1/84
Looks like no tags are added yet.
What is “normal” hearing?
-Normal depends, among other things, on:
How you do your testing to determine “normal”
i.e. your protocol
Who you test to determine “normal”
screened or unscreened sample?
Screening using self report or records review to identify those with risk factors for hearing loss
Carefully controlled laboratory testing
-very precise, but a slow process. So generally small N’s (# of subjects)
-Careful subject screening/selection, instructions and practice
-Threshold often measured multiple times
subject selection criteria may be bias threshold estimates
Mass/survey testing
-Test a lot of people (large Ns) in a short time
-To do this you generally have:
A less controlled setting (e.g. state fair; clinic)
Less instruction/practice prior to testing
Threshold only measured once (or twice)
a less strict screening/selection process.
Screened population
-Screened excludes (based on self report) anyone with risk factors for hearing loss such as
occupational or leisure noise exposure
previous ear infections
head injury
known use of ototoxic medications
known hearing loss
self report of perceived hearing difficulties
“normal” hearing thresholds
-Today, normal hearing is defined by the RETSPL values (AKA audiometric zero) - the “reference” thresholds (in dB SPL) for normal hearing
RETSPL: Reference Equivalent Threshold in dB SPL
-Where do RETSPL values come from?
recall thresholds (the “T” is RETSPL) depend on
how you do the testing (carefully controlled vs mass survey)
And who you test (screened vs unscreened)
RETSPL
“Reference equivalent threshold sound pressure level”
-The mean/modal SPL at threshold (at a specified frequency, measured in a specified coupler), based on data from a large number of “otologically normal” adults (18-25/30 years old)
-Our American standard RETSPLs now mirror European International Standards Organization (ISO) standard RETSPLs
ISO 389-1 for supra-aural earphones
ISO 389-2 for insert earphones
Problem with thresholds in dB SPL
Large array of “normal” thresholds because “normal” values vary with the transducer/coupler combination
What measurement unit do we use on audiograms?
dB HL (dB hearing level)
dB hearing level (dB HL)
-The difference between the hearing threshold of an individual (in dB HL) and the RETSPL, also in dB SPL
i.e. individual threshold minus the mean/modal threshold of a large group of “otologically normal hearing” young adults
-0 dB HL is the softest sound that can elicit a response in a series of ascending trials at least 50% of the time (i.e. threshold) in “otologically normal” individuals
This is true regardless of the transducer or frequency
important points to remember about dB HL and SPL
-0 dB HL does not equal 0 dB SPL
-In dB SPL “normal” varies with frequency and transducer
we’re most sensitive in the 1-4 kHz range
-In dB HL “normal” is constant across freq and transducer
0dB
How to calculate dB HL
-Start with individuals “equivalent threshold in dB SPL” (ETSPL)
-subtract RETSPL from this value
-Difference is threshold in dB HL
Sensation level in dB (dB SPL)
-The level (in dB) above threshold for that sound
SL can be used to describe levels for simple sounds like pure tones
Hearing threshold definition
-Hearing threshold of an individual is “the lowest decibel hearing level (dB HL) at which responses occur in at least one half of a series of ascending trials.”
minimum of two, or three, responses at a single level
The audiogram
-Many different kinds but all have:
freq on the x-axis (octave scale)
Hearing level in dB on the y-axis (dB HL)
Spacing: 20 dB ~1 octave
-Additional space for
speech audiometry
acoustic immittance
-y-axis should also report source for “reference threshold”
Assessing the auditory system
-Behavioral pure tone thresholds
a primary component of the audiologic test battery
important for many purposes
-For diagnostic (site of lesion) purposes we measure and compare two thresholds obtained via the
air conduction pathway and bone conduction pathway
-Dysfunction at any point along these pathways results in HL
Air conduction pathway
-Air conduction (AC)
outer ear —> middle ear —> inner ear and beyond
-Elevated AC thresholds mean dysfunction somewhere along the pathway
but you dont know where!
