Panel Session on Frequency Lowering Technology
Panel chair:
Panel members:
Susan Scollie
Jace Wolfe
Andrea Bohnert
Danielle Glista
Michael Boretzki
Clinical rationale for frequency lowering:
• Bandwidth matters:
– Children need access to the high frequency sounds of speech, to understand and monitor:
• See: Moeller et al, 2007, a review article by Stelmachowicz et al (2004), previous proceedings from this meeting, and this conference’s presentation by Andrea Pittman.
• But:
If audibility cannot be provided via the available bandwidth and gain/output, is it beneficial to lower the cues to an audible frequency range? A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Bandwidth limitations in current‐era devices:
A severe sloping loss.
Hearing aid responses and targets for speech (input at 65 dB SPL)
A Sound Foundation Through Early Amplification 2010
Both fall below threshold above 2000 Hz (speech peaks are audible to about 2500 Hz –
not shown).
Hearing Aids A and B are both modern devices, and are at maximum settings in this region.
Panel Session on Frequency Lowering
Frequency Lowering (FL): two types
• Frequency Compression (FC)
– E.g.: Phonak SoundRecover
• Frequency Transposition (FT)
– E.g.: Widex
AudibilityExtender
• Review: Simpson (2009), Trends in Amplification
A Sound Foundation Through Early Amplification 2010
Below cutoff
Above
cutoff
Frequency (Hz)
Below cutoff
Above
cutoff
Frequency (Hz)
Panel Session on Frequency Lowering
Frequency Lowering (FL): two types
• Frequency Compression (FC)
– E.g.: Phonak SoundRecover
• Frequency Transposition (FT)
– E.g.: Widex
AudibilityExtender
• Review: Simpson (2009), Trends in Amplification
A Sound Foundation Through Early Amplification 2010
Below cutoff
Frequency (Hz)
Below cutoff
Frequency (Hz)
Panel Session on Frequency Lowering
Some studies of frequency lowering:
(other studies have been done in adults, plus other non‐peer reviewed in kids)
• FC in children:
– Glista et al., 2009a
• FC improves detection/recognition (group vs individual)
• Significant candidacy factors (hearing loss, age group)
• FT (various types) in children:
– Miller‐Hansen et al, 2003, MacArdle et al, 2001
• FT improves detection & recognition (group vs individual)
– Auriemmo et al, 2009
• FT + training improved consonant recognition (other outcomes)
– Smith et al, 2009
• FT improved consonant recognition
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Efficacy? Experimental design factors…
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Fitting Method (pediatric):
1) Provide more audibility of high frequency cues than is possible with a well‐fitted device. The frequency response is based on DSL5 child. 2) We verify using measures that show us audibility of specific high frequency speech bands (see Glista & Scollie, AudiologyOnline 2009)
UWO, and Hearts for Hearing, and University Mainz are all following this method.
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Verifit “Speech Bands” with/without:
With FC, 6300 Hz above threshold:
This speech signal has been
notch filtered here.
A high frequency band is left
(6300 Hz in this example.)
Without FC, the 6300 Hz band is below threshold.
The notch lets you observe
lowering of the high
frequency band.
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Today’s panel:
• Jace Wolfe, Oklahoma, USA: Evaluation of FC for moderate hearing losses.
• Andrea Bonhert, Mainz, Germany: Evaluation of FC for moderate to profound losses.
• Danielle Glista, London, Canada: Do children need an acclimatization period after FC fitting? • Michael Boretzki, Staefa, Switerland: Future directions in evaluating SoundRecover.
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Evaluation of frequency compression for moderate hearing losses
Audiologists
Jace Wolfe, Ph.D., CCC‐A Heather Kasulis, AuD, CCC‐A
Kimberly Fox, AuD., CCC‐A Brooke Shoemaker, Au.D., CCC‐A
Speech‐Language Pathologists
Joanna T. Smith, M.S., CCC‐SLP, LSLS Cert. AVT
Teresa H. Caraway, Ph.D., CCC‐SLP, LSLS Cert. AVT
Wendy DeMoss, M.S., CCC‐SLP, LSLS Cert. AVT
Tamara Elder, M.S. CCC‐SLP, LSLS Cert. AVT Darcy Stowe, M.S. CCC‐SLP, LSLS Cert. AVT Natalie O’Halloran, M.S. CCC‐SLP, LSLS Cert. AVT Lindsay Steuart, M.S., CCC‐SLP, LSLS Cert. AVT
Krissa Cummins, M.S., CCC‐SLP
Support Staff
Kris Taylor Pati Burns Kristy Murphy
June Cashion Susan LeFleuer
Megan Miller
A Sound Foundation Through Early Amplification 2010
Kerri Brumley
Sherry Edwards
Katie Edwards
Panel Session on Frequency Lowering
What about children with moderate hearing loss?
