Linical & Experimental Audiology Speech-in-noise screening tests by internet; improving test sensitivity for noise-induced hearing loss Monique Leensen

linical & Experimental

Audiology

Speech-in-noise screening tests by internet; improving test sensitivity for noise-induced hearing loss

Monique Leensen

Wouter Dreschler

Academic Medical Centre, Amsterdam, The Netherlands

[email protected]


Audiology

2

27-2-2010

Screening

Early detection hearing impairment

Raising awareness about hearing loss

Self-assessment screeningtests:

• National Hearing Test by telephone (Smits et al. ’04)

• Online speech-in-noise tests

Easy to administer

Introduction


Audiology

3

27-2-2010

Online speech-in-noise tests

Online speech-in-noise tests

Adjustable presentation level

Spectrum of masking noise is equal to speech spectrum

• Independent of presentation level

Adaptive procedure

Fixed noise level

Speech Reception Treshold (SRT)

• 50% correct word recognition

Introduction


Audiology

4

27-2-2010

Oorcheck (LUMC ’04)

Target population:

youngsters (12-24 years)

Closed set of 9 CVC words

Introduction


Audiology

5

27-2-2010

Oorcheck and NIHL

Aims at youngsters

• Risk for hearing loss after exposure to loud music

• Raising awareness about hearing loss

However, NIHL does not greatly affect SRT in stationary

noise (Smoorenburg 1992)

• Closed set of speech stimuli

• Hearing threshold levels at low and middle frequencies are

(near) normal

Introduction


Audiology

6

27-2-2010

Objectives of this study

1. Investigate the potential of Oorcheck to discover NIHL

in an early stage

2. Examining ways to enhance the test sensitivity for mild

NIHL using different forms of noise modifications

Make existing speech-in-noise screening tests (like

Oorcheck) applicable for NIHL

Objectives


Audiology

7

27-2-2010


1. Investigate the potential of Oorcheck to discover NIHL

in an early stage





Objectives - 1


Audiology

8

27-2-2010

Subjects

n=100 (2 excluded)

NH: 49 (age 27 yrs ± 8.5)

with HTLs ≤ 15 dB HL

HI: 49 (age 56.3 yrs ± 9.4)

- 25 with narrow dip (ND)

- 24 with broad dip (BD)

HTL2kHz affected

Methods - 1


Audiology

9

27-2-2010

Oorcheck

SRT (dB snr)

-16-15-14-13-12-11-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0

HI : BD

HI : ND

NH

good poorin-sufficient

moderate

Test sensitivity for NIHL

Sensitivity 51%

49% HI is classified as NH

Specificity 90%

10% NH is classified as HI

Results - 1


Audiology

10

27-2-2010


1. Investigate the potential of Oorcheck to discover NIHL in an

early stage





Objectives - 2


Audiology

11

27-2-2010

Noise modifications

Spectral filtering

• Low pass filtered noise

Less masking of high frequencies

Temporal modulation

• NH benefit more from interrupted noise than HI (e.g. De Laat & Plomp 1983, Festen & Plomp 1990, Versfeld & Dreschler ‘02)

Both types of modifications combined

Noise parameters were based on model predictions (ESII)

and pilot measurements

Methods - 2


Audiology

12

27-2-2010

Experimental set of noises

test noise filtering modulation noise floor

1 Oorcheck - - -

2 16 Hz - 16 Hz -15 dB

3 LP LP 1.4 kHz - -15 dB

4 LP & 16 Hz LP 1.4 kHz 16 Hz -15 dB

5 HP HP 1.4 kHz - -15 dB

6 HP & 16 Hz HP 1.4 kHz 16 Hz -15 dB

7 OC -15 - - -15 dB

LP HP

Methods - 2


Audiology

13

27-2-2010

Results – group differences

Results - 2


Audiology

14

27-2-2010


Results - 2


Audiology

15

27-2-2010


Results - 2


Audiology

16

27-2-2010


Results - 2


Audiology

17

27-2-2010


Results - 2


Audiology

18

27-2-2010

Results - frequency dependence

Correlations between individual SRTs and HTLS

Highest correlation for LP noise and 2-6 kHz

Results - 2


Audiology

19

27-2-2010

Validation of OC LP

Increased sensitivity &

specificity:

sensitivity 95%

specificity 98%

Oorcheck - Low Pass noise

SRT (dB snr)

-26 -24 -22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0

HI: BD

HI : ND

NH

good in-sufficient

poor

Results - 2


Audiology

20

27-2-2010

Discussion

To be investigated further:

• Extrapolation of lab results

ambient noise

system settings

transducers

• Level dependency

Discussion


Audiology

21

27-2-2010

Conclusions

1. The original Oorcheck showed only small differences

between NH and HI subjects with mild NIHL (narrow dips)

2. Hearing-impaired subjects deviated more clearly from

normal performance when using low-pass filtered

masking noise

This suggests that this noise condition is more sensitive

to discover NIHL in an early stage

Conclusions


Audiology

22

27-2-2010

Generalization

We expect that the approach used will be valid for other

types of speech-in-noise tests as well

If this can be proven, manipulation of the masking noise will

increase the sensitivity and specificity with respect to NIHL

for different types of speech-in-noise screening tests

Conclusions


Audiology

23

27-2-2010

23

Thank you!

Questions?


Audiology

24

27-2-2010


Audiology

25

27-2-2010

Documents

Linical & Experimental Audiology Speech-in-noise screening tests by internet; improving test sensitivity for noise-induced hearing loss Monique Leensen