-Bone conduction testing can help
Bone conduction pathway
-bypasses outer and middle ear
directly stimulates both cochlea simultaneously
-comparing AC and BC thresholds helps isolate site of dysfunction
Air conduction audiometry
-signal source is audiometer
-output transducer options
supra-aural headphones
e.g. TDH-39, 49, 50
insert earphones
e.g. ER-3A, ER-5A
circumaural headphones
e.g. Sennheiser HDA-200
Loudspeaker (sound field thresholds)
Bone conduction audiometry
-signal source is audiometer
-output transducer is a bone oscillator
e.g. radio ear B-71
-mastoid placement is typical
forehead placement may be used with kids, post-mastoidectomy patients, automated audiometers
correct calibration values are required
AC pure tone right ear symbol
red circle
AC pure tone left ear symbol
blue X
AC pure tone masked right ear
Red triangle
AC pure tone masked left ear
blue square
audiogram symbols: Pure tone bone
-place bone conduction (BC) symbols to the right (for the left ear) or left (for the right ear) of the frequency axis
unmasked BC mastoid symbol shows side of transducer placement (not the ear responding)
Or use the ear “unspecified” symbol “^” (placed on the frequency axis)
or place the oscillator on the forehead and use the forehead symbols
Bone conduction-mastoid unmasked left ear
>
Bone conduction-mastoid unmasked right ear
<
Bone conduction-mastoid masked left ear
]
Bone conduction-mastoid masked right ear
[
bone conduction - forehead unmasked
Carrot pointing down
bone conduction - forehead masked

audiogram symbols: no response
-Symbols for “no response” at highest output level
add a downward directional arrow to the existing symbol
to the left (right ear) or right (left ear) of the symbol
Do not connect with lines

* symbol
often used to suggest a “vibrotactile” response to BC
Symbols for unaided soundfield threshold
S
-reflects the threshold in the better ear
-If no response- do not connect with lines to actual responses
-Aided: often use “A” but must be defined on the audiogram
classifying HL
-terms reflect the expected (on average) magnitude of speech understanding problems that someone with that pure tone average (PTA) would have
normal limits
-10 to 15
slight HL
16-25
Mild HL
26-40
Moderate HL
41-55
Moderately-severe HL
56-70
Severe HL
71 to 90
profound HL
91+
Types of audiometers
-Described based on their capabilities
Pure tone generation (type 1, 2, 3, or 4)
Type 1-full diagnostic; Type 4- hearing conservation
Extended high frequency testing (Type HF)
8000-16000
Speech (type A, B, or C)
Input for recorded materials
Microphone for MLV
Earphone and/or loudspeaker output
-E.g. Type 1HFA
Frequency response
-The need for a flat frequency response depends on your purpose
Pure tone threshold measures versus complex stimuli (e.g. speech)
Maximum output
-Need varies with purpose
Screening audiometer versus diagnostic audiometer
-Max output also varies with frequency
-Increasing levels above the maximum creates distortion
Supra-aural
-TDH series (39,49, 50)
-MX 41 AR cushion
-Headband should be tight enough to hold paper between cushions
-Calibrated in a 6cc coupler
Supra-aural Pros
-reliable/durable
-inexpensive
-familiar
supra-aural cons
-poor frequency response
high frequency roll off (same as insert phones)
-Low frequency “leakage”
-Compared to inserts earphones
Less interaural attenuation ~40dB
Increased “occlusion” effect
-Can cause collapsing canal
Circumaural
cushions fit around (not on) the pinna
Circumaural pros
-Some have flat frequency response
-may be used with extended HF audiometry
-very comfortable
-better noise isolation than supra-aural
Circumaural cons
-modified calibration methods used
won’t fit on a 6cc coupler
-big on small heads
-Less interaural attenuation
insert earphones (ER 3A/5A/3C)
-Mimics frequency response of TDH 39 headphones
-calibrated in a 2cc coupler
insert earphone pros
-more hygienic
disposable foam earplugs
-More comfortable for adults and kids
-inter-aural attenuation ~50-70 dB
-Greater attenuation of ambient noise
-Protect against collapsing canals
-Less variability in threshold due to transducer placement
-Less artifact during electrophys testing (ER 3A/3C)
-Ease in relating behavioral audiometric data to hearing aid measures (both calibrated in a 2cc coupler)
insert headphones cons
-mimics frequency response of TDH 39 earphones
NOT FLAT frequency response!
-Costs! ~$400.00 pair
~$1.00 pair for disposable tips
-Cerumen can block receiver output
-connector is easily lost!
