mild hearing loss is serious business harvey dillon sharon cameron, teresa ching, helen glyde, gitte...

Mild Hearing Loss is Serious Business

Harvey DillonSharon Cameron, Teresa Ching, Helen Glyde, Gitte Keidser,

David Hartley, Jorge Mejia

NAL, The Hearing CRC

IHCON, 2010

Slides on the NAL web site:

www.nal.gov.au

What is mild hearing loss?

• Four-frequency average (500, 1000, 2000, 4000 Hz) hearing loss in better ear between 20 and 40 dB HL

• Self-reported disability or handicap within a certain range

• SRT in noise loss of between 3 and 6 dB

Frequency Scatterp lot (Spreadsheet in 30000 audiograms 2010.stw 37v*30132c)

L4FA = 9.5175+ 0.9136*x

< = 0 (0 ,20] (20,40] (40,60] (60,80] (80,100] (100,120] (120,140] (140,160] (160,180] (180,200] (200,220] > 220 0 20 40 60 80 100 120

4FAHL or 3FAHL?

How prevalent is mild hearing loss?

• Davis (1995): 16% of adult population• Wilson (1990): 18% of adult population• Hartley et al (in press): 34% of people aged > 50 years

Blue Mountains 4FAHL better earH istogram (BM H S_audiogram_edited.sta 41v*2956c)

-10 0 10 20 30 40 50 60 70 80 90 100 110 120

4FA_Better

No of ob

Hartley et al (in press)

Aging population - Australia

2010 2020 2030 2040 2050

Population aging, worldwide

0 5 10 15 20 25

Japan Italy

Greece Sweden

France United Kingdom

Canada Australia

New Zealand United States of America

Hong Kong (SAR of China) Korea, Republic of

Singapore China (exl. SARs and

World Viet Nam

Indonesia India

Malaysia South Africa

Philippines Papua New Guinea

Percentage > 65 years

C luster analysis of mild losses

C luster 1 C luster 2 C luster 3 C luster 4 C luster 5 C luster 6 C luster 7 C luster 8 C luster 9 C luster 10250 500 1000 2000 4000 6000 8000

Frequency (H z)

HL in ear w

ith better 4F

L (dB H

Mild losses for study

250 500 1000 2000 4000 8000

Frequency (Hz)

Mild loss (4FA=29 dB)

Mild-mod loss (4FA=39 dB)

~ Minimum loss eligible for

government benefits

~ Median loss newly fitted in Australia

250 500 1000 2000 4000 8000

Frequency (Hz)

Mild 9 dB

Mild-mod 39 dB

Moderate 49 dB

Mod-Sev 59 dB

Severe 69 dB

4FA HL

How common is hearing aid use amongst those with mild hearing loss?

Penetration by hearing loss

• Penetration as a function of hearing loss– Davis (1995)– Wilson et al (1998) – Hartley et al (in Press)– Kochkin

losshearingwithNumber

aidshearingowningNumbernPenetratio

consistent

Aid ownership by hearing lossBlue Mountains Population >55 years

5 15 25 35 45 55 65 75 85 95 105 115

4FA Hearing threshold (dB HL)

5 15 25 35 45 55 65 75 85 95 105 115

lePopulation

Ownership

Hartley et al (in Press)

What are the characteristics of mild hearing loss?

Characteristics of mild loss

• Threshold elevation √• Loss of OHC/OAE• Loss of frequency resolution – TEN, PTC, FRI• Loss of temporal (envelope) resolution• Loss of fine temporal information• Loss of spatial processing• Loss of SRT in noise• Increased disability and handicap

Loss of active process in cochlear:OHC and OAE

-20 0 20 40 60 80 100 120

H L at 2 kH z (dB)

E at 2 kH

z (dB S

Loss of frequency resolution

HTL (dB HL)

0 20 40 60 80 100 120

Ching & Dillon (unpublished data)

HTL (dB HL)

0 20 40 60 80 100 120

350 Hz

HTL (dB HL)

0 20 40 60 80 100 120

HTL (dB HL)

0 20 40 60 80 100 120

HTL (dB HL)

0 20 40 60 80 100 120

Frequency resolution

Loss of temporal resolution

HTL (dB HL)

0 20 40 60 80 100 120

Temporal resolution

HTL (dB HL)

