
recorded at Cz.
The “curves” for adults (Figure 2) usually have an
initial positive peak at about 50 msec. followed by
an obvious trough (negative peak) near 100 msec.
This peak and trough are poorly developed or
completely absent in infants. By adult years (over
the age of twenty) the dominant component is a
negativity (80 – 120 msec) that is preceded and following by positive components (i.e., P1 at 50 to 70
msec, and P2 at 150 – 200 msec) (Davis, 1965).
Notice that the vertical scale is 1.25 µV in the adult
graphic in contrast to 5 µV for the infants. The response of the infant is more robust.
Infants often have a broad, rounded peak at about
200 msec. This is the “curve” HEARLab seeks to
detect. When infants with normal hearing are tested, these curves are often seen. The intensity level
of the infant response is much higher than that of
the adult. To summarize:
Adult CAEPs:
• Have a positive peak near 50 milliseconds.
• And an obvious trough (negative peak)
near 100 milliseconds.
• And a large positive peak near 200 milliseconds.
• The end of the curve—the tail—seen between 400-500 msec. is near zero or it curves
upward slightly.
• The intensity of these responses is “subdued”, i.e., much smaller than those seen
with infants.
Infant CAEPs:
• Have a positive peak near 200 milliseconds
• The early peak and trough seen in adults
are usually missing.
• The end of the curve (the tail) seen between
400-500 ms is decidedly negative, curving lower on the graph. This “negativity is
easier to see if you overlay adult and infant
curves.
• The size of the response is robust, a lot
higher than an adult.
In normal hearing infants, the overall shape (morphology: amplitude and time at peak) of the curves
can change a little when different test signals are
used. These differences tell us that the cortex of
little babies is mature enough to tell the differences
between various speech sounds. This is very exciting because all speech and language development
16
depends on the differentiation abilities of the brain.
We want to adjust amplification to maximize these
differences.
According to Suzanne Purdy, “CAEPs in infants
evoked by some different speech phonemes differ
in latency and morphology. This indicated different underlying neural representations of speech
sounds and suggests that the information needed
to differentiate the stimuli is available to the listener.” However, this is really only reliable at the
group level. For the individual infant, therefore, a
lack of difference between the waveforms for two
different speech sounds is no cause for alarm.
Different stimuli create different responses. You
can see this in Figure? by comparing the highest
curve on the chart—the black line that represents
/t/—with the one peaking farthest to the right—
the green line that represents /m/. Low, mid, and
high pitched speech-like sounds are included in
the test to give us some indication regarding the
infant’s hearing in different frequency zones.
The “p” Value
A statistical analysis of each response is automatically calculated and displayed on the screen as the
“p” value for each sound. This is a sophisticated
statistical test (based on the Hotellings-t2 statistic)
that helps the clinician determine the likelihood
(probability) that the waveform observed was
caused by the patient detecting the test signal. This
statistic runs and updates itself in real time. Research on this “indicator” has shown it to be very
reliable; often better than the clinician running the
test. Do not rely solely on this analysis. The graphical display also shows the time region in which the
peak should fall if the child’s auditory system has
been exposed to sound for all of his/her life. If the
child spent significant time without adequate auditory stimulation (e.g. prior to receiving an effective hearing aid fitting or a cochlear implant), then
peak latencies longer than those in the shaded region may be observed. This is “normal” for a child
who has been deprived of sound, and the latencies
should decrease in the months following effective
aiding or implantation.
HEARLab also displays p values describing the
likelihood that the responses to different sounds
are different from each other. These are based on
another statistic – MANOVA, or multi-variate
analysis of variance.
Cortical vs. Brain Stem Tests
There is considerable published research on the attempt to use ABR in fitting hearing aids. Most publications point to the complications. Few authors
report success. The primary test signal for ABR is
the “click,” a sound that is very short in duration.
Clicks tend to saturate the amplifier and they are
incompatible with the processing delay of the digital hearing aid. A click cannot engage the compression circuits in a hearing aid in a realistic manner.
The cortical response is seen around 200 milliseconds in infants, and the stimulus used to elicit the
response is about 30 ms long. This stimulus duration is considerably longer than the attack time in
most (but not all) digital hearing aids, thus allowing the compression circuits to stabilize. For hearing aids with very slow compression, the child may
have lower audibility for connected speech than
he/she does for the isolated speech sounds used
within HearLab.
