Transcript Fitting and Evaluation of FM Systems for HA Users Guidelines for

Fitting and Evaluation
of FM Systems for HA
Users
Guidelines for Fitting and Evaluation
ANSI S3.22-Specification of Hearing Aid
Characteristics (2003 approved)
ANSI WG 81-Electroacoustic Evaluation of
Hearing Assistance Devices (Approval Pending)
AAA Task Force-Developing Guidelines for
Fitting and Evaluation of FM Systems
Assumptions in Fitting FM Systems

Properly Functioning Hearing Aid


Properly Fit Hearing Aid


Matched to the needs of the child or adult with
hearing loss
Properly Configured to Receive the FM Signal


Meets Specifications-ANSI S3.22
Contacts removed on Direct Audio Input
Connections
Binaural Ear Level System is Optimal
Arrangement

Microphones at Ear Level
1. The auditory signal should be as consistent
as possible


This facilitates learning
Ideally this is achieved with minimal changes
of equipment
2. There are different signal input levels to the
hearing aid mic and to the FM mic in typical
use situations


Typical input to HA mic - 65 dB SPL
Typical input to FM mic - 80 dB SPL
3. Think in terms of output rather than gain


Goal is to match SPL levels at the eardrum so that signal
is consistent (exceptions will be discussed)
Therefore, for same output, gains will differ because input
levels are different
EXAMPLE
using overall signal levels


Say that Mary hears best when level at the eardrum is
100 dB SPL
For the HA Fitting


65 dB SPL input + 35 dB gain = 100 dB SPL
For the FM Fitting

80 dB SPL input + 20 dB gain = 100 dB SPL
4. Use speech-weighted noise for frequency response
and use tone sweep for SSPL90

All of the energy in complex signal sums to an
overall level, there is not 70 dB at every frequency,
but 70 dB overall, therefore…..
The SPL at 1 kHz will be lower for 70 dB speech
weighted noise than for 70 dB sweep frequency
tones. But….
 For a pure tone sweep, there IS 70 dB at every
frequency

5. Use FM settings first, then HA, to match output for
HA and HA+FM conditions
 Ideally
want the child to put on HA and
not change settings when using FM
 Need to check electroacoustically to
determine if this is possible
6. Optimize the testing
 Try to do electroacoustic measures before patient arrives
 Perform real ear measures if possible
 When you have options to receive FM only &
FM+ENV, start first with the signal arrangement used
most often
EVALUATION OPTIONS
for
FM SYSTEMS
A. Audio-Visual Checks
B. Electroacoustic Measurements
Verifying Performance
Matching FM’s to HA’s
C. Real Ear Evaluation
Matching FM’s to HA’s
D. Behavioral Evaluation
Speech Recognition
NEW TERMINOLOGY


While in a sea of strips from the test box… it
occurred to me that we needed new terms
Start Simple….First Letter


E for Electroacoustic Measures
R for Real ear Measures
Then…


HA for Hearing aid
FM for FM system in FM mode
There are four basic evaluation
steps!
1) HA - 65 dB SPL complex signal
2) HA & FM - 80 dB SPL complex signal
3) HA – 90 dB pure tone signal
4) HA & FM – 90 dB pure tone signal
EHA65
EFM/HA80
EHA90
EFM/HA90
SIMPLIST CASE
MILD TO MODERATE HEARING LOSS
100
dB SPL
Amplified
Speech
Gain is
40 dB for
HA
Unamplified
HA Mic
Auditory
Area
Gain is
25 dB for
FM
Speech Input to
FM Mic
ENV signal
10 dB below
FM and still
Audible
0
ENV MIC
FM MIC
FM+ENV MIC
MORE DIFFICULT CASE
SEVERE TO PROFOUND HEARING
LOSS
dB SPL
100
Amplified
Speech
Gain is
60 dB for
HA due to
Feedback
Unamplified Speech Input to
HA Mic
0
ENV MIC
Auditory
Area
MORE DIFFICULT CASE
SEVERE TO PROFOUND HEARING
LOSS
dB SPL
100
Amplified
Speech
Auditory
Area
Gain is
60 dB for
HA due to
Feedback
Gain is
45 dB for
FM
Unamplified Speech Input to
HA Mic
FM Mic
ENV signal
10 dB below
FM and NOT
Audible
0
ENV MIC
FM MIC
FM+ENV MIC