0 20 40 60 80 100 120

350 Hz

HTL (dB HL)

0 20 40 60 80 100 120

HTL (dB HL)

0 20 40 60 80 100 120

HTL (dB HL)

0 20 40 60 80 100 120

Ching & Dillon (unpublished data)

Loss of fine temporal information

• 20 synapses per IHC • Synapse loss or IHC loss reduced

averaging temporal jitter

Inspiration:• Bodian, Lieberman, Moore,

Pichora-Fuller, Spoendlin,

A digression into “normal” hearing

Spatial Processing Disorder

Speech

Sharon Cameron

Listening in Spatialised Noise - Sentences(LiSN-S) Conditions

Total Advantage

Talker Advantage

atial A

High Cue

Low Cue

Same voices Different voices

Same direction

Differentdirections

Cameron & Dillon (2009)

LiSN-S Diagnostic Screen

6 7 8 9

-60A ge Group

Spatia

l Advanta

Spatial Advantage (≡ Spatial Release from Masking)

Australia

Nth America

Spatial AdvantageB

Results profile: spatial processing disorder

Spatial processing remediation Pre vs. Post (n=9)

LC SRT - p = 0.158

Talker Advantage - p = 0.981

HC SRT - p = 0.0002

Spatial Advantage - p = 0.0002

Total Advantage - p = 0.001

L C S RT HC S RT T a l ke r S p a ti a l T o ta l

L iS N-S Co n d i ti o n

P re -T ra i n i n g P o st-T ra i n i n g 3 M P o st-T ra i n i n g

Application to people with hearing loss

Spatial hearing loss in hearing-impaired people:LiSN-S Prescribed Gain Amplifier

Deficit in SRTn with hearing loss

Cameron, Glyde & Dillon, unpublished data)

Low C ue defic it (dB) = 0.0147-0.0631*xH igh cue defic it (dB) = 4.0175-0.3029*x

20 25 30 35 40 45 50 55 60

4FA H L w orse ear (dB H L)

ficit re n

Low C ue defic it (dB ) H igh cue defic it (dB )

R = -0.87

P = 0.000006

Cause of deficit in SRTn T alker Adv defic it (dB) = 3.9432-0.0061*x

Spatia l Adv defic it (dB ) = -3.4051+ 0.2273*x

20 25 30 35 40 45 50 55 60

4FA H L w orse ear

ficit re n

T alker Adv defic it (dB ) Spatia l Adv defic it (dB)

R = -0.86

P = 0.00001

Talker advantage deficit versus age

T alker Adv defic it (dB) = 1.6806+ 0.0407*x

0 10 20 30 40 50 60 70 80 90

lker A

eficit (d

Binaural processing

ILDITD

Executive

control

~Sensorineural

hearing loss

ILDITD

SO / IC / A1

Loss of SNR in understanding speech

“Basic” loss of 0.6 dB per 10 dB of loss

Loss of Spatial release from masking of 2.3 dB per 10 dB of loss

Loss of Talker cue release from masking of 0.5 dB per 10 years of age

-10 0 10 20 30 40 50 60 70 80

Test ear 4FA (500 to 4k Hz)

Version 1 Version 2

Loss of SRT in noiseCommonly 1.5 dB increase in SNR per 10 dB of hearing loss

1 dB / 10 dB

1.8 dB / 10 dB

SNR = -4 dB

Carter, Zhou & Dillon, unpublished data)

Should mild hearing loss interfere with speech perception?

Speech and noise levels

30 40 50 60 70 80 90

Background noise level (dBA)

Source: Pearsons, Bennett and Fidell (1977)

Calculation of SII

100 1000 10000

Frequency (Hz)

Noise = 60 dBA Speech = 64 dBA SIInh=0.65 SIIhi=0.46

Effective audibility

0 10 20 30 40 50 60 70

Sensation level (dB)

ity 40

Transfer functionSII Percent correct

0102030405060708090

0 0.2 0.4 0.6 0.8 1

ConnectedSpeechTest

250 500 1000 2000 4000 8000

Frequency (Hz)

Mild loss (4FA=29 dB)

Mild-mod loss (4FA=39 dB)

Predicted speech intelligibility

30 40 50 60 70 80 90

Background noise level (dBA)S

Normal hear

Mild unaided

Mild-modunaided

30 40 50 60 70 80 90

Norm hear

Mild unaided

Mild-mod unaided

Greatest problems in noisy places !!

