1

DSL v5.0: A Presentation for the Ontario Infant Hearing Program Associates

Richard Seewald, Marlene Bagatto and Sheila MoodieNational Centre for Audiology

The University of Western OntarioLondon Ontario Canada

Early development . . . .

1980Goal: To develop a computer-assisted

electroacoustic-based procedure for the fitting of amplification in pre-verbal children.

The Desired Sensation Level (DSL) Method

Early Development . . . .

Looked to the literature for studies of . .• Speech perception in children• Preferred Listening levels and Loudness

discomfort in children

Speech perception (e.g., Erber & Witt 1977)

History of the DSL Method

• Version 1 was completed in 1984 (pencil and paper)

• The first software system (DSL v3.0) was released in 1990(DSL for fitting linear gain instruments)

• DSL[i/o] for Windows was released in 1996(DSL[i/o] v4.0 for linear and WDRC instruments)

• DSL m[i/o] v 5.0 was released to hearing instrument manufacturers in 2005

How well does this work???

??Validation Studies

2

Preferred Listening Levels of ChildrenWho Use Hearing Aids:

Comparison to Prescriptive Targets

Scollie, Seewald, Moodie and DekokJAAA 2000

Funding provided by Phonak AG

Preferred Listening Levels in Children

• N = 18 Mean age = 10 yearsMild to Profound SN hearing loss

• The subjects listened to average conversational speech and adjusted their VC to the level they preferred.

• The subjects preferred VC setting, for an average speech input, was compared to DSL & NAL prescribed settings

Preferred Listening Levels in Children

PLL / DSL Comparison

0

20

40

60

80

Rec

omm

ende

d Li

sten

ing

Leve

l (dB

)

0 20 40 60 80 Preferred Listening Level (dB)

DSL

Preferred Listening Levels in Children

PLL / DSL Comparison

• On average, the DSL prescribed setting was 2 dB lower than the subject’s PLLs

• The PLLs were within 5 dB of the DSL prescribed settings for 70% of subjects

Preferred Listening Levels in Children

PLL / NAL Comparison

0

20

40

60

80

Rec

omm

ende

d Li

sten

ing

Leve

l (dB

)

0 20 40 60 80 Preferred Listening Level (dB)

NAL-RP/NL1

Preferred Listening Levels in Children

PLL / NAL Comparison

• On average, the NAL prescribed setting was 11 dB lower than the subject’s PLLs

• The PLLs were within 5 dB of the NAL prescribed settings for 9% of the subjects

3

Studies of Aided Speech Perception and Loudness with DSL[i/o] algorithm

• Jenstad et al. (1999) Speech perception

• Jenstad et al. (2000) Aided loudness measures

Studies of Aided Speech Perception and Loudness with DSL[i/o] algorithm

• Jenstad et al. (1999) Speech perception:

Mean sentence perception scores were high (>90% correct) across a wide range of speech input levels (i.e., 48 – 83 dB SPL)

Studies of Aided Speech Perception and Loudness with DSL[i/o] algorithm

Jenstad et al. (2000) Aided loudness measures:

Aided growth of loudness was normalized for speech and other environmental sounds through the DSL[i/o] algorithm.

DSL is just about right!DSL is just about right!

Different Prescriptionsfor Adults and Children

Background

Currently, there are currently differing opinions regarding the electroacoustic requirements for adults vs. children

• Prescriptive procedures developed for adults can be applied with young children (Ching, et al, 2001)

• Infants and young children require different prescriptive procedures (e.g., Stelmachowicz, 1991; 2000, Seewald et al. 1996).

4

Studies . . . .

• MacRae (1986)• Snik & Hombergen (1993)• Laurnagaray & Seewald (2005)

Relationships Between the Hearing Threshold Levels, Aided Threshold Levels and Aided

Speech Discrimination of Severely and Profoundly Deaf Children

John MacRaeNAL Report No. 107

January 1986

MacRae (1986)

• Subjects: 100 children with moderate through profound hearing loss.

