Prescribe hearing aids to:

Harvey Dillon

Audiology NOW! 2010

Prescribing hearing aids for adults and children


Adult

Child

Measure hearing thresholds (dB HL)

Measure hearing thresholds (dB HL or dB SPL)

Measure individual RECD (or estimate from age)

Harvey Dillon, Gitte Keidser, Teresa Ching, Matt Flax, Scott Brewer

Enter into manufacturer software (hearing aid auto adjusted to approximate prescription)

HEARing CRC National Acoustic Laboratories

Enter into manufacturer software (hearing aid auto adjusted to approximate prescription)

Verify with real ear measurement Audiology NOW! San Diego

Adjust amplification to better match prescription

2010

Infant Fitting Procedure Electrophysiological hearing threshold with insert phones (dB nHL)

Behavioural hearing threshold with insert phones (dB HL)

Calculate hearing threshold level (adult equivalent dB HL or dB SPL in ear canal) Verification of REAG ?

Adjust hearing aid in coupler via computer to better match prescribed coupler gain

Prescribe hearing aids to: Measure individual RECD, (or estimate RECD from age)

Apply prescription to derive coupler gain targets



Make speech intelligible Make loudness comfortable



Prescription affected by other things



– – – –

localization, tonal quality, detection of environmental sounds, naturalness.

Adjust hearing aid via coupler/programmer to achieve coupler gain targets

Evaluation !

The rationale for NAL proceudres

Deriving optimal gains - step 1 Speech spectrum & level

Maximize calculated speech intelligibility , but Keep total loudness less than or equal to normal

Loudness model

Gain-frequency response

Normal loudness

Compare

Intelligibility achieved Intelligibility model

NAL-NL1 (1999)  empirical studies  psychoacoustic studies

Amplified speech spectrum

NAL-NL2

 speech intelligibility models Audiogram

NAL – The Hearing CRC

Loudness model

Loudness (hearing impaired)

Harvey Dillon

Audiology NOW! 2010

Overall approach to prescription


Limiting compression ratio

Psychoacoustics

The maximum C

Theoretical predictions

Assumptions, rationale

Final formula 3.0 Ca

2.0 1.5

Compare

Speech science

6

2.5

4 f log2 125

1.0 0

H 50

Empirical observations

2

100

Multi-dimensional equation:

Effect of language

A neural network H250

H500

H1000

H2000

H8k



SPL





Gain at each frequency depends on importance of each frequency Low frequencies more important in tonal languages Two versions of NAL-NL2 – Tonal languages – Non-tonal languages

G500

G1000

G2000

G8k

Calculating loudness Loudness model of Moore and Glasberg (2004) Allowance for hearing loss

External & middle ear

Filtering into auditory bands

1/3 octave SPL

G250

30

Nois e

Threshold

Freq Calculate loudness per band

Sum across bands

Audibility:

... 5

x Importance:

Free field speech level

Input to cochlea

Excitation level

Loudness per band

Total loudness

...


... 16 x

0.001

0.002

=

=

... 17 x

... 0 x

0.003 0.002 =

=

0.005 ... 0.032 ... 0.051 ... 0

= 0.30

Harvey Dillon

Audiology NOW! 2010


Speech Intelligibility Index

The transfer function Percent Correct

Sum

SII = ∑ Ai Ii Importance

Audibility

Sentences

100 80 60

Nonsense syllables

40 20 0 0

But intelligibility gets worse if we make speech too loud!

Observed and Predicted performance

0.2

0.4

0.8

1

Audibility and Speech intelligibility – H.I.

1400 - 5600 Hz

VCV

1.0

100

MS

MF

SF

SS

80

0.8

60 Proportion correct

Percent correct

0.6

Speech Intelligibility Index (SII)

40 20 0

12

24

36

12

24

36

12

24

36

12

24

0.6

0.4

0.2

36

Sensation level (dB)

0.0 0.0

0.2

0.4

Ching, Dillon & Byrne, 1998

Deficit = Sansii - SIIeff

0.6

0.8

1.0

LP7: Q LP7: N LP14: Q LP14: N LP28: Q LP28: N LP56: Q LP56: N HP28: Q HP28: N HP14: Q HP14: N HP7: Q HP7: N

Calculated SII

Intelligibility and audibility

Percent Correct

100 80

Deficit = 0.6 - 0.4 = 0.2



60

1

40 m

20

SIIeff

SIIansi p

0 0

0.2

0.4

0.6

0.8

Speech Intelligibility Index (SII)