Predicted variation of SRTn with hearing loss

0 20 40 60 80

4FA hearing threshold (dB HL)

) 50607080

Modified SII model predicts only 0.4 dB loss per 10 dB of hearing loss

Speechlevel

Modification of SII

• Assume normal hearers get 6 dB advantage from spatial separation of speech and noise

• Assume hearing impaired listeners lose spatial advantage at a rate of 1.3 dB per 10 dB of loss (above SII predictions)

total loss of SNR is 1.7 dB per 10 dB of loss

Calculation of intelligibility

Speech spectrum

Noise spectrum

Threshold

Sensation Level

Effective audibility

Importance function

(Information received)

Percent correctMax

Do hearing aids help people with mild hearing loss?

current

Benefit of hearing aids

Predict increase in speech intelligibilitywith

the modified Speech Intelligibility Index

Calculation of aided benefit

Noise = 50 dBA

Speech = 58 dBA 0

100 1000 10000Inse

100 1000 10000

Frequency (Hz)

100 1000 10000

Frequency (Hz)

Speech intelligibility (mild loss)

30 40 50 60 70 80 90

Norm hearDirectOmniUnaided

30 40 50 60 70 80 90

Norm hear

Direct

Unaided

• Hearing aid “helps” in quiet places

• Tiny additional benefit from directivity

Conditions:

DI = 3 dB when REIG > 3 dB

n.h. spatial adv = 6 dB

∆SNR re n.h. = -3.7 dB

Speech intelligibility (mild-moderate loss)

30 40 50 60 70 80 90

Norm hear

Direct

Unaided

• Hearing aid “helps” in quiet places

• Tiny additional benefit from directivity

Conditions:

Speech intelligibility (moderate loss)

30 40 50 60 70 80 90

Norm hear

Direct

Unaided

Conditions:

Speech intelligibility (moderate-severe loss)

30 40 50 60 70 80 90

Norm hear

Direct

UnaidedConditions:

Speech intelligibility (severe loss)

30 40 50 60 70 80 90

Norm hear

Direct

UnaidedConditions:

∆SNR loss re n.h. = -9 dB

Summary of benefit versus HL

0 20 40 60 80

4FA Hearing Threshold (dB HL)

0 20 40 60 80

40 dB A

60 dB A

80 dB A

Background noise level

Why don’t directional microphones help more in noise?

(and adaptive noise reduction)

1. Impact of open fittings on directivity

125 250 500 1k 2k 4k 8kFrequency (Hz)

dB) Combined path

Amplif iedpath

Ventpath

DI = 3 dB 1.5 dB

DI = 10 dB 5 dB

DirectionalOmni-directional

Frequency range of directional mic (mild loss)

100 1000 10000

Frequency (Hz)

5560657075808590

Effect of dynamic

noise reduction on

directional mic

Effect of aiding at 60 dBA (mild loss)

100 1000 10000

Frequency (Hz)

Noise = 60 dBA

Speech = 64 dBA

100 1000 10000

Frequency (Hz)

100 1000 10000

Frequency (Hz)

Noise = 80 dBA

Speech = 76 dBA

100 1000 10000

Frequency (Hz)

100 1000 10000

Effect of aiding at 80 dBA (mild loss)

2. Impact of reverberation on directivity (and vice versa)

0 1 2 3 4 5

Distance from source (m)

DirectReverberant

0 1 2 3 4 5

Distance from source (m)

DirectReverberant

057.0RT

VQQdistcritical ms

effective

2. Impact of reverberation on directivity (and vice versa)

Benefit of directional microphones

Benefit affected by:– Directivity pattern of microphone– Distance and direction of talker and noise sources– Acoustics of the room– Frequency range over which the hearing aid is

directional– Frequency range over which the wearable has usable

hearing

– Nothing else

(OK, Measurement error!!)

Conclusion of acoustic analysis

1. People with mild loss need help in noisy places

2. Hearing aids increase the speech information available mostly in quiet places!

3. Mics directional only where there is gain high frequencies and lower levels Where audibility is limited by threshold, not noise

So, objective benefit questionable, and increasing with degree of loss, but …….