Bone conduction transducers
-Radio Ear B70A, 71, 72, and B81 (new) are common
Headband tension affects force (levels)
-Calibrated using an “artificial mastoid”
-Compared AC responses frequency responses are peaky and don’t have much low frequency
-Lower output and dynamic range compared to AC transducers
~30-50 dB more output via AC
-Newer “B81” model gives more LF output with less distortion than B71
Audiologists describe threshold results in terms of:
-Degree of loss
normal, mild, … profound
-Configuration of loss - how thresholds vary with frequency
flat, sloping…
-Type of loss (or site of lesion- outer, middle, or inner ear, neural or higher)
conductive, sensorineural or mixed
-Symmetry of loss - hearing loss is (or not) similar between ears
unilateral loss is asymmetric
bilateral loss can be symmetric or asymmetric
Configuration of loss
-no consensus on how to define configuration
-flat loss
-sloping loss
-rising/reverse slope loss
-trough loss
-peak(ed) loss
Flat loss
-thresholds (in general) ± 20 dB
Sloping loss
-general trend of poorer thresholds as frequency increases
LF<HF by >20 dB
May add qualifiers e.g. mild, moderate, or severe based on thresholds
steeply sloping loss
a large change over a narrow frequency range
ski slope loss
precipitous loss
rising/reverse slope loss
-same as sloping but reverse direction
HF<LF by >20 dB
trough loss
-worst thresholds in mid-frequencies
cookie bite loss
Peak(ed) loss
-best thresholds in mid frequencies
reverse cookie bite loss
inverted V audiogram
Conductive loss
-bone conduction thresholds are normal and are better than air conduction thresholds
-conductive losses have an air-bone gap (ABG)
ABG (in dB) =AC threshold-BC threshold
-Conductive loss means the site of lesion is in the outer or middle ear
an ABG>10 dB is significant even if AC thresholds are normal
Sensory/neural (sensory or neural)
-also sensorineural & SNHL
-with SNHL AC and BC thresholds are similar (ABG<10 dB)
-generally BC<AC (meaning BC is better than AC) thresholds but BC thresholds can be worse
-Sensory (cochlea) or Neural (VIII nerve) or higher
Mixed loss
-Bone conduction thresholds are not normal
-ABG is >10
-Site of lesion is outer/middle ear and cochlea (or higher)
-Any HL that has both a conductive and SN component in the same ear
Why do we care about asymmetry?
-consistency
-Diagnostic utility
Asymmetric defintion
-Between ear difference of
>10 dB at 3 or more frequencies
>15 dB at 2 or more frequencies
>20 dB at 1 or more frequencies
-These criterion were based on best match to “expert” judges
optimizing consistency not diagnostic utility
-Pretty conservative criteria resulting in high prevalence of asymmetry (i.e. >50% cases)
Diagnostic utility of asymmetry
-Asymmetric SNHL can be a red flag for vestibular schwannoma (VS)
but no standardized criterion/definition of asymmetry for detecting such tumors
-Asymmetry alone is not an accurate diagnostic indicator of retrocochlear pathology
-But in conjunction with other test results, the presence of asymmetry can (and should) influence your clinical decision.
How much asymmetry is “too much?”
-Answer varies across clinicians and sites
-Starting out remember Mangham theory: Average difference >20 at 1,2,4 and 8 kHz was one of the most accurate predictors
But used alone, sensitivity is not great, and false positives will be high! This will be the case with any asymmetry criteria
-Remember, asymmetry alone is not an accurate diagnostic indicator of retrocochlear pathology
-PTA protocols cannot be used for proper screening or diagnosis of vestibular schwannoma despite their good sensitivity, and MRI remains the gold standard for this purpose
Measuring thresholds
-Multiple methods can be used to measure thresholds (e.g. hearing or visual thresholds)
-Current methods are based on early work by Gustav Fechner
He developed the “method of limits”
basis for current clinical method
Method of limits (MOL)
-A series of ascending and descending runs. Each run consists of a series of “yes-no” trials
After each trial, the subject says “yes” or “no”
-The threshold is the average of “crossover points”
-A valid and reliable method, but it’s time-consuming, and it’s susceptible to subject bias and criterion effects
Subject bias: anticipation bias, habituation bias
Subject criterion: strict, lax
Current threshold search procedure (ASHA)
-Current method is a “modified method of limits”
First pass- Hughson and Westlake
modification of MOL but only “ascending” runs contribute to threshold
Current version- Carhart and Jerger “modify Hughson and Westlake” procedure
Keep only ascending responses contribute to threshold criteria