• Stelmachowicz and colleagues (2000, 2001, 2002, 2004) have shown that children with moderate to moderately severe SNHL need a wider bandwidth for optimal speech recognition.
• Young children with moderate to moderately severe SNHL show delays in fricative production (Moeller et al., 2007; Stelmachowicz et al, 2004).
• Children with access to high‐frequency information (i.e., >4K Hz) demonstrate better short‐term word learning (Pittman, 2008).
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Study Objectives
• Does non‐linear frequency compression (SoundRecover in the Nios hearing aid) improve speech recognition for children with moderate SNHL?
• Does non‐linear frequency compression (SoundRecover in the Nios hearing aid) improve speech production for children with moderate SNHL?
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Methods
• 18 children with moderate to moderately severe high‐frequency SNHL fitted with Phonak Nios micro‐sized behind‐the‐ear hearing aids.
• Today, we will be reporting on results for 15 children. A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Mean Audiogram
N = 15
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Subject Characteristics
• Full‐time users of digital behind‐the‐ear hearing aids.
• No ANSD
• No previous experience with frequency lowering technology
• Oral‐Aural communicators with English as primary language
• 5‐13 years of age (Mean Age: 10 years, 6 mths)
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Procedures
• Measured unaided audiometric thresholds with insert earphones coupled to foam eartips.
• Measured RECD with same foam eartip.
• Used Audioscan Verifit to calculate threshold at TM in dB SPL.
• Fit hearing aid to appropriate earmold.
• Entered thresholds (dB HL) into Phonak iPFG fitting software.
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Step 1: Fit to target without frequency compression
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Step 2: Ensure that high‐frequency sounds are audible
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Procedures
•
Evaluated speech production, speech recognition, and aided thresholds with subjects’ own hearing aids and Phonak Nios BTE hearing aids.
•
Subjects wore Phonak Nios BTE hearing aids for two 6‐week periods:
– NLFC Off
– NLFC On
•
Order in which NLFC was used was counter‐balanced across subjects.
•
After completion of the two 6‐week trials, the subjects wore the hearing aids with NLFC enabled for 6 months. A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Procedures
• Aided Thresholds
– 4000, 6000, & 8000 Hz
– Recorded /sh/ & /s/, Univ Western Ontario
• Speech Recognition
– University of Western Ontario Plural Test
– Phonak Logatome Test
– BKB‐SIN
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
• Results
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Aided Thresholds (dB HL)
NLFC Off vs. NLFC On
*: p < .05
**: p < .01
***: p < .001
NLFC provides a statistically significant
improvement in aided thresholds.
A Sound Foundation Through Early Amplification 2010
Wolfe
et al. (in press), J Am Acad of Audiol
Panel Session on Frequency Lowering
UWO Plural Test
NLFC Off vs. NLFC On
NLFC improves speech recognition on
UWO Plural Test by 16% points.
A Sound Foundation Through Early Amplification 2010
Wolfe
et al. (in press), J Am Acad of Audiol
* P < .001
Panel Session on Frequency Lowering
Speech Recognition Threshold (dB SPL) for 7 VCV Tokens
*
*
* : p = .05
A Sound Foundation Through Early Amplification 2010
Wolfe
et al. (in press), J Am Acad of Audiol
Panel Session on Frequency Lowering
Speech Recognition in Noise
NLFC Off vs NLFC On
Non‐linear Frequency Compression does not degrade speech recognition in noise.
A Sound Foundation Through Early Amplification 2010
Wolfe
et al. (in press), J Am Acad of Audiol
Panel Session on Frequency Lowering
Logatome Thresholds
Improvement in speech recognition in
quiet observed at 6-month interval
*
*
*
*
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Speech Recognition in Noise on BKB‐SIN
NLFC provides significant improvement
in noise after 6 months!