What do hearing aid users say ?

Experimental evaluation of self-reported benefit

• 400 clients sampled from national database– 41,521 new clients fitted Feb to Sept, 2004

• Audiometric and other details obtained from clients’ files

• Questionnaire sent to clients 5 to 12 months after fitting

– International Outcome Inventory for Hearing Aids

– Plus 6 purpose-designed questions

• Non-responders followed up by phone or additional mail to get a high response rate (effectively 86%)

Hearing loss distribution

5 15 25 35 45 55 65 75 85 95 105 115

leSample distribution

Population distribution

Usage of hearing aids

None< 1 hr/day

1-4 hrs /day4-8 hrs /day

> 8 hrs /day

Q 3: Daily us age

f clie

Factor analysis of questionnaireFactor Loadings

Factor 1 Factor 2 Factor 3

Q1: want aids 0.69 0.30 0.21

Q2: difficulty unaided 0.70 0.41 0.25

Q3: use 0.74 -0.18 0.08

Q4: benefit 0.82 -0.32 0.00

Q5: residual difficulty 0.03 -0.76 -0.33

Q6: Worth it 0.83 -0.33 -0.00

Q7: Residual handicap -0.29 -0.56 -0.06

Q8: Bother to others -0.18 -0.68 -0.26

Q9: Quality of life 0.82 -0.32 0.02

Q10: Replace them 0.34 -0.15 -0.23

Q11: Face vision -0.29 -0.42 0.74

Q12: paper vision -0.22 -0.47 0.70

Proportion of variance 0.32 20 0.12

Composite benefit Composite difficulty Vision

International Outcomes Inventory

for Hearing Aids

Effect of hearing loss on benefitCurrent effec t: F (3, 281)= 1.1951, p= .31194

V ert ic al bars denote 0.95 c onfidenc e intervals

10-19 20-29 30-39 40-49

4FA H L in better ear (dB H L)

posite benefit

If hearing loss does not determine benefit, then what does?

Wishes And Needs Tool

1. How strongly did you want to get hearing aids? Wanted it very much Wanted it quite a lot Wanted it moderately Wanted it slightly Did not want it

2. Overall how much difficulty do you have hearing when you are not wearing your hearing aids?

Very much difficulty Quite a lot of difficulty Moderate difficulty Slight difficulty No difficulty

Q 2: No diffic ulty unaided

1 2 3 4 50

Q 2: S light diffic ulty unaided

1 2 3 4 5

Q 2: M oderate diffic ulty unaided

1 2 3 4 5

Q 2: Q uite a lot of diffic ulty unaided

1 2 3 4 50

Q 2: V ery m uc h dific ulty unaided

1 2 3 4 5

Difficulty hearing unaided and wish to get hearing aids

Unaided difficulty related to wish to get

hearing aids

Want hearing aidsVery muchNot at all

Need increases with hearing lossC urrent e ffect: F (3, 275)= 8.5017, p= .00002

Vertica l bars denote 0 .95 confidence in tervals

10-19 20-29 30-39 40-49

B etter ear 4FA range (dB H L)

Cu rre n t e ffe ct: F(8 , 1 0 9 )=7 .9 1 0 9 , p = .0 0 0 0 0

1 1 .5 2 2 .5 3 3 .5 4 4 .5 5

S tre n g th o f n e e d

fit: =("Q

fit" + "Q

Benefit versus need strength

Why don’t more people with mild hearing loss even try hearing aids?

Factors affecting benefit experienced (and hence the reports of others)

• Degree of pure-tone loss

• Self-reported disability and handicap

• Acceptable Noise Level

• Stigma / cosmetic concern

• Manipulation and management

• Age

• Tinnitus

• Personality ….

Personality

People more likely to acquire hearing aids are:– Open– Non-obsessive– Non-neurotic– Internal locus of control

People more likely to report benefit are:– Extroverted– Agreeable

Health Belief Model

People act rationally, in their best interests, based on their beliefs

Weighing up of beliefs for and against a health decision

What are the disadvantages of the solution?

Do I have a problem?

Is it serious enough for me to want to remove it?

Is there a solution

that works?AdvantagesDisadvantages

“My hearing loss is not bad enough to need them” (Kochkin, 1993)

Motivation comes from ….