suggest 5 dB step size
Use 1-2 second stimulus duration with quiet intervals in between
Terminate ascending run at first response
decrease level and start next ascending run
Threshold defintion
“the lowest hearing level at which responses occur in 1) at least one-half of a series of 2) ascending trials, with a 3) minimum of two responses out of the three required at a single level.” ANSI
-Repeat at 2, 3, 4, 6, and 8 kHz
recheck 1000 Hz. if thresholds differ by >5 dB use best threshold and recheck threshold at another frequency
Finish, in order, 500, 250, and then 125 Hz (if needed)
Minimal response level (MRL)
-Infants and young children are not “little adults”
-MRL: lowest level of stimulus that produces the desired response
actual thresholds may improve over time
stimulus characteristics
-1-2 seconds duration (all types)
Vary the time intervals between presentations
no shorter than 1-2 seconds
-Pulsed tones (~200-250 ms on/off duration) may reduce false positives and be “preferred”
-Warble tones (or narrow band noises) are required for sound field testing but can affect thresholds, especially for steeply sloping losses
Response characteristics
-Multiple options are available based on what the patient can do
Press a button
Raise a finger, arm
verbal response
any consistent response that is time locked to your stimulus presentation is acceptable
-Your interpretation of a patient response can affect your threshold
patient response and your interpretation must be consistent
slow responders, “off” responders are o.k.
look for a faster response at higher levels
watch for false positives or false negatives
reinstruct as needed
Dealing with false positives
-False positive: Person says “yes I hear it” when no tone was presented
Increase signal level to remind them of the tone
Make sure your timing of presentations is varied
Reinstruct or vary instruction
e.g. respond “ONLY” if you are sure you hear the tone
Try pulsed or warbled (FM) tones
Dealing with false negatives
-Person doesn’t respond “yes” even though the signal is above threshold
Could have a conservative criterion
reinstruct
e.g. make sure you respond even if it is very faint
Could be giving you a response that you are not seeing
watch the person not just the response light
Look for body movement (e.g. partial hand raise)
Change test method
say yes instead of hand raise
May need a physiologic measure
Test-retest variability
-We assume clinical thresholds are reliable ± X dB
Assumes a 5 dB step size and a 5 dB standard deviation of difference values
~38% of diff’s = 0dB
1 st. deviation
~86% of diff’s ± 5 dB
2 st. deviation
~98% of diff’s ± 10 dB
3 st. deviation
-Threshold differences >10 dB rarely reflect simple test-retest differences
Intrinsic factors affecting thresholds
-Breathing/movement/cardiovascular activity
-cognitive factors
e.g. attention, developmental delay
-Criterion (patient)
-Bias (tester)
Extrinsic factors affecting thresholds
-Environment: temperature/Ambient noise levels
-Equipment issues: calibration
-Procedural factors
instructions, response task, and your interpretation of response and others
Additional factors affecting thresholds
-Poor supra-aural earphone placement
especially a problem for
children and when measuring high-freq thresholds
Center receiver over ear canal
adjust position for comfort and loudness
-Insert earphone placement
inserts should be deeply inserted
-Bone oscillator placement
again, big problem for kids (oscillator is too big)
Move it around until subject hear tone the loudest
check headband tension
this weakens over time (reduces applied force & signal)
-Vibrotactile response
-Tester Bias
-Background noise
Vibrotactile response
-Most problematic for BC testing, especially at LFs
Can lead to “false” ABGs at freq <1 kHZ
-Can occur via AC at high levels
~80-120 dB HL at 250-1000 Hz
Tester bias
-Assumptions can bias your measurement
Background noise affecting thresholds
-ANSI standard for Maximum permissible noise levels (MPNLs) tells you “how much is too much”
-MPNLs limit noise levels such that negligible masking (<2 dB) occurs for signals presented at RETSPLs
i.e. 0 dB HL; suggests measuring thresholds that are better than the RETSPL may not always be possible
-MPNLs vary based on transducer type
supra-aural versus inserts (MPNLs “ears covered”)
Sound field/BC (MPNLs “ears not covered”)
Higher MPNLs for inserts vs. supra-aurals
Lowest MPNLs for “ears not covered”
Test booth
-Sound treated, not soundproof
custom built or modular
-floor may be spring mounted
reduce vibrations
-double or single walled
double walled booths provide greater ambient noise reduction
but are more expensive