Note: Descriptions are shown in the official language in which they were submitted.
CA 02235067 1998-06-19
-1-
SPATIALIZATION FOR HEARING EVALUATION
BACKGROUND OF THE INVENTION
TECHNICAL FIELD
The present invention relates to hearing evaluation and hearing aid fitting.
More
particularly, the present invention relates to virtual electroacoustic
audiometry for unaided,
simulated aided, and aided hearing evaluation.
DESCRIPTION OF THE PRIOR ART
The human auditory system processes sounds from a complex three-dimensional
space via
the external, middle, and inner ear, as well as via the complex neural
pathways that lead to
the auditory cortex within the brain. A measurable hearing loss, due to
various conductive,
sensorineural, or central auditory disorders, affects a significant percentage
of the human
population, particularly elderly persons. Rehabilitation via hearing aids
remains the only
viable option for those types of hearing impairments that cannot otherwise be
medically
treated or surgically alleviated.
Advances in hearing aids and fitting technologies are continuously being made.
Today's
ear-level hearing aids, i.e. in-the-ear (ITE), behind-the-ear (BTE), in-the-
canal (ITC), and
completely-in-the-canal (CIC) types, are more cosmetically appealing due to
improvements
in electronic and mechanical miniaturization. More significant, however, is
the increasing
availability of advanced hearing aid signal processing schemes, such as
adaptive filtering
and mufti-band dynamic compression.
As manufacturers are continuously developing new hearing aids with unique
signal
processing schemes, a hearing aid dispensing professional is faced with the
increasingly
difficult task of prescribing and selecting a hearing aid for a hearing-
impaired individual
from the available selection. A cursory look at available hearing aid
processing schemes
reveals an impressive array of categories, sub-categories, and associated
acronyms that are
baffling to most hearing aid dispensing professionals (see Mueller, H.G., A
Practical
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126 .
-2-
Guide To Today's Bonanza of lJnderused High-Tech Hearing Products, The
Hearing Journal, vol. 46, no. 3, pp. 13-27, 1993).
Today, optimal fitting of prescription hearing aids remains an elusive goal in
,
auditory rehabilitation. The fundamental problem is that there are numerous
electrical, acoustic, physical, and other parameters that affect hearing aid .
performance. These parameters include signal processing schemes,
electronic circuit adjustments, size of hearing aid, insertion depth, venting
size, patient controls, and life-style related factors that must be considered
't 0 when prescribing and fitting a hearing aid. These hearing aid parameters
are not only complex and highly interrelated, but also vary according to the
unique interaction of the hearing device with the hearing-impaired
individual.
y 5 Generally, the in situ performance characteristics of a hearing aid cannot
be
predicted with today's conventional fitting instrumentation and methods.
Dissatisfaction among hearing aid user's, partially due to poor hearing aid
prescription fitting, is manifested by a high return rates, often exceeding
20%
according to industry reports.
Factors that Contribute to Unsatisfacto~i Hearing Aid Results
1. Inaccuracy of conventional diagnostic audiometry.
Assessment of hearing is the first step in the prescribing and fitting of a
hearing aid. Accurate assessment of the individual's hearing function is
important because all hearing aid prescriptive formulas depend on one or
more sets of hearing diagnostic data (see Mueller, H. G., Hawkins, D. B.,
Northern, J. L., Probe Microphone Measurements: Hearing Aid Selection
and Assessment, Singular Publishing Group, lnc., 1992: Ch. 5).
The hearing aid prescription process involves translating the diagnostic data
into target hearing aid electroacoustic parameters that are used in the
selection of the hearing aid. Traditional hearing evaluation methods and
instruments employ a variety of air-conduction transducers for coupling
acoustic signals into the ear. Commonly used transducers include supra-
aural earphones, such as TDH-39, TDH-49, TDH-50, insert earphones, such
as ER-3A, and free-field speakers {see Specification of Audiometers, ANSI-
S3.6-1989, American Standards National institute).
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-3-
A threshold measurement obtained with such transducers is referenced to a
mean threshold obtained by testing a group of otologically normal
. individuals. This mean threshold, by definition, is referred to as the zero
decibel hearing-level or 0 dB HL. With this zero reference concept,
threshold measurements of otologically normal persons can vary by 20 dB or
more. These variations can be attributed to following factors:
1. Variability d~e to transducer type used and placement with respect to
the ear.
In a study by Mowrer, et a! discrepancies of 10 dB were found in 36% of
threshold measurements (see Mowrer, D. E., Stearns, C., Threshold
measurement variability among hearing aid dispensers, Hearing Instrument,
vol. 43, No. 4, 1992). Another major disadvantage of measurements
obtained using a traditional transducer is that results are not
iriterchangeable with measurements taken with another transducer for a
given individual (see Gauthier, E. A., Rapisadri, D. A., A Threshold is a
Threshold is a Threshold... or is it?, Hearing Instruments, vol. 43, no. 3,
1992).
2. Variability due to transducer calibration methods that employ couplers
that do not represent the human ear.