MORE DIFFICULT CASE
SEVERE TO PROFOUND HEARING
LOSS
Charlie
So in this case, we want to amplify the FM signal
to a HIGHER intensity than that achieved with
the Hearing Aid
Electroacoustic Measurements
MORE DIFFICULT CASE
SEVERE TO PROFOUND HEARING
LOSS
100
Amplified
Speech
dB SPL
Gain is
60 dB for
HA due to
Feedback
Auditory
Area
Gain is
60 dB for
FM
Unamplified Speech Input to
HA Mic
FM Mic
ENV signal
10 dB below
FM and NOW
IS Audible
0
ENV MIC
FM MIC
FM+ENV MIC
MORE DIFFICULT CASE
SEVERE TO PROFOUND HEARING
LOSS
100
Amplified
Speech
dB SPL
Gain is
60 dB for
HA due to
Feedback
Auditory
Area
Gain is
60 dB for
FM
Unamplified Speech Input to
HA Mic
FM Mic
ENV signal
10 dB below
FM and NOW
IS Audible
0
ENV MIC
FM MIC
FM+ENV MIC
Let’s Consider ….
Steps to Verify HA and FM Outputs
Preliminary Setup
1. Set HA to User Settings
2. Set Test Equip for Complex Signal
3. Set Test Equip to display OUTPUT,
not gain
4. Go into Multicurve Function
Electroacoustic Measurements
Set up for testing HA alone
Steps to Verify HA and FM Outputs
Billie
Change setting to Curve 1-Put HA in test box
Curve 1. Test HA alone with 65 dB SPL complex
input
Electroacoustic Measurements
Set up for testing HA & FM
Steps to Verify HA and FM Outputs
Billie
Change setting to Curve 2-Put FM
Transmitter in test box, Attach FM
Receiver Boot to HA
Curve 2. Test HA&FM system (set to FM only)
with 80 dB SPL complex input
Electroacoustic Measurements
Steps to Verify HA and FM Outputs
Billie



Compare Curves 1 and 2 : EHA65 and
EFM/HA80
Adjust volume if necessary to match HA alone
output,
If a programmable hearing aid , make and new
program and refer to that as “FM Program”
Electroacoustic Measurements
Steps to Verify HA and FM Outputs
Checking Maximum Output- Billie
Change setting to Curve 3-Put HA in Test Box
Curve 3. Test HA alone with 90 dB SPL pure-tone sweep
input
Electroacoustic Measurements
Steps to Verify HA and FM Outputs
Change setting to Curve 4-Put FM
Transmitter in test box, HA with FM
Receiver Boot and 2cc coupler outside
Curve 4. Test HA&FM system (set to FM only)
with 90 dB SPL pure-tone sweep input
Match Max Output- Billie


Compare Curves 3 and 4, EHA90 and
EFM/HA90
Adjust max output if necessary in the “FM
Program”, to match HA output
Electroacoustic Measurements
SUMMARY
These curves are
closely aligned
except for the
high frequencies
at the typical
Input levels (1&2).
Additional high
frequency info. Is
often desirable.
Goals for Verfication

FM Advantage of about 10 dB


EFM/HA80 minus EHA65
FM Transparency of about 0 dB

EFM/HA90 minus EHA90
Real Ear Measurements

Same principles re: matching output apply

Main difference is to place FM mic as close to
reference mic as possible

Curves: RHA65, RFMHA80, RHA90, RFMHA90
Real Ear Measurements
Set up for Leveling
Set up for testing HA alone
Set up for testing HA & FM
EVALUATION OPTIONS
A. Audio-Visual Checks
B. Electroacoustic Measurements
Verifying Performance
Matching FM’s to HA’s
C. Real Ear Evaluation
Matching FM’s to HA’s
D. Behavioral Evaluation
Speech Recognition
Behavioral Measurements
There are two main reasons why it is
not advisable to do threshold testing
with the FM system.
The first one stems from the suggestion
to place the FM mic 6 inches from the
loudspeaker and place the child outside
the booth…..
Behavioral Measurements
Behavioral Measurements
1) There is too much variability in
vertical and horizontal dimensions
which may result in variability in
the signal levels arriving at the FM
mic.
Behavioral Measurements
Behavioral Measurements
2) In addition, the FM systems may
operate differently for low level
signals than for higher level ones
like speech because of
compression.
THEREFORE…..
Threshold testing with the FM system
is not recommended!
Rather……….
Behavioral Measurements
Speech Recognition Testing


Hearing Aid alone
FM System
FM Only
 FM+ENV Mode


Test each arrangement in Quiet and in
Background noise if time permits
Behavioral Measurements
Speech Recognition Testing





Select appropriate test with multiple lists
Set FM and HA controls to user settings
Place FM mic on examiner but turned off
Place HA and FM receiver on Patient
Seat Patient in soundbooth at 45 degrees
azimuth
Behavioral Measurements
INTERPRETATION
Relate Laboratory findings to Clinical Applications

How do you determine if a score is significantly low?
Or
significantly different from another score?