Cortical testing also has one other major advantage
in contrast with ABR. The electrical events we are
measuring are much larger and closer to our electrodes. On page three of his, “New Handbook of
Auditory Evoked Responses,” James W. Hall discusses the size (voltage) of the various test data:
ABR, Cortical (Auditory Late Responses), etc. He
says, “Activity arising from the higher regions of
the auditory system (the cerebral cortex) involves
hundreds of thousands, perhaps millions of brain
cells. The electrodes are also relatively close to the
sources of this activity. Therefore, these responses
tend to be somewhat larger in size (amplitude), on
the order of 5 to 10 µV. In contrast, activity generated by the ear, auditory nerve and brainstem,
which involves fewer neural units and may arise
at a further distance from the electrodes, may be
extremely small, on the order of 0.10 to 0.5 µV.”
Ok, changing the subject; let’s look at a couple of
older patients.
John Doe.
John Doe is a thirty-eight year old gentleman with
a pending Workman’s Compensation case. He
worked at a noisy fabrication plant for fourteen
years. He says the noise exposure hurt his ears and
he cannot hear. The “Workers Comp specialist” at
17
the plant sent him to North Shore Audiology. They
gave John several hearing tests and fitted him with
a pair of BTE hearing aids on trial.
John’s first hearing test showed a sloping 65-85 dB
hearing loss bilaterally for air conduction, the bone
conduction scores were near 50 dB. He was referred to an ENT for the “conductive” component,
but pneumo-otoscopy done with the microscope
and tympanomentry had been normal. The ENT
ruled out middle-ear involvement.
North Shore Audiology gave John a second hearing
test three months later and found about the same
65-85 dB hearing loss bilaterally for air conduction.
The bone conduction thresholds (air-bone gap) observed during the first hearing test “disappeared”
on the second audiometric evaluation. John and
the audiologist talked about this. John said maybe
he was responding to the vibration not the sound.
North Shore Audiology was uncomfortable with
this case and wanted cortical tests to see how accurate the audiogram and hearing aid fitting were.
They referred John to Dr. Judy Smith. John cooperated with the test. He had signed the release-ofinformation form and all other appropriate paperwork.
Dr. Judy talked with John. He gave her a copy of his
latest hearing test done by North Shore Audiology.
He said the hearing aids were working fine and he
could hear well when he wore them. Without the
hearing aids, he said, he could not hear. John had
done his homework.
Dr. Judy started the “threshold” (CTE: cortical
threshold estimate) tests with the Right ear using
insert earphones. She observed good cortical responses at 65 dB SPL; she lowered the input level
to 55 dB, then 45 dB and good responses were observed. She switched ears: Repeated the sequence:
Once again good responses were observed.
At this stage John was becoming uncomfortable
and he showed his irritation and he asked to “end
the tests.”
Dr. Judy told John that her test results were inconsistent with the hearing tests he had received
at North Shore Audiology. She scheduled him for
another visit. He never returned for the follow up
visit.
Dr. Judy sent North Shore Audiology a report that
said in effect, “My current tests indicate marked inconsistencies with your earlier hearing test, on the
basis of these tests we cannot estimate John’s level of hearing loss; however, because we obtained
good CAEPs at 45 dB SPL bilaterally unaided, we
believe his hearing is significantly better than the
levels indicated by the behavioral audiogram.”
North Shore Audiology forwarded Dr. Judy results
to all interested parties and waited for the results
of the additional studies but John never showed up
for these studies.
Tiffany
Tiffany is a sweet, very quiet, six year old girl living
in a foster care home. She has been in several foster care families the last couple of years. She is in a
special education classroom at school. She does not
speak, she responds when asked to do things by
teachers and her foster parents. The school audiologist, Margaret, gave her several hearing tests. Tiffany’s responses were inconsistent; she responded
at very high levels, e.g. 100 dB HTL. But, Margaret,
the audiologist, wrote on the hearing tests, “I don’t
trust these values! We need additional tests.”
Tiffany had been seen several times by a local pediatrician, Dr. Sanchez, who treated her for chronic
otitis media with constant purulent drainage. Dr.
Sanchez observed large perforations in both her
ear drums.