Acknowledge loss

Experience difficulty

Experience handicap

Self-image

Expect benefit

Expect to manage them

$ Cost OK

Try them

Health belief model

Hearingloss

Difficulties experienced:

frequency, severity

Self-image

Ability to manage

Hearing aideffectiveness

Externalimage

Incon-venience

Health belief model

Hearingloss

Difficulties experienced: frequency, severity

Self-image

Ability to manage

Hearing aideffectiveness

External image

Inconven-ience

Health belief model

Hearing

Difficulties

experienced:

frequency, severity

Ability to

manageHearing aid

effectivenessExternal

Inconven

-ience

How can the balance of benefits to disadvantages be improved?

Improving advantages and removing disadvantages

Problem 1 Solution 1

• WDRC

• Slim-tube, miniaturization

• Feedback cancellation

• Low-level expansion

• Re-chargeable batteries

• Auto telecoil

• Frequency lowering

• Bandwidth extension

• Trainable responses

• Adaptive noise reduction

• Transient noise reduction

• Directional microphones

Working better in noise

Speech intelligibility in noise

Binaural-Processing Super-directional Microphone (Mejia et al., 2007)

Rear- directional array

Masking threshold

Cross-fading process

Outputs with spatial reconstruction

HRTFL HRTFR

d = 3 ms delay

(Precedence effect)

DOA- reconstruction

Q1 Q2 Q3 Q4

Binaural beamformer

Main directional signal

Subsidiary signal

Super-directional microphones

Speech reception threshold in noiseReverberant room: crit dist = 0.4m, radius = 1 m

NH Dir BF NH Dir BF NH Dir BF

2-speaker 4-speaker 140-babble

Mejia and Johnson, unpublished data

ListenerListener

Hearing Aid

Linked binaural hearing aid technology

Female talker

Male talker

Childrenplaying

Blind-source separation binaural noise reduction

Implication: People with mild or moderate hearing loss were not making much use of spatial cues.

n.h. benefit from re-insertion of spatial cues

h.i.gain most from

directivity

Effect of super-directivity (mild loss)

30 40 50 60 70 80 90

Norm hear

Direct

Unaided

Super-directional

30 40 50 60 70 80 90

Norm hear

Direct

Unaided

Super-directional

• Super-directional mic not directional at all over a broader and broader range as noise levels rise

Conditions:

So super-directivity alone is no use ……..

What if we could achieve directivity at low frequencies?

Effect of low-frequency directivity (mild loss)

30 40 50 60 70 80 90

Norm hear

Direct

Unaided

Extended direct

30 40 50 60 70 80 90

Norm hear

Direct

Unaided

Extended direct

• Now directional over entire frequency range in noisy places

Conditions:

DI = 3 dB when REIG ≥ 0 dB

Occlusion with closed moldsVent or leak transmission in

Active Occlusion Reduction

HearAid C

Active occlusion reduction

Mejia, Dillon, & Fisher (2008)

In combination?

Active occlusion reduction

(closed mold)

Super-directivity = ?+

Low-frequency super-directivity (mild loss)

30 40 50 60 70 80 90

Norm hear

Direct

Unaided

Extend/super

30 40 50 60 70 80 90

Norm hear

Direct

Unaided

Extend/super

• Super-directivity over entire frequency range

super-normal hearing

Conditions:

100 1000 10000

Frequency

Low-frequency super-directivity (mild-moderate loss)

30 40 50 60 70 80 90

Norm hear

Direct

Unaided

Extend/super

30 40 50 60 70 80 90

Norm hear

Direct

Unaided

Extend/super

• Super-normal hearing for the median hearing aid wearer

Conditions:

100 1000 10000

Frequency

Low-frequency super-directivity (moderate loss, 4FA = 49.8 dB HL)

30 40 50 60 70 80 90

Norm hear

Direct

Unaided

Extend/super

30 40 50 60 70 80 90

Norm hear

Direct

Unaided

Extend/super

• Super-normal hearing in very noisy places for even a moderate-severe loss

100 1000 10000

Frequency

Conditions:

Adaptive noise reduction

• Gain dependent on SNR correct in principle, but room for improvement:– Gain reduction can reduce directional mic

effectiveness– No point in reducing noise below threshold– Gain increase where SNR is best is just as

important

Jump to summary

Self-fitting hearing aid

Audiometer

NAL-NL2 Prescription

Formula

Real-ear to coupler difference

Adjust Hearing Aid

Automatic

Developing countries: 1 audiologist per 500,000 people, to 1 per 6,000,000

Australia, USA: 1 audiologist per 10,000 people

Plus tra

inability

-10 0 10 20 30 40 50 60 70

M anual thresho lds (dB H L)

atic th

s (dB H

Automatic versus manual audiometry

Test-retest standard deviations

250 1000 4000

Frequency (Hz)

Manual audiometry (5 dB Hughson-Westlake)

Automatic audiometry (2 dB final step size)

NAL-NL2

Adjust

Auto Aud

Real-Ear to Dial Difference:Inter-subject standard deviations

0123456789

100 1000 10000

Frequency (Hz)

B) Insert

ClosedOpen

Real-Ear to Dial Difference:Inter-subject standard deviations

0123456789

100 1000 10000

Frequency (Hz)

Insert

Closed

Saunders & Morgan,InsertValente et al, ER3A

Valente et al, TDH39

Hawkins etal, TDH39

Hawkins et al, ER3A

Insert

Saunders & Morgan

Valente et al

Hawkins et al

Supra-aural

Valente et al

Hawkins etal

Trainable Hearing Aids

30 40 50 60 70 80 90

Input level (dB SPL)

) CRGain

In summary…

In summaryvery

not very

many, including spatial hearing loss

yes, in noise

only in quiet places

….. expected benefit too small re need

closed-ear, binaural processing

How prevalent is mild hearing loss?

How common is hearing aid use amongst those with mild hearing loss?

What are the characteristics of mild hearing loss?

Is mild hearing loss a problem to people?

Do hearing aids help people with mild hearing loss?

Why don’t more people with mild hearing loss even try hearing aids?

How can hearing aids provide greater benefit where it is most needed?

Messages for ….

Public health authorities:

• Increase hearing awareness (prevention, rehabilitation)

• Increase hearing screening opportunities

Messages for ….

Clinicians:

• Discern primary reasons why unmotivated clients are unmotivated

• Provide information to change unrealistic beliefs

• Understand and diagnose the fundamental problem that clients are presenting with SRT loss

Messages for ….

Researchers:

• Better understanding of the components and causes of SNR loss

• Prescription procedures for adaptive noise suppression

• Time constants• Relationship with thresholds• Relationship with noise spectrum and level

Messages for ….

Manufacturers:• Achieve better performance in noise

Binaural processing Closed fittings Wireless Smarter adaptive noise suppression

Hearing aids of the future

Convergence:

• hearing aid/enhancer,

• phone interface,

• hearing protector,

• computer interface (in and out),

• music player,

• GPS interface

Thanks for listening

For the slides from this talk ..…

www.nal.gov.au

Amplification and directivity

100 1000 10000

Frequency (Hz)

100 1000 10000

Frequency (Hz)

100 1000 10000

50 dBA 80 dBA

Unaided

mild hearing loss is serious business harvey dillon sharon cameron, teresa ching, helen glyde, gitte...

db of loss slide

hearing loss slide

db slide

db loss

noise loss

spatial hearing loss

snr slide

press slide

Documents

the audiometric threshold: measured in-situ, automated, and...

stop brake pad squeal! - ags company...sil-glyde® and...

henry r. glyde - delaware physicsglyde/hrgcv06_2005.pdf ·...

receiver-in-canal hearing instruments ·...

hearing instrument for hearing loss patients | bengal speech...

study guide - national board for certification in hearing...

guidelines for fitting multiple memory hearing aids ·...

about hearing aids | hearing aids delhi | center for hearing...

detrital zircon geochronology of the glyde and favenc...

hearing aids & implantable hearing devices

code enforcement hearing agenda hearing date & time...

omega-glyde ® polytetrafluoroethylene (ptfe) pc) ·...

hearing loss hearing impairment

modern hearing protection & modern hearing aids€¦ ·...

hearing, touch, taste and smell. hearing audition – the...

hearing-impaired normally hearing

hearing and hearing conservation for musicians hearing and...

about hearing aids | center for hearing aids | buy hearing a

localized bose-einstein condensation in liquid 4he in...

optica - glyde architectural acoustic systems