*
* P < .05
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Jax
• 13‐year old boy
• Congenital hearing loss of unknown etiology
• Previously wore Perseo
211 Behind‐the‐Ear hearing aids
• Excels in typical classroom placement
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Summary
• NLFC improves speech recognition and speech production for children with moderate hearing loss.
• Research needed to examine pros and cons of mild losses!
• Verification is key
– Probe microphone measures with calibrated stimuli designed for verification of frequency lowering hearing aids or with live voice stimuli (/sh/, /s/).
– Ensure adequate sound quality
– Aided speech recognition
• Remember earmold acoustics!
• Children may need to acclimate
– May require time to develop speech recognition and production.
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Acknowledgements
• Susan Scollie, Ph.D. & Danielle Glista, M.Sc., University of Western Ontario
• Teresa Carway, Ph.D., SLP, LSLS, Hearts for Hearing
• Andrew John, Ph.D., University of Oklahoma Health Sciences Center
• Erin Schafer, Ph.D., University of North Texas
• Myriel Nyffeler, Ph.D., Michael Boretzki, Ph.D., and Christine Jones, Au.D., Phonak
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
• Thank you for your attention
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Evaluation of frequency compression for moderate to profound hearing losses
Andrea Bohnert
University Medical Center of the Johannes Gutenberg‐University Mainz
Department for Oto‐Rhino‐Laryngology, Division for Communication Disorders
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Children with a severe to profound loss:
ƒ Can we demonstrate speech recognition benefits?
ƒ In quiet as well as in noise conditions?
ƒ Can articulation be improved?
Which configurations of hearing loss will benefit….
• Steep or flat losses???
A Sound Foundation Through Early Amplification 2010
34
Panel Session on Frequency Lowering
Clinical field trial
Audiogram right / left
ƒ 4 female, 9 male
(6 – 15 years)
ƒ Average age: 10,5 years
Audiogram right
125 Hz 250 Hz 500 Hz 1 kHz
2 kHz
4 kHz 6 kHz
8 kHz
0
20
40
60
ƒ All experienced HA
users
ƒ Fitted on DSL basis with
high-quality HA
Mean
80
100
120
140
Audiogramm left
125 Hz 250 Hz 500 Hz 1 kHz
2 kHz
4 kHz 6 kHz
8 kHz
0
20
ƒ 4 main stream school
ƒ 9 school for HI
40
60
80
Mean
100
120
140
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
35
Study ‐ Group results – GII T1 vs T5
Göttinger II quiet 65 dB
Göttinger II quiet 55 dB
T1 own HI
T2 SR on
T2 SR off
T5 SR on
T1 own HI
T5 SR off
T2 SR off
T5 SR on
T5 SR off
90
90
80
Speech intelligibility (%)
Speech intelligibility (%)
T2 SR on
100
100
70
60
50
40
30
20
80
70
60
50
40
30
20
10
10
0
0
1
2
3
4
5
6
7
8
9
10
means
1
2
3
4
5
6
7
8
9
10
means
Göttinger II in noise 65 dB
T1 own HI
T2 SR off
T2 SR off
T5 SR on
T5 SR off
100
90
Speech intelligibility (%)
80
70
60
50
40
30
20
10
0
A Sound Foundation Through Early Amplification 2010
1
2
3
4
5
6
7
8
9
10
means
Panel Session on Frequency Lowering3
6
Study ‐ Group results – GII T1 vs T5
80
Speech intelligibility (%)
70
Göttinger II quiet 55 dB
70
p < 0.001
p < 0.01
60
80
Göttinger II quiet 65 dB
p < 0.01
p < 0.01
80
Göttinger II noise 65 dB
70
60
60
p < 0.001
50
50
50
40
40
40
30
30
30
20
20
20
10
10
10
0
0
0
Means
1
A Sound Foundation Through Early Amplification 2010
Means
1
p < 0.05
Means
1
Panel Session on Frequency Lowering3
7
Study – Group results – AAST Spondee in quiet
Spondee Test in quiet
- high frequency words -
SPIQ HI ow n T1
SPIQ SR on T2
SPIQ SR off T2
SPIQ SR on T5
SPIQ SR off T5
0
10
SRT (d B)
20
30
40
50
60
70
Subject 1
Subject 2
Subject 3
Subject 4
Subject 5
Subject 6