Tiffany has been fitted with BTE hearing aids
years before by another clinic, but Tiffany refused
to wear them. Dr. Sanchez referred Tiffany to Dr.
Judy Smith for a comprehensive evaluation.
Tiffany was a joy to work with and sat very quietly
while Dr. Judy did the tests. Dr. Judy started with
the “unaided” ACA test and she was delighted to
see good cortical responses without a hearing aid
at 65 dB SPL. Without question, Tiffany had some
hearing loss due her middle ear problems but her
hearing was much better than reported.
Dr. Judy switch to the CTE (cortical estimate assessment) and she presented “tones” to Tiffany using the bone vibrator. The responses were close to
normal hearing levels.
This information was shared with:
18
• Dr. Sanchez, the Pediatrician, who referred
Tiffany to an ENT to manage the middleear problem.
• The foster family.
• And Margaret, the school audiologist, who
referred Tiffany to counseling.
It turns out that Tiffany was a severely abused
child who lived in a “shell.” All reports about her
being “deaf” were inaccurate. She needed considerable time in counseling. A year later, her speech
and language skills were improving. Her ears are
being closely monitored by the ENT, and the discharge has stopped. Quickly, with the love and
support of teaches, counselors, and foster parents,
she learned to come out of her protective “shell.”
She was placed in a “normal” classroom.
Not all people have wonderful lives. Sweet Tiffany
had had a terrible life; one that tears your guts out
when you think about it. Tiffany was surviving the
only way she knew how. She acted “deaf” even
though she had pretty good hearing.
Tiffany was not able to respond accurately to the
Margaret’s tests. Margaret had been told by the
teachers that Tiffany could not “talk” so she assumed Tiffany had a severe hearing loss and believed incorrectly that Tiffany needed high power
hearing aids.
The beauty of “cortical” tests is that while not all
people can raise their hand when they hear a test
signal the response can be seen with only passive
cooperation from the patient. The cortical response
correlates highly with perceived speech. We may
prefer to use direct speech tests but these the patient to respond by repeating the words. But often
the patient cannot do this. Tiffany could not do this
when she was first seen by Margaret at school.
The story of HEARLab is more than a story about
a new machine. It is a story about bringing hearing to
infants, babies and older people. It is a beautiful story
about helping children like Tiffany. Today, Tiffany
can raise her hand when Margaret presents a pure
tone and asks her to indicate when she hears the
tone. But, there was a time when Tiffany could not
raise her hand, so we turned to her “cortical” responses that said in effect to us, “I can hear! I can
hear!”
A new chapter in the profession of Audiology
has just opened. Major advances in research by
scientists around the world, and particularly in
the USA, have shown us the importance of cortical responses in indicating not only the detection
of sounds but the maturity of the auditory system
in the cortex. NAL have added to this research by
focusing on the automated detection of responses
to speech sounds in infants. All of this research has
been brought together into modern computer technology and hardware. Frye Electronics Inc. have
produced this as an easy-to-use, clinically practical, reasonably priced, machine that can measure
and analyze evoked cortical responses. Now the
fun begins! We have the opportunity to do a better, more certain, job of bringing hearing to infants
and people who have problems with responding in
regular auditory testing.
The authors predict this new technology will generate so much excitement that it will not be long
before you see audiologists and their little patients
on many of the most watched TV talk shows.
NOTE: Lyndal intends to edit the following information and add as an Appendix to the final version
of the user manual (Initially prepared for report of
clinical evaluation by LC):
19
Appendix A: Practical aspects of CAEP testing with infants
The HEARLab system has been designed to make
objective audiological assessment as easy and efficient as possible. However, testing young children,
regardless of the hardware and software employed,
presents practical challenges.
The application of CAEPs to a pediatric population
may be relatively new to mainstream clinical practice, however, the general techniques and strategies used in other areas of pediatric audiology are
still highly relevant. Experienced clinicians will be
well aware of how to best manage the test environment, and will have developed many of their own
solutions to overcoming the issues that inevitably
arise when working with infants.
The clinical validation trials of the new HEARLab
system involved repeated CAEP measures, conducted in a controlled and systematic way. This
provided a valuable opportunity to investigate different approaches and test techniques, and to consider their relative merits, without the usual restrictions of a routine clinical assessment appointment,
in which a diagnostic outcome must be achieved
within a limited time frame.
will be suitable activities/supervision away
from the test room. The parent should be free
to focus their attention on the child having the
test.