• Measured the real-ear gain they were using and compared this to that prescribed by the NAL-R algorithm

MacRae (1986)

MacRae (1986) MacRae (1986)

5

Hearing Aid Fitting of Preschool and Primary School Children

Snik and HombergenScandinavian Audiology (1993)

Snik and Hombergen (1993)

• Subjects: 40 Adults; 60 Children

• Measured the preferred insertion gain as a function of the PTA hearing loss.

Snik and Hombergen (1993)

Adults Children

Snik and Hombergen (1993)

• The results showed that overall the mean use insertion gain was 7 dB less for the adults relative to that used by the children in this study

Adult / Child Preferred Listening LevelsAdult / Child Preferred Listening Levels

Scollie, Seewald, Cornelisse, Moodie, Bagatto, Laurnagaray, Beaulac & Pumford

Trends in Amplification (2005) Trends in Amplification (2005)

Procedure

• Subjects (n=72)– 24 children who were full-time hearing aid users– 24 adults with experience wearing hearing aids– 24 adults who were new hearing aid users

• Prescriptive targets DSL[i/o] 4.1• Participants had a variety of nonlinear hearing

instruments• New users provided with 15 to 20 day period of hearing

instrument use

6

Procedure

• Volume control wheel set to minimum

• Speech (RMS 60 dB SPL) via a loudspeaker in sound field

• Participant set VCW until talker sounded the best (repeated measure)

0

10

20

30

40

50

60

70

0 10 20 30 40 50 60 70

Recommended Listening Level

Pref

erre

d Li

sten

ing

Leve

l

0

10

20

30

40

50

60

70

0 10 20 30 40 50 60 70

Recommended Listening Level

Pref

erre

d Li

sten

ing

Leve

l

0

10

20

30

40

50

60

70

0 10 20 30 40 50 60 700

10

20

30

40

50

60

70

0 10 20 30 40 50 60 70

Adult/Child Preferred Listening Levels

□ Children

New Adult Users

Experienced Adult Users

Adult/Child Preferred Listening Levels

• Repeated Measures:

The average test-retest difference was1 dB for all three groups and no participant varied by more than 6 dB on repeated measures.

Adult/Child Preferred Listening Levels

The three groups differed in their PLLs:

• Children: The mean PLL was 2 dB below the DSL 4.1 target

• Experienced Adults: The mean PLL was 9 dB below the DSL 4.1 target

• New Adult Users: The mean PLL was 11 dB below the DSL 4.1 target

DSL[i/o] 4.1 is too loud !DSL[i/o] 4.1 is too loud !

DSLDSL®® v5.0: Hearing Aid Prescription for v5.0: Hearing Aid Prescription for Adults with Acquired Hearing Loss Adults with Acquired Hearing Loss

7

DSL v5.0: adult targets for range of speech inputs

60

52

74

0

10

20

30

40

50

60

70

80

90

100

110

120

130

140

100 1000 10000

Frequency (Hz)

Ear C

anal

Lev

el (d

B S

PL)

0

10

20

30

40

50

60

70

80

90

100

110

120

130

140

100 1000 100000

10

20

30

40

50

60

70

80

90

100

110

120

130

140

100 1000 100000

10

20

30

40

50

60

70

80

90

100

110

120

130

140

100 1000 10000

70

Preferred Listening Levels (PLLs) of Adults and Children Who Wear Hearing Aids Compared to DSL v5.0 Target Criteria

Seewald, Moodie, Scollie, Moodie & Polenko(200 ?)

Adult PLL / target comparisons study

• N = 9 adults with sensory hearing loss

• MEASURES:

1. RECD measures2. Set hearing aid in S-REM using DSL v5 target criteria for

REAR and RESR3. Set VCW to minimum, ask client to set to PLL using running

conversational speech 4. Measure S-REM at PLL setting5. Repeat6. Make desired electroacoustic changes per client report7. Measure S-REM at final setting8. Compare deviation from PLL

-20

-15

-10

-5

0

5

10

15

20

100 1000 10000

Frequency (Hz)

Dev

iatio

n fr

om P

LL (d

B)

-20

-15

-10

-5

0

5

10

15

20

100 1000 10000

Frequency (Hz)

Dev

iatio

n fr

om P

LL (d

B)

-20

-15

-10

-5

0

5

10

15

20

100 1000 10000

Frequency (Hz)

Dev

iatio

n fr

om P

LL (d

B)

PLL / target comparisonsPLL / target comparisonsDSL 4.1 DSL 5.0

Average deviation from PLL 500 – 4000 Hz:

DSL 4.1 = 12.1 dB

DSL 5.0 = 0.5 dB

Now that’s more like it !!!Now that’s more like it !!!