1 30


Harvey Dillon

Audiology NOW! 2010


Desensitisation for hearing loss

BKB, VCV and CUNY

1 0.8

0 20 40 60 80 100 120

0.7 0.6 0.5 0.4 0.3 0.2 0.1

1

0

0.8

0

20

40

m parameter

Effective audibility

0.9

60


0.6 0.4 0.2 0 0

20

40

60

80

100

Hearing threshold (dB HL)

Frequency resolution AFRI (dB)

60 350 Hz

40 20 0 -20 0

20

40

60

80

100

120

HTL (dB HL)

AFRI (dB)

60

Psychoacoustics

1 kHz

40 20 0 -20 0

20

40

60

80

100

120

HTL (dB HL)

Why measure only pure tone thresholds? AFRI (dB)

60 2 kHz

40 20 0 -20 0

20

40

60

AFRI (dB)

40 20 0 -20 20

40

60

ATRI (dB)

Other measurements

350 Hz

40 20 0 -20 0

20

40

60

80

100

120

HTL (dB HL)

ATRI (dB)

60



1 kHz

40 20



0 -20 0

20

40

60

80

100

120

HTL (dB HL)

ATRI (dB)

60 2 kHz

40



20 0 -20 0

20

40

60

80

100

120

HTL (dB HL)

ATRI (dB)

Reason: OHCs lost

60 4 kHz

40

0 -20 0

20

40

60

HTL (dB HL)


80

100

120

Hearing threshold levels Outer hair cell function – click-evoked otoacoustic emissions Frequency resolution – psychophysical tuning curves – cochlear dead regions – TEN test

Cognitive ability  Age 

20

120

4 kHz

HTL (dB HL)

60

100

60

0

Temporal resolution

80

HTL (dB HL)

Reason: OHCs lost

80

100

120

120

Harvey Dillon

Audiology NOW! 2010


Healthy PTC – no dead region

Poor PTC: Dead region at 4 kHz Psychophysical tuning curve

Psychophysical tuning curve

120 A29

110

Masker Level (dB SPL)

Masker Level (dB SPL)

Masker

100 90 80 70

110

100

90

Signal

60

80

50

500

1000

1500

2000

2000 3000 Masker Frequency (Hz)

4000

5000

Off-frequency listening: TEN test

Off-frequency listening: PTC Basilar membrane vibration

1000

Basilar membrane vibration

Masker Frequency (Hz)

Threshold Equalizing Noise (TEN)

Frequency or position

Frequency or position

Based on Moore (2004)

TEN and PTC (non) agreement 2 kHz

TEN: Alive

TEN: Dead

TEN uncertain

PTC: Tip in place

60

1

PTC: Tip shifted

4

3

PTC uncertain

1


1

2 2

1

Can we better predict intelligibility if we use psychoacoustic results?

Harvey Dillon

Audiology NOW! 2010


Likely intermediate effects

Yes, a little – speech deficit increases as frequency selectivity gets broader

Cognition

?

But not once we fully build HL into the SII prediction

Mechanical

PTC

Stria

OAE

Cardiovascular

Age

OHC TEN

Noise IHC

Implications for prescription

Why are hearing thresholds so useful? Frequency selectivity

Pure tone thresholds critical Knowledge of temporal resolution, frequency resolution, dead regions adds relatively little to prediction of intelligibility

Age and cognitive ability affect all frequency bands similarly  no effect on gain needed

National Acoustic Laboratories, Sydney, Australia

HL

Keidser and Dillon

Temporal resolution Speech Perception proficiency

Hearing thresholds Central auditory processing Age

Other

Cognitive ability


Keidser and Dillon

Gain; adults, medium input level (N = 187)

Gain preference over time N = 11

Source: Keidser, O’Brien, Yeend, & McLelland (submitted)