†† Ask the parent/care giver to bring food, drinks,
or dummies (“pacifiers”) for the child, to the
appointment. Some favorite toys, that are suitable as quiet distracters, can also be useful in
making the child feel more secure in the test environment. DVDs that the child enjoys can also
provide familiarity and useful distraction.
†† Suggest that the child be dressed for the as-
sessment in layers of clothing that can be easily
removed. Electrode contact can be comprised
if the child becomes overheated and “sweaty”
and it may be necessary to remove clothing
to cool them down. It is better not to have to
pull clothes over the child’s head once they are
“wired up”.
†† Ask parents/care giver to ring and postpone
the appointment if their child is unwell, particularly if the child has a temperature. A restless
and irritable state is not conducive to quality
recordings.
Some general suggestions, based on observations
made during the study, are summarized in this
Appendix. It is hoped that this information may
provide useful guidance, particularly for clinicians with less experience in pediatric audiology,
or those who are new to using electrophysiological
assessment techniques with young children.
Test environment
Before the appointment
†† Make the test environment “child friendly”. For
†† Prior to the appointment, provide parents/care
giver with information (written and verbal)
about the procedure. This will reduce the time
spent in explanation at the assessment.
†† When arranging the appointment, ask about
the child’s routines. Try to book the test at a
time of day when the child is likely to be in a
“good” mood, and less likely to be overtired
and irritable. Allow plenty of time so that appointment is not rushed, and it is possible to
take breaks if needed.
†† Check whether the parent/care giver intends to
bring sibling/s to the appointment. If it is necessary for siblings to attend, make sure there
†† Call the parent/care giver to confirm the ap-
pointment the day before, and take the opportunity to check whether they have any questions or concerns they would like to discuss.
example, decorate the test booth and surrounding areas using items such as mobiles, displays
of soft toys (out of the child’s reach), and fabric
motifs. Avoid hard reflective objects that will
cause sound reflections. Minimize technical
“clutter”. Keep wires out of view and laboratory supplies in drawers. Children, particularly
if they have undergone medical treatment or
hospitalization in the past, may associate such
items with unpleasant procedures.
†† Keep the test area clean and tidy. A plastic
backed ‘Draped’ sheet (available from medical suppliers) is a useful surface for arranging
preparation materials, and for wrapping used
electrodes, cotton tips etc. afterwards for disposal.
20
†† Provide a chair that is as large enough for the
child to sit comfortably, either on their parent’s lap or beside them, during testing. Some
children become irritable if they feel overly
restrained or restricted in their movements.
Where possible, try to let them settle into a position which they prefer. A recliner chair works
well as it can be used for adult testing as well.
†† Use washable covers on the chair (eg, bath
Preparation for testing
†† Children generally have a short attention span,
and their mood and state can change quickly.
Have all test equipment switched on, checked
and calibrated before the child arrives. Have
the recording system software open and ready
on the impedance check screen to avoid unnecessary delays.
toweling) and change between assessments to
maintain hygiene. This makes “dribbles” and
food spills easier to contain and reduces parents inclination to perform immediate “clean
ups”, which can disrupt the testing. Have tissues or baby wipes at hand if needed.
†† If testing is with hearing aids on, change the
†† Have a container on hand to collect items that
require cleaning, according to infection control guidelines (eg, toys that have been in the
child’s mouth).
volved in the test. The child is more likely to be
relaxed and cooperative during the assessment
if their parent is confident and relaxed about
what is happening.
†† Some younger infants may be comfortable in a
†† Make the parent comfortable. Providing a hot
rocker/”Fraser chair”, but don’t rock the baby
during recordings. “Bounces” can be evident
in recordings and rocking can make a child
sleepy.
batteries and check the devices on arrival. Having another staff member do this while you are
interviewing the parent, this will minimize delay in commencing testing.
†† Ensure that the parent understands what is in-
drink or glass of water can help put parents at
ease. It is a good idea to have a safe place to
put a drink beside a parent so that it cannot be
knocked onto baby or electrical equipment.