Subject 7
Subject 8
Subject 9
Subject 10
Subject 11
Subject 12
Subject 13
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering3
8
Study – Group results – AAST Spondee in quiet
Spondee Test in quiet
- high frequency words 1
2
3
4
5
0
10
SRT (dB)
20
30
p<0.01
40
50
60
70
80
mean
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering3
9
Study – Group results – AAST Spondee in noise
Spondee Test in noise
- high frequency w ords (trochee) SPIN SR on T5
20
SPIN SR off T5
16.7
11.5
SNR (dB)
10
9
8.5
6.5
5.5
3
2
7.7
12
3
1
7.7
7.5
4
3
12
2.5
2
3.5
5.8
0
-3
-10
-8
-5 -4.5
-5
-8.5
-11.5
-6.5
-6.5
-6
-7.5
-10
-20
Sub j ect 1- 13
A Sound Foundation Through Early Amplification 2010 40
1.5
-4
-10
-9
T1 versus T5 on p<0.01
T5 on versus off p<0.05
Improvement
SPIN ow n HI T1
Panel Session on Frequency Lowering
Two examples…
Sub A steep loss
Sub B flat loss
Frequency in Hz
-20
-20
0
0
20
20
40
40
dB HL
dB HL
Frequency in Hz
60
60
80
80
100
100
120
120
250
500
1000
2000
4000
8000
250
500
1000
2000
4000
8000
10 y, good speech development
8 y, good speech development
Own HI = Eleva 411
Own HI = Siemens Artis P
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
41
Two examples… Sub A steep loss
Own HI
A Sound Foundation Through Early Amplification 2010
NLFC HI
Panel Session on Frequency Lowering
42
Two examples… Sub B flat loss
Own HI
A Sound Foundation Through Early Amplification 2010
NLFC HI
Panel Session on Frequency Lowering
43
Case Studies
ƒ Speech Scores open and closed sets
Subj A
Open set
words
55 dB
65 dB
Closed set
quiet 65 dB
SRT
Closed set
noise 65 dB
SNR
Trad HA
Subj B
Freq Comp
T2
Freq Comp
T5
Open set
words
55 dB
Trad HA
Freq Comp
T2
Freq Comp
T5
0%
0%
40%
10 %
40%
70%
30 %
50 %
60%
60%
70 %
90%
36 dB
32 dB
27 dB
Closed set
quiet 65 dB
SRT
56 dB
42 dB
42 dB
-3,5 dB
Closed set
noise 65 dB
SNR
9 dB
7,7 dB
3 dB
2 dB
-5 dB
A Sound Foundation Through Early Amplification 2010
65 dB
Panel Session on Frequency Lowering4
4
Case studies:
Subject A
Subject B
Hears new sounds, birds etc.
Teacher can be heard with less effort
More relaxed after school
More relaxed after school
Rather relaxed facial expression
Audio books can be heard with
Trivial sounds are recognized
earlier
TV set to normal volume
Speaks with clearer voice –
more self confident
A Sound Foundation Through Early Amplification 2010
normal volume
More open-minded – takes part in
holiday camps with 50 children
Does not accept everything in
conversation, but argues
Panel Session on Frequency Lowering
45
Summary
ƒ Good spontaneous acceptance for all kids
ƒ ‐ Kids can hear many new sounds (birds, bells etc.)
ƒ ‐ Improved communication
ƒ More activities after school
ƒ Families judge children‘s speech as clearer and more
precise
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering4
6
Clinical implications ‐ future questions….?
9 Viable and robust technology for profound hearing losses
9 It does need to be individually fitted
We still need to learn more……..for example:
- Cochlear implant candidacy
- Asymmetrical hearing loss
- Auditory neuropathy disorders
- Bimodal fittings
A Sound Foundation Through Early Amplification 2010
47
Panel Session on Frequency Lowering
Clinical implications – future questions?
9 Test results maybe not always consistent
9 Do we have the right tests to show all effects of
modern technology?
We should always listen to our children…..
A Sound Foundation Through Early Amplification 2010
48
Panel Session on Frequency Lowering
Acknowledgements…
Clinic-Team University, Mainz
Sabine Müller, Petra Brantzen, Martina Dammeyer,
Bianka Schramm, Prof. Annerose Keilmann
Audiology-Team Phonak, Stäfa
Dr. Myriel Nyffeler, Kai Hessefort, Steffi Kalis
and to all children!
