View
0
Download
0
Embed Size (px)
A FRESH approach to pediatric audiometry; Should we stop using NBN now?
Presentation given at AudiologyNOW!, Anaheim 2013
Johannes Lantz, Otometrics A/S, HEARsound Laboratories ‐ CopenhagenLund
Hannah Holmes, Institute of Sound and Vibration Research, SouthamptonSouthampton
What this is aboutWhat this is about 10
Frequency (kHz) 0.5 2 3 41
‐10
0
10
20
Average Left Ear HTLs as a Function of Stimuli20
30
40
50e l ( dB
)
as a Function of Stimuli
Pure tone
F50
60
70
80
H ea rin
g Le ve Frequency
modulated tone
Narrowband noise
FRESH80
90
100
FRESH
110
120
Masking NoiseMasking Noise Dr Robert Bárány, MD, Vienna (1876 to 1936)
se um
.c om
se um
.c om
he ar in ga id m us
he ar in ga id m us
“The Barany Box is inserted in the hearing ear and w w w .h
w w w .h
“It is scarcely necessary to enumerate the manyThe Barany Box is inserted in the hearing ear and creates a loud buzzing sound while the examiner shouts in the deaf ear to determine if the patient can hear anything. If the patient cannot hear the words shouted, then the ear is considered "Barany Deaf“”
It is scarcely necessary to enumerate the many objections to the use in audiometry of such masking devices as the Bárány noise box and jets of air or water. They are unpredictable in effect and awkward in use.” (Denes & Naunton 1952)then the ear is considered Barany Deaf.
(Source: www.hearingaidmuseum.com) (Denes & Naunton, 1952)
The sound level produced by this Barany Box measured a whopping 110 dB if ´ B D f b f110 dB, so if you weren´t Barany Deaf before…
NBN MaskingNBN Masking ’ h d f d d d• 1950’s: The advantage of NBN over Wide Band Noise was recognized
– Masking Efficiency: the relation between a sound's ability to mask and its loudness. A sound with high masking efficiency is one with good masking ability but minimal loudness.
– 1/3 octave Narrow Band Noise
dB
Wide Band Narrow Band
Hz
Narrow Band
Effective Masking Level CorrectionsEffective Masking Level Corrections
Critical Ratio: SNR at threshold when shifted by the noise
Since the NBN masking noise is wider than the critical band, some energy that is ”wasted” outside must be accounted for in the calibration
Amounts in Decibels (dB) to be Added to the Reference Equivalent Threshold Sound Pressure Level (RETSPL) to achieve Effective Masking (dB EM) for One‐Third Octave Band Masking Noises
(Extract from ANSI S3.6‐2004 American National Standard Specifications for Audiometers)
Standing Waves in Sound FieldStanding Waves in Sound Field ( ) d f• Pure tones (sine waves) gives rise to standing waves at certain frequencies
• What are the available options that may also be of interest for a hard‐to‐ impress toddler?p
• Warble and Narrow Band Noise have been used extensively for many years as sound field stimuli for pediatric audiometry h l i li l d h i i di i di• These popular stimuli also made their way into pediatric audiometry under headphones and inserts
Standing Waves in Sound Field H Dillon and G Walker Stimuli for audiometric testing
Standing Waves in Sound Field H. Dillon and G. Walker, Stimuli for audiometric testing. J. Acoust. Soc. Am., Vol. 71, No. 1, January 1982
More concerns about NBN filter slopesMore concerns about NBN filter slopes
• Orchik and Mosher (1975) – “…realize that the noise parameters, especially bandwidth and filter slope, can result in
a significant overestimate of threshold sensitivity in patients with sloping audiometric configurations.”
• Orchik and Rintelmann (1978) – ”...for subjects with sharply sloping high frequency sensorineural hearing losses...
...narrow band noise may substantially overestimate pure tone threshold sensitivity.”
• Stephens and Rintelmann (1978)
Average difference from normalized pure tone thresholds pertone thresholds per stimulus type for sharp configurations
Principle illustrated using the OTOsuite Hearing Loss Simulator
Pure tone stimulus
Test frequency
Principle illustrated using the OTOsuite Hearing Loss Simulator
FRESH noise stimulus
Test frequency
Principle illustrated using the OTOsuite Hearing Loss Simulator
NBN as stimulus
If we present a Narrow Band masking noise as stimulus atmasking noise as stimulus at the same level, the patient will respond to the circled area where the narrow band noise spills over into the audible range. Hence we will continue decreasing the stimulus level until the patient stops responding...
Assumed test frequency
Principle illustrated using the OTOsuite Hearing Loss Simulator
NBN as stimulus
The patient stops responding and weresponding and we mark the assumed threshold and thus underestimate the hearing loss
Assumed test frequency
Principle illustrated using the OTOsuite Hearing Loss Simulator
NBN as stimulus
This picture illustrates why there is no evident problem when the hearing loss isevident problem when the hearing loss is relatively flat. The stimulus remains within the inaudible range across all frequencies q
FREquency Specific Hearing noiseFREquency Specific Hearing noise h ” ” d f h d d (• The ”recipe” used for FRESH noise in the Madsen Astera Audiometer (GN Otometrics) from Walker, Dillon and Byrne (1984)
Sloping AudiogramsSloping Audiograms
(Walker, Dillon and Byrne, 1984)
FRESH Noise in Astera control panelsFRESH Noise in Astera control panels
Classic mode Sunshine mode
-15
-10
-5
0
p ea
k)
-45
-40
-35
-30
-25
-20 A
m p
lit u
d e
(d B
r e
NBN 100 1000 10000
-50
-45
Frequency (Hz)
5
0
NBN
-30
-25
-20
-15
-10
-5
tu d
e (d
B r
e p
ea k)
100 1000 10000 -50
-45
-40
-35
Frequency (Hz)
A m
p lit
Warble
-15
-10
-5
0
ea k)
-40
-35
-30
-25
-20
A m
p lit
u d
e (d
B r
e p
e
100 1000 10000 -50
-45
Frequency (Hz)
FRESH
Sound ExamplesSound Examples
NBN 500 Hz FRESH 500 Hz NBN 1000 Hz FRESH 1000 Hz
Pilot study Subject #1Subject #1
Pilot study Subject #1Subject #1
Erling B L1 NBN, R2 NBN
Pilot study Subject #1Subject #1
Erling B L1 NBN, R2 NBN
MATLAB ModellingMATLAB Modelling
0.1 1 10
Frequency (kHz)
b. Dead region: 1-4 kHz
0.1 1 10 -10
0
10
20 H
L )
30
40
50
60
h re
sh o
ld (d
B H
70
80
90
100
T
tone
NBN
FRESH
First subject at ISVRFirst subject at ISVR Frequency (kHz)
‐10
0
Frequency (kHz) 0.5 2 3 41
10
20
30
B)
Average Left Ear HTLs as a Function of Stimuli
Pure tone 40
50
60
ar in g Le ve l ( dB
Frequency modulated tone
Narrowband 70
80
90
H ea noise
FRESH
100
110
120
Consequence of NBNConsequence of NBN h h d d h l h l ll ?• What consequence may the underestimated hearing loss have clinically?
Underestimated hearing loss True hearing loss (NBN)
DSL 5 Aided Response TargetDSL 5 Aided Response Target
Underestimated hearing loss True hearing loss (NBN)
ReferencesReferences American National Standards Institute ANSI S 3 6 2004 American national standard specification forAmerican National Standards Institute. ANSI S 3.6‐2004, American national standard specification for
audiometers”
Denes, P., Naunton, R.F. (1952). Masking in Pure‐tone Audi