Implications for DSL v5.0

• DSL v5.0 acknowledges that adults and children not only require, but also prefer different listening levels, thus:

Thus, DSL v5.0 is the first generic hearing instrument prescription algorithm to prescribe different sensation levels of amplified speech for children with congenital or early onset hearing loss versus adults with late onset hearing loss.

8

DSLDSL®® v5.0: The New DSL Method for v5.0: The New DSL Method for Hearing Instrument Fitting Hearing Instrument Fitting

The DSL Method v5.0The DSL Method v5.0• Compatibility with ABR

Assessment Data• Updated RECD Norms

AUDIOMETRICASSESSMENT

• Multi-Level TargetsHEARING AIDVERIFICATION

• The DSL m[i/o] Algorithm• Multi-channel • Targets for Quiet & Noise• Modifications for Conductive

Losses and Binaural Fittings• Targets for Children & Adults

ELECTROACOUSTICPRESCRIPTION

The DSL Method v5.0The DSL Method v5.0

• Compatibility with ABR Assessment Data

• Updated RECD Norms• Updated RECD

measurement procedure

AUDIOMETRICASSESSMENT

Compatibility with ABRAssessment Data

nHL to eHL

• Diagnostic Audiologists correct ABR data by applying frequency-specific values

• Data reported in eHL on ISCIS form

• eHL data entered into DSL software for hearing aid fitting

Application in DSL v5.0

Three options:

• Enter nHL value and use DSL v5.0 default corrections

• Enter nHL value and use own correction values

• Enter eHL values– This assumes correction has already been applied

9

Setup

Ontario Infant Hearing Program Corrections

500 1000 2000 4000AC 15 10 5 0

*Equipment, parameter, and calibration specific

Updated RECD Predictions for Infants

Updated Average RECD Values

Real-Ear-to-Coupler Difference (RECD) Predictions as a Function of Age for Two

Coupling Procedures

Marlene Bagatto, Susan Scollie, Richard Seewald, Shane Moodie & Brenda Hoover

2002, JAAA, vol 13(8)

10

Predicted RECDs

DSL v4.1 predicted values:- foam tip coupling only- based on 12 month age ranges

Newly developed predictions:- foam tip and earmold coupling- to the nearest month

Predicted RECD Values

Subjects- 392 infants & children- ages 1 month to 16 years- 141 ears used immittance tips- 251 ears used earmolds- normal otoscopic and immittance findings

Results- high variability in RECD measures for children of the same age

Predicted RECD values: Predicted RECD values: EarmoldsEarmolds Updated RECD Predictions

i)

-10

-5

0

5

10

15

20

25

30

0 20 40 60 80 100

Age (months)

REC

D (d

B)

ii)

-10

-5

0

5

10

15

20

25

30

0 50 100 150 200 250

Age (months)R

ECD

(dB

)

Foam / immittance tips Earmolds

Predicted RECD values: Predicted RECD values: EarmoldsEarmolds Predicted RECD Values

Limitations:– all subjects had normal middle ear

function– high variability in RECD measures

associated with children of the same age

Therefore, whenever possible, predicted values should NOT replace the more precise RECD measurement.