A Sound Foundation Through Early Amplification 2010 49
Panel Session on Frequency Lowering
Do children need an acclimatization period?
Danielle Glista, Ph.D.
Child Amplification Laboratory
National Centre for Audiology,
University of Western Ontario
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Acknowledgements
• This work was supported by:
Canadian Institutes of Health Research, Masons
Help‐2‐Hear Foundation and Phonak AG
• Special thanks to: Susan Scollie, Richard Seewald, Meg Cheesman, David Purcell and Jacob Sulkers for their contributions
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
What is auditory acclimatization?
New acoustic information
Time to acclimatize
Example: new audibility of speech cues post hearing aid fitting
Systematic change in auditory performance
From the Eriksholm workshop on Auditory Deprivation and Acclimatization (Arlinger et al., 1996)
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Why study auditory acclimatization?
• Work by Stuart Gatehouse and the Eriksholm
Workshop on Auditory Deprivation and Acclimatization (1995):
– Auditory acclimatization is a real phenomenon with important research/clinical implications
– Evidence suggests the mean reported improvement in benefit over times ranges from 0 to 10% (across speech materials and presentation conditions)
(Arlinger et al., 1996)
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Why study auditory acclimatization?
• Previous research on nonlinear frequency compression (FC) and speech perception benefit suggests:
– Considerable performance variability at the level of the individual ‐ adult and child data (Simpson, 2009)
– Pediatric pilot data provides informal evidence of an acclimatization effect for some listeners (Glista et al, 2009)
• As with all fittings involving new, complex signal processing, adaptation time becomes important
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Study design
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Case Study
• 11 years of age
• Exposure to ototoxic
medication
• Long‐term, full‐
time HA user
• Suspected dead regions (TEN test: Moore, Glasberg & Stone, 2004)
A Sound Foundation Through Early Amplification 2010
CNT
DR
DR
DR
Panel Session on Frequency Lowering
Fitting details
y
Study worn aids = Naida IX SP, SoundRecover setting = 1600 Hz cut‐off, 4:1 ratio
y
DSL v5.0, FC setting individualized (refer to AudiologyOnline: Glista & Scollie, 2009)
6300 Hz with FC
4000 Hz with FC
4000 Hz without FC
Is this enough?
Live /∫/ is
broader in
bandwidth…
6300 Hz without FC
/∫/ Filtered high‐frequency speech bands
/s
/
Screen captures from the Audioscan Verifit
A Sound Foundation Through Early Amplification 2010
Live speech with FC enabled
Panel Session on Frequency Lowering
Results – Speech recognition: Plurality
UWO Plurals Test
“Crayon”
A Sound Foundation Through Early Amplification 2010
“Crayons”
Panel Session on Frequency Lowering
Results – Speech recognition: Plurality
Improvement
Baseline
Treatment
Withdrawal
Time course
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Results – Speech recognition: Consonants DFD Test (Cheesman & Jamieson, 1996)
“aTil”
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Results – Speech recognition: Consonants A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Results – Discrimination of /s/ vs. /∫/
“see”
A Sound Foundation Through Early Amplification 2010
“she”
“see”
Panel Session on Frequency Lowering
Results – Discrimination of /s/ vs. /∫/
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Results – Detection of /s/ and /∫/
Ling‐6 Sound Test
“mmm”
“ooo”
“aaa”
“sss”
“eee”
“shh”
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Results – Detection of /s/ and /∫/
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Clinical implications
• Overall, significant speech perception benefit was reported with FC compared to without FC
• Acclimatization trends with FC:
– Benefit change ranged from 0 to 17%, across measures
– Significant acclimatization trends were observed after approximately 6 weeks or longer
– Two unique acclimatization patterns where exemplified: • Gradual improvement over time
• Improvement after a specific period of acclimatization (S‐shaped curve)
• Further cases are currently under analysis
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Clinical implications
• Speech perception testing administered on more than one appointment, and after allowing a period of acclimatization can assist in validating FC benefit
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Thank you for listening
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Future directions in evaluating SoundRecover
Michael Boretzki, Ph.D.