†† Fluorescent lighting can cause problems of †† Ensure the parent feels in control of the situaelectrical interference. Incandescent lighting
should be used in preference. As well as providing a technical advantage, incandescent
lamps can create a pleasant, relaxed ambience
for children and parents. Novelty lamps (eg,
artificial fish tanks, “lava” lamps), as well as
providing illumination, can provide the child
with visual distraction.
†† It is important that the tester is able to moni-
tor the test environment and the baby’s state
throughout the test. A strategically placed
video camera can be extremely helpful if the
arrangement of the test booth makes it difficult
to maintain a clear view. A video camera that
works well in low light is recommended.
†† Where the tester is in a separate observation
room, an audio monitor is essential in monitoring the ambient noise level in the test environment, to ensure consistent stimulus delivery,
and is also useful in communicating with the
distracter.
tion. Seek their advice about the child’s preferences, and the best strategies for preparing
them for testing. Respect their opinion and follow their suggestions.
†† Try to build some rapport with the child. A
little physical interaction with the child while
you are interviewing the parent (eg, patting
or stroking their arm or head) can be useful in
gauging how they will react to the preparation
for electrode sites, and may possibly help the
child accept it more readily.
†† Make sure the child’s physical needs (eg, diaper changing) are attended to before starting
preparation for the electrode placement.
†† Have any items that might be needed (eg,
bottles, food, toys) close at hand, in order to
minimize noise and disruption during testing.
Check with parents to see if bottles need warming or food needs preparation before you start.
†† Attempt otoscopy and tympanometry first if
indicated, but don’t persevere if it causes the
child to become too active or distressed.
21
†† Ask parents to switch off mobile ‘phones or
pagers as these may cause distraction to the
child if they ring during testing, and may also
be a potential source of electrical interference.
†† Some children will be reassured by watching
the preparation in a mirror, but this may make
others more apprehensive.
†† Television can be a good distraction during
challenging part of the test procedure. Approach the preparation confidently but not too
forcefully. Smile, and talk to the child reassuringly.
electrode placement. Use a range of children’s
DVDs with lots of color and movement. If the
child becomes interested, the DVD can be left
playing (with the sound muted) when testing
starts. This is the time to use your noisy, fun
toys, before the real testing begins! Make sure
you put these out of sight before you start.
†† Start the preparation in a position where the
†† Wherever possible have a trained distracter
Preparation of electrode sites
†† Attaching the electrodes is potentially the most
child is comfortable and not overly restrained.
For example, try starting while an older infant
is playing on the floor, or at a child’s table and
chair.
†† Try not to physically “stand over” the child
while doing skin preparation. Working from
behind the child may be a good option. For your
own health and safety, try to maintain a posture that is ergonomic and doesn’t place strain
on your back or neck. Sitting on the floor beside
baby whilst chatting to Mom or Dad and gently
rubbing the skin can be a very non-threatening
way to get the skin prepared.
†† Electrode sites are generally prepared by abrad-
ing with a cotton applicator ‘bud’ and a medical gel intended for the purpose. Rub firmly
and vigorously enough to cause a slight redness on the skin surface, but not so hard that
the child becomes obviously distressed by the
sensation. Rubbing gently, but firmly, back and
forth works better than “dabbing” at the skin.
Rub on the back of Mom or Dad’s hand first so
they know what it feels like.
†† Work as quickly as possible and minimize the
number of physical contacts with the child.
Don’t fuss or “overdo” it, but be mindful that
it is better to prepare skin thoroughly than to
have to repeat the whole process.
†† If the child needs reassurance about the prepa-
ration, modeling the procedure (eg, by rubbing
the forehead of the parent or a doll with a cotton bud, and sticking on an electrode) can be
helpful. Try letting an older child have a “turn”
at putting an electrode on a toy or on their parent.
(in addition to the parent), to interact with
the child during electrode placement, as well
as during testing. Toys that involve some fine
motor manipulation (eg, block stacking, button
pressing) can help keep hands away from the
electrode sites.
†† Cleaning skin with an isopropyl alcohol prep
swab after abrading is sometimes recommended, and may improve contact, but it can make
the electrode stick very firmly and make it difficult to remove. It can also feel “stingy” – try
it on yourself if you don’t believe this! Preparation with an alcohol wipe is not needed and
is not recommended for the delicate skin of infants.