11

Comparison of RECD Norms

RECD Norms: 2 month old

0

5

10

15

20

25

30

100 1000 10000Frequency (Hz)

REC

D (d

B)

Old Norm

New Norm

New IHP Procedure

• If have DSL v5.0, use of the table of RECD values is no longer necessary

• Choose foam tip or earmold coupling

• Measured RECDs are always preferable

The DSL Method v5.0The DSL Method v5.0• Compatibility with ABR

Assessment Data• Updated RECD Norms

AUDIOMETRICASSESSMENT

• Multi-Level TargetsHEARING AIDVERIFICATION

• The DSL m[i/o] Algorithm• Multi-channel • Targets for Quiet & Noise• Modifications for Conductive

Losses and Binaural Fittings• Targets for Children & Adults

ELECTROACOUSTICPRESCRIPTION

The DSL Method v5.0The DSL Method v5.0

• The DSL m[i/o] Algorithm• Multi-channel • Targets for Quiet & Noise• Modifications for Conductive Losses

and Binaural Fittings• Targets for Children & Adults

ELECTROACOUSTICPRESCRIPTION

The New DSLMulti-Stage Input-Output

Algorithm

The DSLv5.0 m[i/o] algorithm:The DSLv5.0 m[i/o] algorithm:

INPUT (dB SPL re: Free Field)

Expansion Stage Linear Stage Limiting Stage

Targets for OUTPUT (dB SPL in the Ear Canal)

WDRC Stage

12

0

20

40

60

80

100

120

140

0 20 40 60 80 100 120 140

`

Upper Limit of Comfort

Threshold

Input range for WDRC

The DSL m[i/o] algorithm

Input dB SF (SPL)

Out

put d

B R

E (S

PL)

WDRC: Compression Threshold

0

10

20

30

40

50

60

70

80

0 20 40 60 80 100

dB HL

Com

pres

sion

Thr

esho

ld (d

B S

PL)

Loud Speech

Average Speech

Soft Speech

In DSL v5.0, a channel-specific compression threshold prescription recommends a low compression threshold for milder hearing losses. The prescribed CT is about 50 dB SPL for moderate hearing losses, and is gradually raised as hearing level increases, to about 70 dB SPL for severe to profound losses.

DigitalProcessing

MultichannelCompression

DSL

H E A R I N G A I D V E R I F I C A T I O N

SELECTION

Working with MultichannelHearing Aids

Modifications for Conductive Hearing Loss

• Will increase the upper limit of comfort by 25% of the air-bone gap. Resulting targets are slightly more linear, and have slightly more gain.

• Targets are not allowed to exceed 140 dB SPL (ear canal).

Modifications for Conductive Hearing Loss

0

10

20

30

40

50

60

70

80

90

100

110

120

130

140

100 1000 10000

Frequency (Hz)

Ear C

anal

Lev

el (d

B S

PL)

0

10

20

30

40

50

60

70

80

90

100

110

120

130

140

100 1000 10000

Frequency (Hz)

Ear C

anal

Lev

el (d

B S

PL)

Targets for Aided Speech

Targets for Aided Speech with 15 dB conductive hearing loss

DSL 5.0: Binaural correction

0

10

20

30

40

50

60

70

80

90

100

110

120

130

140

100 1000 10000

Frequency (Hz)

Ear C

anal

Lev

el (d

B S

PL)

0

10

20

30

40

50

60

70

80

90

100

110

120

130

140

100 1000 10000

Frequency (Hz)

Ear C

anal

Lev

el (d

B S

PL)

Targets for Aided Speech for monaural fittingTargets for Aided Speech for binaural fitting

↓ 3 dB

13

Binaural Correction Notes

• The literature is not conclusive on whether a gain reduction is needed for binaural fittings in children. This element of DSL v5.0 requires further investigation. Until conclusive results are available, it is recommended that IHP associates choose ‘NO’ binaural correction even for those children fit with binaural aids.

MPO Targets

• The ULC target ( ) is really intended to be matched by fully saturated hearing aid responses, therefore a slightly lower target may be more appropriate for use with the MPO test signal. For this reason, the target input/output function within DSL v5.0 can be used to compute a level-dependent target for either 85 dB SPL (in the real ear) or 90 dB SPL (in the coupler, using simulated real ear measurement). This new target will be somewhat lower than ULC for most hearing losses.