Program manager R&T projects – Audiology and fitting methods
Phonak AG
Co-authors:
Nicola Schmitt
Andrea Kegel
Harald Krueger
Julia Rehmann
Frederik Eichhorn
Katrin Meisenbacher
Juliane Raether
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Overview
1.
2.
3.
4.
5.
Background and goals
Design and development
Prototype evaluation
Test revision
Future directions
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Goals in Development of the Logatome test
Development Goals:
1)We wanted a computer‐based test with high sensitivity and specificity to high frequency phoneme intelligibility!
2)We wanted a language‐neutral test!
3)We wanted a test that would be applicable with mild‐to‐moderate hearing losses!
Application Goals:
1)Suitable for comparison of different hearing aids, 2)Suitable for evaluation of different settings of a hearing aid
¾
For example, frequency compression on versus off
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
What does Logatome mean?
„A logatome is an artificial word of one or more
syllables which obeys all the phonotactic rules of a language but has no meaning. Examples of English logatomes would be the nonsense words snarp or
bluck.“ from: Wikipedia
In our test, all of the Logatomes are /aCa/:
For example: Asa, ata, asha
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Test construction: What are we measuring?
Processing
capabilities
Phoneme
recognition
Hearing
impairment
Contextual cognition
Prediction
more needed
Phonemic memory
Identification
less intelligible
Auditory resolution
Discrimination
less distinct
Auditory sensitivity
Detection
less audible
Sound
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Test construction: What are we measuring?
Processing
capabilities
Speech
Speech
Recognition
Contextual cognition
Recognition
Threshold
Threshold
(SRT)
(SRT)
Phoneme
recognition
Hearing
impairment
Prediction
more needed
I Iheard
heard„dog“
„dog“
Phonemic memory
Identification
Auditory resolution
Discrimination
Speech
Speech
Detection
Detection
Auditory
sensitivity
Threshold
Threshold
(SAT,
(SAT,SDT)
SDT)
Detection
Difference
Difference
threshold
threshold
less intelligible
I Iheard
heard
different
different
sounds
less distinct sounds
I Iheard
heard
something
something
less audible
Sound
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Is the Logatome test sensitive to Frequency compression?
SRT dB relative to SRT unaided
n=12 adult subjects with mild hearing loss
SoundRecover on
SoundRecover off
Boretzki, M.; Kegel, A. (2009). The benefits of nonlinear frequency compression for people with mild hearing loss. Audiology Online, November 2009
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Logatome Test Design: Factors Evaluated
1)
2)
3)
4)
5)
6)
Minimize phoneme predictability!
Minimize non‐consonant cues!
Challenge high frequency hearing loss!
Minimize floor and ceiling effects!
Maximize valid responses!
Improve consistency! (revised test)
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Goals 1 and 2: Reduce confounds
• Goal 1: Minimize phoneme predictability!
– By using logatomes (asa, asha, afa) we can reduce guessing from context.
• Goal 2: Minimize non‐consonant cues!
– Embed „asa“, „asha“ etc. in identical vowels, we can prevent guessing from
vowel cues.
Initial /a/ of
“ama”
Consonant
Final /a/ of
“ama”
/s/ from “asa”
/sh/ from “asha”
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Goal 3: Challenge high frequency hearing loss! 3rd octave level dB
• Unvoiced fricatives from a female talker
• Created /s/ at both 6 and 9 kHz
-10
-15
-20
-25
-30
-35
-40
-45
-50
/f/
/s/ 9 kHz
/s/ 6 kHz
/sh/
100
1000
10000
Frequency Hz
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
¾ Is not a speech detection
threshold
¾ Is a speech recognition
threshold, in dB(A) per consonant.
60
50
40
Asa
Not understood
Asa
Not understood
Not understood
Not understood
Asa
• We use an adaptive tracking
procedure to measure the
levels needed for
understanding
• The score:
Presentation level dB
Goal 4: Minimize floor and ceiling effects! 30
Trials Æ
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Goal 5: Maximize valid responses! • Un‐forced choice procedure reduces guessed answers
• Listeners can indicate that they don‘t know
• Listeners can repeat a trial
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Goal 5: Improve consistency
• Providing repetitions of each sound improved
consistency by 0 to 4 dB per Logatome:
– „asa, asa, asa“ rather than just „asa“.