Optimizing and maintaining electrode contact
†† Use a liberal amount of electrode paste under
the vertex electrode even if a disposable electrode, that already contains conductive gel, is
used.
†† If using disposable electrodes, a spot of double
sided tape (the type used for retaining hearing
aids) on the underside of the plastic tab of the
electrode stud, can give a firmer hold, particularly for mastoid or forehead sites (ie, where
the skin is free of hair).
†† A headband is very helpful in keeping the elec-
trodes in place, particularly at the vertex, but
some children are less accepting of wearing a
headband than others.
†† To make wearing a headband more appealing
to the child, choose colorful, soft and stretchy
materials. Give older children a choice of colors
or designs (eg, have a selection of different mo-
22
tifs sewn on a selection of headbands). Having
a choice of “girls” or ‘”boys’” styles can be important to the child, and sometimes also to the
parent. Use fabrics that are easy to wash and
dry after use.
†† Dividing the top of the headband before use (by
cutting a slit a few inches long across its centre)
allows a section of fabric to be stretched forward to hold the forehead electrode in place.
†† Passing the leads under the headband can help
reduce pulling and strain on the electrode site
during the test.
†† Elastic bandage, particularly of material that al-
lows the ends to adhere without pins or tape
(eg, “peg” bandage), can be a reasonable alternative, but tends to be more “fiddly” to put on
than a headband. If an inexpensive type is chosen it can also be disposed of after use which
can be an advantage.
†† Micropore tape (either on its own or with a
headband) is appropriate for keeping re-usable
type electrodes attached.
†† Once the electrodes are attached, try to drape
the leads behind the child, avoiding contact
with the child’s face or neck. If the child can
feel them, they will be more inclined to pull at
them. Try to keep the leads away from clothing (eg, don’t let them become tangled in bibs
or collars).
†† Loosely taping the leads to the back of the
child’s clothing (using micropore tape) may be
helpful, but ensure they are not taped so tightly
that the electrode leads are pulled off the head
if the child suddenly leans forward.
†† Make sure that if the parent is holding the
child, that the electrode leads are not cramped
or pulled under the parent’s arm. Try directing
the leads up and over the parent’s shoulder
†† Avoid the child leaning back onto the electrode
leads, or making sudden large movements,
such as lunging forward. Strategic use of distraction toys can help in this respect.
†† If the child starts to touch the leads or elec-
trodes don’t over-react (eg, grab suddenly at
the child’s hands). In preference, try to distract
the child by offering them an alternative item
to play with.
†† Once the electrodes are in place, avoid touch-
ing them unless really necessary (ie, they are
obviously slipping/becoming unstuck). Drawing child’s attention to them will often result in
renewed efforts to remove them.
†† Don’t let the child overheat. This can result in
electrodes lifting and the reject rate increasing.
If the child gets very restless it can be better to
suspend testing than let it proceed until they
are “hot and bothered”.
Distraction techniques
†† The distracter is best seated in a comfortable
position at, or below, the child’s eye level. Care
must be taken to maintain an appropriate position in relation to the speaker if the stimulus is
presented free-field.
†† Have a wide selection of age appropriate toys
that are not too noisy. Keep toys and other distraction aids in easy reach, to minimize noise
and disruption as the test proceeds. Choose
toys that can be cleaned according to infection
control procedures. Toys that can’t be cleaned
(eg, soft toys) should be kept out of the child’s
reach.
†† For very young infants, the main aim of distrac-
tion is to keep them alert and awake (they do
not have the motor skills to pull at the leads or
electrodes). Mobiles, hand and finger puppets
are all useful items. Visual novelties (eg, toys
with lights and motion) can be excellent. Some
mechanized toys are too noisy to use while the
testing is in progress, but can be good to use
during breaks, in order to regain a child’s interest and increase alertness.
†† Toys that are used for VROA distraction are
generally suitable for children in the 7- 24
month age group. Examples include; stacking
plastic rings or cups, farm animals, large counting and threading beads, puzzles, colorful
teething rings and so on. Items that keep hands
occupied are ideal for children old enough to
manage them.
†† For younger children or those who are not
developmentally ready to “play” with items,
23
toys with texture (eg, spiky plastic balls, plastic
animals etc) can be interesting for the child to
touch or mouth (make sure you put them in the
“washing” container after use).