MPO Targets

DSL v5: Comparison to DSL 4.1 Target Criteria

+ + + + ++ + + +*

* * * * * * * *

+ + + + ++ + + +*

* * * * * * * *

14

+ + + ++

++

+ +

+

+ + + ++

+ + +

+

+ ++

+

For Infants

For Young Children For Adults

15

Child Care Settings are Noisy Learning Environments

• Noise levels in occupied preschool classrooms in child care centers can range from 66 to 94 dB(A)

• Noise levels in occupied infant/ toddler classrooms range from 58 to 68 dB(A) and the SNR ranges from -3 to +12 dB.

• up to 50% of young children in child care have chronic problems with fluid in their middle ear, where they experience this condition almost 50% of the time during the first 3 years of life

Manlove, Frank & Vernon-Feagans (2001).

Seewald: Signal-to-Noise Considerations

• If one believes in an auditory approach to speech and language development then the following must occur:

1. The desired signal must be delivered within the child’s residual hearing capacity.

2. The speech signal must be perceived in order for the normal course and sequence of speech and language development to occur.

3. The speech signal cannot be constantly buried in a background of noise.

DSL v5: Prescriptions for Noisy Environments

• Importance-based reduction in gain– Loudness comfort– Speech Intelligibility

• Derived from the Speech Intelligibility Index (ANSI, 1997) 0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

100 1000 10000

Frequency (Hz)

Imp

ortan

ce

Prescription for Noisy Listening Environments

0

10

20

30

40

50

60

70

80

90

100

110

120

130

140

100 1000 10000

Frequency (Hz)

Ear C

anal

Lev

el (d

B S

PL)

Targets for QuietTargets for Noise

Prescription for Noisy Listening Environments

• The Noisy Listening Environment targets require validation

• It is not a minimum implementation requirement for manufacturers in this release of DSL v5

• Many manufacturers, including Audioscan, have not implemented these noise targets in their first release of DSL v5

Case Examples

16

• Sample verification data for hearing aid fitting to DSL v5.0 targets.

Outcomes

• 100 % vowel recognition

• 72% high frequency consonant recognition

• Loudness perception within normal range

• Real-world listening rated good to great on CHILD questionnaire

Evaluation of DSL v5.0

• 10 children– 7 to 17 yrs

• Mild to severe SNHL

• Binaural fittings– 55, 65, 75,

MPO

-20

0

20

40

60

80

100

120

100 1000 10000Frequency (Hz)

dB H

L

Preliminary Data from Children

• 3 preferred responses slightly over target in low to mid frequencies

• 2 preferred responses slightly under target in low frequencies

• 5 satisfied with hearing aid set to target

Loudness Outcomes

0

1

2

3

4

5

6

7

8

45 50 55 60 65 70 75 80 85

Input dB SPL

Loud

ness

Rat

ing

DSL v5.0 and the SII

• SII summarizes the proportion of audible speech

• SII in Verifit based on real-ear SPL thresholds and measured REAR of the aid

• Overall audibility present in the fitting as a percentage

17

DSL v5.0 and SII

• SII value does not indicate the proportion of speech that will be understood by the user

• Speech understanding depends on development, type of speech materials, listening environment, and other factors

What is a good SII?

• SII has a level distortion factor that is applied for speech levels higher than 62.4 dB SPL

• Applied in severe to profound fittings

• Makes SII value appear lower than one would expect

• 75 dB HL and better SII >60%• 80 dB HL and above SII 20 to 60%

0

10

20

30

40

50

60

70

80

90

100

20 40 60 80 100 120

Better Ear HF-PTA (dB HL)

Bet

ter E

ar S

II (%

)

75 dB SPL65 dB SPL55 dB SPL

IHP Protocol Review

• An individual with congenital hearing loss can be fit with DSL child targets regardless of age.

• Transducer is eHL • Measure RECDs or

choose average• Binaural is set to NO

e

IHP Verification Review IHP Verification Protocol

18

DSL Website Update

• Total website redesign will begin in January 2007

• Goal for completion: Late Spring 2007• Discussion: suggestions for website

Protocol Review

• Discussion re: challenges in clinical practice in following recommended IHP amplification protocol.