– Near threshold, repeated stimuli may sound different
Click here if the 3 sounds are not all the
same, or if the sounds are too soft to be
heard.
The software will increase the test level
automatically.
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Goal 5: Improve consistency
• Carefully selecting the set of Logatomes:
3rd octave level dB
– Including a wide range of sounds improves consistency
– Our final set for clinical use includes six Logatomes, 3 to 9 kHz region:
-10
-15
-20
-25
-30
-35
-40
-45
-50
/f/
/t/
/s/ 9 kHz
/s/ 6 kHz
/sh/ 5 kHz
/sh/ 3 kHz
100
1000
10000
Frequenzy Hz
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Goal 5: Improve consistency
• Evaluate across languages:
– This testing will continue, early results shown for 25 listeners with NH
– 10 German, 6 English, 9 Thai native speakers
– Logatomes that vary by language (e.g., aka) excluded (final set circled)
SRT dBA (level of /a/)
70
German subgroup
60
English subgroup
Thai subgroup
50
40
30
20
A Sound Foundation Through Early Amplification 2010
Ata
Asha
Asa 9 kHz
Asa 6 kHz
Apa
Ana
Ama
Ala
Aka
Aha
Aga
Afa
Ada
Aba
10
Panel Session on Frequency Lowering
Development of a Clinical Logatome Test:
Order of Tasks
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Development of a Clinical Logatome Test:
RESPONSE SCREEN
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Summary and Future Directions
1) Goal: To develop a language‐neutral intelligibility test that is
sensitive and specific to high frequency phoneme
intelligibility
2) Method: Female talker, unforced choice, non‐consonant
cues mimimized, adaptive SRT measurement for each
stimulus, multiple presentation
•
This method may be challenging for listeners with profound losses
3) Validation studies: Appears sensitive to the effects of frequency compression in mild and moderate hearing losses, other evaluations are in progress (normative data, data
across losses & languages)
4) Adaptation for use with children: A next step .... Feedback?
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering
Thanks for your attention!
Selected References for panel session:
Boretzki, M., Kegel, A. 2009. The benefits of nonlinear frequency compression for people with mild hearing loss. Audiology Online, November.
Glista D, Scollie S. 2009. Modified verification approaches for frequency lowering devices. Audiology Online, November.
Glista, D., Scollie, S., Bagatto, M., Seewald, R., Parsa, V., Johnson, A. 2009a. Evaluation of nonlinear frequency compression: Clinical outcomes. International Journal of Audiology, Vol. 48, No. 9 , Pages 632‐644.
Glista, D., Scollie, S., Polonenko, M. and Sulkers, J. 2009b. A Comparison of Performance in Children with Nonlinear Frequency Compression Systems. Hearing Review, November. MacArdle, B. M., West, C., Bradley, J., Worth, S., Mackenzie, J., and Bellman, S. C. 2001. A study of the application of a frequency transposition hearing system in children. British Journal of Audiology 35: 17‐29.
Miller‐Hansen, D. R., Nelson, P. B., Widen, J. E., and Simon, S. D. 2003. Evaluating the benefit of speech recoding hearing aids in children. American Journal of Audiology 12(2): 106‐113.
Moeller, M. P., Hoover, B., Putman, C., Arbataitis, K., Bohnenkamp, G., Peterson, B., et al. 2007. Vocalizations of infants with hearing loss compared with infants with normal hearing: Part I‐‐phonetic development. Ear and Hearing, 28(5), 605‐627.
Pittman, A. 2008. Short‐Term Word‐Learning Rate in Children With Normal Hearing and Children With Hearing Loss in Limited and Extended High‐Frequency Bandwidths. Journal of Speech, Language, and Hearing Research Vol.51 785‐797.
Stelmachowicz, P. G., Pittman, A. L., Hoover, B. M., Lewis, D. E., and Moeller, M. P. 2004. The importance of high‐
frequency audibility in the speech and language development of children with hearing loss. Archives of Otolaryngology ‐ Head and Neck Surgery 130(5): 556‐562.
Wolfe, J., John, A., Schafer, E., and Caraway, T. (in press). Evaluation of non‐linear frequency compression for children with moderate hearing loss. Journal of the American Academy of Audiology.
Time for discussion (15 minutes)
A Sound Foundation Through Early Amplification 2010
Panel Session on Frequency Lowering