National Acoustic Laboratories

1
Infant Hearing Aid Evaluation Using Cortical
Auditory Evoked Potentials
Suzanne C Purdy, Harvey Dillon, Mridula Sharma,
Richard K Katsch, Lydia M Storey, Teresa Y Ching
National Acoustic Laboratories Cooperative
Research Centre for Cochlear Implant Hearing
Aid Innovation, Speech Science, The University
of Auckland, New Zealand
2
Motivation for investigating aided auditory
evoked potentials

• Objective hearing aid validation techniques
  needed for young infants and difficult-to-test
  children

Dillon, NAL
3
Cortical responses in children with normal hearing
Dillon, NAL
4
Research Goals
For infants with normal hearing

• to characterise cortical responses to
  supra-threshold speech and tonal stimuli
• to determine whether responses are sensitive to
  differences in stimulus characteristics

Dillon, NAL
5
Methodology

• N 42 normal hearing
• 3-7 months old
• Awake
• Speech and tonal stimuli via loudspeaker
• Recording electrodes at C3, Cz, C4
• 0.1-30 Hz filter, -100 to 500 ms time window

Dillon, NAL
6
CAEP in Infants with Normal Hearing

• Experiment 1 Stimulus Effects
• N20 (12 F, 8 M)
• 3-7 months (mean 5.0)
• /mae/, /tae/, /gae/, 500 Hz, 2 kHz

Dillon, NAL
7
Grand average infant cortical responses recorded
at Cz electrode
Dillon, NAL
8
P1 amplitude
Dillon, NAL
9
P1 latency
Dillon, NAL
10

• But what about each individual subject?

Dillon, NAL
11
Multivariate Analysis of Variance
Voltage
Time

• Divide each record into 50 ms time bins
• Average data points within each time bin
• Use these averages as variables in MANOVA
  analysis
• MANOVA finds the combination of variables that
  best distinguishes two or more stimuli
• Result is probability of two stimuli coming from
  different distributions

Dillon, NAL
12
Number of infants (N20) with significantly
different cortical responses to pairs of stimuli
Based on MANOVA at Cz, 101 to 500 ms post-onset,
in eight bins each 50 ms
Dillon, NAL
13
Number of infants (N20) with significantly
different cortical responses to pairs of stimuli
Based on MANOVA at C3, 101 to 500 ms post-onset,
in eight bins each 50 ms
Dillon, NAL
14
Dendrogram Average Intra-subject dissimilarity of
the cortical responses
Dillon, NAL
15
CAEP in Infants with Normal Hearing

• Experiment 1 Conclusions
• responses reliably present
• significant latency and amplitude differences
  across stimuli
• Significant differences in waveshape for mae and
  tae for 19 out of 20 children (20 out of 20 at C3
  instead of Cz)

Dillon, NAL
16
CAEP in Infants with Normal Hearing

• Experiment 2. Stimulus Duration
• N11 (6 F, 5 M), 3-7 months (mean 5.1)
• /tae/ and /mae/ 141 versus 79 ms
• duration not significant for latency or amplitude

• Experiment 3. Inter-stimulus Interval
• N10 (5 F, 6 M), 3-6 months (mean 4.4)
• interstimulus intervals 750, 1125, 1500 ms
• trend for greater response amplitudes with longer
  ISI (p.052), no change in latencies
• medium (1125 ms) ISI used in subsequent studies

Dillon, NAL
17
Cortical responses in children with hearing
impairment
Dillon, NAL
18
Research Questions

• Aided CAEP in hearing-impaired infants and
  children
• recordable in infants and children with
  moderate-profound loss?
• consistent with hearing loss and hearing aid
  characteristics?
• show consistent changes with altered hearing aid
  settings?

Dillon, NAL
19
Hearing Impaired Infants Children

• N40 aged 2 months to 18 years (31 right, 33
  left ears)
• variety of etiologies
• one conductive ear, others sensorineural
• 40 had other disabilities such as autism or
  developmental delay
• stimulus 65, 75, or 85 dB SPL via loudspeaker at
  45º azimuth
• hearing aids fitted using NAL-NL1 formula

Dillon, NAL
20
Range of hearing impairments (4-frequency pure
tone average) versus childs age
Dillon, NAL
21
of ears tested with aided cortical response to
65 dB SPL speech stimulus
Number per category
13 18 8 8 10 7 13
20 7
Dillon, NAL
22
Enhanced cortical P1 response to /gae/ with
increased hearing aid gain where initially there
was no response when hearing aid was set
conservatively relative to the measured
tone-burst ABR thresholds
Dillon, NAL
23
Are /tae/ /mae/ cortical responses different in
hearing impaired children?

• 19 subjects (23 ears)
• 8 infants 6-20 months, 11 children 4-12 years
• Mild to profound (mostly severe) hearing loss

• Conclusion
• 60 had different responses based on individual
  ANOVA

Dillon, NAL
24
Are cortical responses sensitive to hearing aid
fine tuning changes?Subjects N9 aided
children, 6-12 years, moderate-severe hearing loss
Dillon, NAL
25
Effect of filter response variation
-6, -3, 0, 3, 6 dB/oct
-3, 0, 3, dB/oct
Dillon, NAL
26
Cortical auditory evoked responses

• reliably present in young infants
• sensitive to changes in stimulus characteristics
  within individual normal-hearing infants
• present in most aided infants and children with
  moderate-profound hearing loss
• consistent with and sensitive to changes in
  hearing aid characteristics in many hearing
  impaired children
• further work underway to determine impact of
  hearing aid prescription on both cortical
  responses and on functioning of infant

Dillon, NAL
27
Our next goal

• Find the hearing aid response that best enables
  differentiated cortical responses to different
  speech sounds
• Find the hearing aid response that is best for
  the child in real life
• Compare!

Dillon, NAL
28
Thanks for listening

• Thank you
• Subjects and families
• Gladesville Baby Health Centre
• Australian Hearing Centres
• Royal Institute for Deaf and Blind Children
• Sydney Cochlear Implant Centre
• The Deafness Centre, Childrens Hospital at
  Westmead