Thornton and Raffin (1978) determined
the degree to which a second speech
recognition score must vary from an
initial score to be significantly different
Critical differences for speech recognition
scores (Thornton and Raffin,1978)
Initial
90
80
70
60
50
40
30
20
100
50
25
10
81-9676-9872-100
50-100
68-8964-9256-9640-100
57-8152-8648-9230-90
47-7342-7836-8420-90
37-6332-6828-7610-90
27-5322-5816-6410-80
19-4314-4812-5610-70
11-328-36
4-44
0-60
Disproportionately Low Scores?

Data for these comparisons were based
on the 95% confidence limits of the PB
Max for NU6 word lists from a sample of
407 ears with a wide range of pure-tone
averages (500, 1000 and 2000 Hz)
(Dubno et al. 1995).




SPeech Recognition INTerpretation
SPRINT Chart
May be used:
1) to compare two speech recognition scores or
2) to compare PBmax score on NU6 lists to
normative data
SPRINT Chart for 25-Word NU6 Lists
0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80 84 88 92 96
P
T
A
d
B
H
L
0
4
8
12
16
20
24
28
32
36
40
44
48
52
56
60
64
68
72
76
80
84
88
92
96
0
4
8
12
16
20
24
28
32
36
40
44
48
52
56
60
64
68
72
76
80
84
88
92
96
0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80 84 88 92 96
First Percent Correct Score
S
e
c
o
n
d
P
e
r
c
e
n
t
S
c
o
r
e
SPRINT Chart for 50-Word NU6 Lists
0
10
20
30
40
50
60
70
80
90
0
0
10
10
20
20
30
30 e
S
c
o
n
d
P 40
T
A
40
d
B 50
H
L
P
e
r
c
50 e
n
t
60
60 c
S
o
r
e
70
70
80
80
90
90
0
10
20
30
40
50
60
First Percent Correct Score
70
80
90

Speech recognition scores are often
interpreted based on some personal internal
norms because validated normative
information is not conveniently available.
Speech Recognition Testing
ASHA Guidelines

Test Conditions
HA-Quiet (Speech at 55 dB HL)
 HA-Noise (Add Speech Noise at 50 dBHL)
 HA&FM-Noise (Keep levels, Turn on FM Mic)
 HA&FM-Quiet (Keep levels, Turn off noise)

Behavioral Measurements
Speech Recognition Testing
Sample Scores
 HA-Quiet
 HA-Noise
 HA+FM-Noise

If score decreases significantly, then FM gain too low
 HA+FM-Quiet

80%
50%
85%
80%
If score increases significantly, then FM gain too low
Behavioral Measurements
You thought EHA65, EFM/HA90,
EHAFM65, EFMHA80 were too
Much!

Hearing Aid Alone
 BHA55
 BHA55/50

Hearing Aid + FM System
 BFM/HA55/50
 BFM/HA55
Behavioral FM Verification
Case Illustration
Third Grader who resisted wearing FM system.
Testing done to show him benefits.
 Hearing Aid Alone
BHA55
 BHA55/50


90%
60%
Hearing Aid with FM
BFM/HA55/50 (unlit FM)
 BFM/HA55/50 (bilat FM)
 BFM/HA55

80%
90%
100%
Getting Desired Results

If the desired FM Advantage and FM
Transparency are not obtained, adjustments may
be necessary to Hearing Aid/FM settings via
Manual adjustments-screwdrivers and trimpots
 Programmable adjustments-software, Noah,
TOASTER?

WATCH WHAT YOU SAY!!!
SUMMARY



Goal of Fitting FM is to allow improved signalto-noise ratio so that signal from the FM
transmitter is higher intensity than background
noise
Electroacoustic Verification is necessary to
determine optimal signal to noise ratio
Behavioral Verification with FM/HA
combination useful for Counseling and
Motivating Student