Harvey Dillon

Audiology NOW! 2010



Keidser and Dillon

Adjustments to prescription to allow for experience


Gain; adults, low and high input levels Smeds et al. 2006 Zakis et al. 2007

Preferred gain deviation from NAL-NL1 re gain preferred at 65 dB SPL in dB

Gain adjustment

1.5

12 10 8 6 4 2 0 -2 0 -4 -6 -8

New Experienced Experienced New

20

40

60

80

100

120

1 0.5 0 -0.5

-2

50

Input level in dB SPL

Keidser and Dillon

Compression ratio preferences: severe and profound hearing loss

Binaural loudness correction 7 Gain variation (dB)


1.1

1:1

1.0

Average 1/CR in LF band

Suggest that the compression ratio should be slightly higher, at least for clients with mild 80 and moderate hearing loss

-1 -1.5

Hearing threshold (dB HL)

0.9 0.8 0.7 0.6

1.8:1

0.5

6 5 4 3 2 1 0

0.4

3:1

0.3 0.2 45

Keidser and Dillon

50

55

60

65

70

75

80

85

0

20

40

90

60

80

100

120

Input level (dB)

Average HTL in LF band (dB HL) Source: Keidser, Dillon, Dyrlund, Carter, and Hartley (2007)

RECD in infants (own mold; HA2)

Directional microphones for infants and toddlers?

35 30 25 Anna O’Brien

SNR improvement 0.9

Front

15

Side

Back Proportion forward-looking

10

2

5

0 -2

0

-4 -5 0.6

2.0 0.8

6.0 4.0

20.0 8.0

60.0 40.0 80.0

400.0

Age (months)


0.7 0.6 0.5 0.4 0.3 0.2 0.1

SC1

SC2

SC3

Scenario

200.0

NH HI

0.8

4

dB

RECD_4k

Teresa Ching

20

SC4

0.0 SC1

Yes ! More to gain than there is to lose.

SC2

SC3

SC4

Scenario

Ching, O’Brien et al, 2009

Harvey Dillon

Audiology NOW! 2010


Language at 12 months after fitting • Effect of age of fitting: p = 0.0001* • Effect of hearing loss: p < 0.0001* • Effect of prescription: p =0.9

Covariate means: F6AV3FAMD: 55.64857

Language skills at 3 yrs Effect of age of implant: p = 0.02

EC-AC*"Cat2Fit"*"Cat2Pres"; LS Mea ns EC-AC * Cat2 Fit: F(1, 1 70 )=7 .7 41 , p =.00 5 Vertical bars denote 0.95 confidence intervals

EC-AC*"Cat2CI"; LS Means Current effect: F(1, 32)=4.2223, p=.04814 "Cat2CI": F= 5.32, p = 0.02

120 EC-AC EC-AC

A12P_EC A12P_AC

EC-AC EC-AC

PLS-4 Standard score (%)

PLS-4 Standard score (%)

DSL

NAL

110

100

90

80

CA36P_EC CA36P_AC

100

80

60 (n = 55)

( n = 4 2)

(n = 57)

(n = 21)

< 6 mo

>= 6 mo

< 6 mo

>= 6 mo

70 Fitting age category

< 12 mo (n=9)

>=12 mo (n=25) Implant age

Fitting age category

Consonant perception at 55, 70, 80 dB

Preference rating in real life Preference rating - average of trials 3 and 4

Consonant perception

18

Current ef f ect: F(2, 92)=1.9551, p=.14739

SITE: Aus

Vertical bars denote 0.95 conf idence intervals

SITE: Can

16

1.00 14 12

No of obs

Proportion correct

NAL-NL1 DSL[i/o]

0.95 0.90

10

17

8 6

0.85

4

0.80

11 8

2

3

0

0.75

dsl

55

70

80

Level (dB SPL)

55

70 Level (dB SPL)

no pref

nal

dsl

T Ching, NAL, CRC CI&HA INN

Intelligibility in quiet

Wireless (e.g. FM)

Amplification / tone controls

Directional microphones

Feedback cancelling

Comfort & quality

Output level

Adaptive noise suppression

Children

NAL-NL1 Adults

Multi-program

ADRO

Convenience

WDRC

Bilateral manual control

Phone interface AutoOpen fittings Auto-telecoil program Implantable hearing aids Rechargeable battery Expansion Auto-gain Bilateral feedback control Alerting tones and messages adaptation Active occlusion reduction Trainability Bilateral auto control Echo reduction Data logging

Input level

Integrated RECD

Clinicians


nal

Preference rating

80

Intelligibility in noise

Empirical evidence: variations from NAL-NL1

no pref