†† Action toys (eg, water-wheels, small spinning
tops, “pecking” birds, ‘”wobbly” animals, clear
plastic balls that contain moving toys) can be
useful as long as they are not too noisy. If children are allowed to hold the items they should
not contain small parts that may be a choking
hazard, and they must be easy to clean. Watch
that water-filled toys don’t leak.
†† For older infants (around 24 months and over)
try coloring-in books with big colorful crayons, play-dough (if past eating it, or well supervised), paper with stickers or self-inking
stamps.
†† Books can be good for children of all ages. Heavy
cardboard books of various shapes, or with
flaps/pop-up features, can provide “hands on”
activity. Books made of plastic (ie, intended for
bath time) can be ideal for very young children
as they are easy to clean if mouthed.
†† Almost every child enjoys watching bubbles
being blown. The small bottles used for parties are inexpensive and easy to use. Avoid
“sticky” bubbles that are designed not to burst.
They tend to leave messy residue in the test environment.
†† Eating and drinking are excellent distracters.
Good choices include baby bottles or infant
sipping cups, and soft foods such as banana,
custard, fruit gel, or sultanas. Avoid hard foods
(eg, crunchy crackers) or large pieces of food of
that require a lot of chewing, as the resulting
noise and jaw movements can affect recordings.
†† Breastfeeding is good for calming infants, but
often can induce sleep. If the baby must feed
during the assessment, watch very carefully and rouse the child gently if they begin to
doze off, or appear their eyes start to appear
“unfocussed”. Be prepared to pause testing if
the child’s state becomes inappropriate. Sometimes a short break to have a feed can gibe baby
a boost to keep them going for a bit more testing.
†† Avoid distraction activities that are too stimu-
lating or that encourage increased vocalization, for example, physically vigorous play,
or games/gestures that encourage the child to
answer questions or name objects. Parents may
sometimes need some guidance about activities
that are inappropriate in this respect.
†† If the child is content and quiet it can be best
for the distracter to sit quietly or withdraw and
leave them to their own devices. If the child
is unsettled sometimes it can help for the distracter to get right out of their view and let the
parent try to settle the child before proceeding.
†† Try to end the test on a ‘”happy note”, rather
than persisting until the child (and potentially
the parent) is distressed. This is especially important if the child will have to attend on another occasion.
24
Appendix B: Translation of research to HEARLab, a clinically practical instrument.
Research finding or practical problem addressed
Clinical implementation or solution
Cortical response shapes vary widely from
person to person
An automated statistical detection method
that does not require any apriori assumptions
about the “normal” response shape, that combines information from different parts of the
response, and that shows the probability of
the response, being evoked by the stimulus.
Response latency varies with the maturity of
the auditory system.
Normative data built into the software display
the range of latencies expected for babies (or
children or adults) for auditor system maturity appropriate to the age of the person being
tested.
Inadequate averaging, or excessive movement by the children, creates variability in
the age waveform that can mask the underlying cortical response.
Residual noise in the waveform is continuously calculated, and traffic light indicators
show when the residual noise is sufficiently
low that the absence of a cortical response can
be interpreted as indicating inadequate sensation level.
Predicting audibility of speech sounds from
pure tones measurements of threshold and
hearing aid performance is very complex
and error prone.
Use speech sounds as a stimuli, and automatically correct for the response of the loudspeaker and the room transmission characteristics.
ABR is best measured when the baby is
asleep, but babies are not always asleep
when needed. Cortical responses are best
measured when the baby is awake, making
the techniques complementary.
Incorporation of pure tones, delivered through
insert earphones, enable hearing thresholds
to be estimated for an awake baby (or adult)
through measurement of cortical response to
tonal stimuli, still with automatic statistical
detection of responses.
Electrical interference from the room adds to
measurement noise and can make it slow or
impossible to measure a valid response.
Miniature pre-amplifiers on the active and
reference electrode connectors greatly reduce
capacitive pickup of interference within the
electrode leads.
26
FRYE ELECTRONICS, INC. P.O. Box 23391 • Tigard, OR 97281-3391 • (800) 547-8209 • Fax: (503) 639-0128 • www.frye.com • e-mail: support@frye.com