Although recently developed couplers more closely match the acoustic
impedance characteristics of an average human ear, there is still
disagreement as to the accuracy of this artificial ear (see Katz, J., Handbook
of Clinical Audioioav, Third Edition, 1985, pp. 126). Most calibration
methods today rely on 6-cc or 2-cc couplers that are known to have
considerable acoustic characteristic discrepancies from real human ears
(see Specification of Audiometers, ANSI-S3.6-1989, American Standards
National Institute). Furthermore, even if an agreement was made regarding
an average artificial ear, variability among individuals is significant due to
individual acoustic characteristics of pinna, ear canal, concha, and to a
lesser extent, the head, and the torso (see Mueller, H. G., Hawkins, D. B.,
Northern, J. L., Probe Microphone Measurements: Hearing Aid Selection
end Assessment. 1992, pp. 49-50). In one study, inter-subject variability was
up to 38 dB across six standard audiometric frequencies when sound
pressure levels (SPL) were measured at the tympanic membrane for 50 ears
CA 02235067 1998-04-16
WO 97!23117 PCT/US96lI3126
_q._
of 25 adults (see Valente, M., Potts, L., Valente, M., Vass, B., Intersubject
Variability of Real-Ear SPL:TDH-39P vs ER- 3A Earphones, In press. JASA).
3. Conventional audiometric measurement methods do not provide a ,
means of self-calibration even though transducer characteristics are
known for changes due to wear or damage of the moving diaphragm. ,
Clinicians who use regular subjective listening methods simply cannot
detect gradual changes in transducer sensitivity.
Although errors due to the above factors are not likely to be accumulative in
all cases, the potential for substantial errors is always present.
Furthermore,
these errors are not consistent across all frequencies and therefore cannot
be simply compensated for during the fitting process via an overall volume
adjustment.
Il. Lack of realistic listening conditions in the unaided and aided hearing
assessment.
1. Lack of Binaural Advantage Considerations.
Many studies have demonstrated the advantage of binaural versus
monaural listening (see Cherry, E. C., Some Experiments on the
Recognition of Speech with One and Two Ears, JASA, vol. 25, no. 5, 1953,
pp. 975-979; Cherry, E. C., and Tylor, W, K., Some Further Experiments on
the Recognition of Speech with One and Two Ears, JASA, vol. 26, 1954, pp.
549-554). These studies have focused on the advantages offered by the
Binaural Masking Level Difference (BMLD) and Binaural Intelligibility Level
Difference (B1LD).
Early studies of BMLD and BILD involved the presentation of signal and
noise to one or both ears at various phase relationships. Tone detection and
speech intelligibility were shown to vary as much as 15 dB, depending on
the signal/noise phase relationship. Even though many of these studies
suggest the significance of binaural considerations, today's hearing
assessment methods, unaided and aided, primarily deal with monaural test
conditions, i.e. testing one ear at a time.
CA 02235067 1998-04-16
WO 97/23117 PCTlUS96/I3126
_g_
2. Lack of Spatiaiized Sound Considerations.
When audiometric signals such as speech and/or noise are delivered to the
ear via a conventional audiometers and associated transducers, the sound
perception by the test subject is not localized to any particular point in
space
(see Specification of Audiometers, ANSI-S3.6-1989, American Standards
National Institute). For example, in speech audiometry evaluation, the
speech stimuli level is adjusted for one ear and speech noise level is
separately adjusted in the opposite ear. The test subject perceives sounds to
be within the head and localization is limited to left/right direction. This
type
of signal presentation and perception is referred to as intracranial and is
unlike the way humans normally perceive natural sounds. Recent studies by
Bronkhorst and Plomp, and Begault expanded on previous binaural
interaction advantage studies by employing headphone localization
techniques (see Bronkhorst, A. W., Plomp, R., The Effects of Head-Induced
Interaural Time and Level Differences on Speech intelligibility in Noise,
Journal of the Acoustical Society of America, vol. 83, no. 4, 1988, pp. 1508-
1516; Bronkhorst, A. W.; Plomp, R., The Effects of Multiple Speech-like
Maskers on Binaural Speech Recognition in Normal and Impaired Hearing,
Journal of the Acoustical Society of America, vol. vol. 92, no. 6, 1992, pp.
3132-3139; and Bagault, D. R., Call Sign Intelligibility Improvement Using a
Spatial Auditory Display, Ames Research Center, NASA Technical
Memorandum 104014, April 1993). The results of these studies conclude the
speech perception is not only dependent on intensity levels but also on the
spatial relationship between speech and noise.
3. Lack of Evaluation Methods in Realistic Listening Environments.
Speech intelligibility and discrimination deteriorates in the presence of
competing speech and other environmental sounds. Furthermore, the
acoustic properties of a room, e.g. its walls and objects within the room, all
play an important role in the filtering process subjected to the original
signal
source. These filtering effects are especially significant for hearing-
impaired
individuals who typically have a limited frequency response and dynamic
range in their hearing function.
Today's methods of presenting competing and environmental sounds via
conventional transducers fail to represent the acoustic reality of the typical
listening condition. Recorded sound material presented via tape players,
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-6-
compact disks, or computer digital playback are subject to filtering effects
of
the transducer employed andlor the room acoustics of the clinical setup.
There are no hearing assessment methods today that can evaluate or
predict the hearing performance of an individual in a specific and realistic ,
listening scenario.
For example, the hearing performance of a hearing-impaired child in a
typical classroom in the unaided condition, and the hearing performance of
the child with a specific hearing aid, i.e. aided hearing, in the same
classroom environment. These and other auditory experiences are
presently considered a fact of life that can not be dealt with in a clinical
setup
{see Mueller, H. G., Hawkins, D. B., Northern, J. L, Probe Microphone
lVleasurements~ Hearing Aid Selection and Assessment, 1992, pp. 69}.
lll. Limitations of current real-ear measurement {REM) equipment and
methods.
In recent years, real ear measurement (REM) systems were developed to
assess the in situ performance of a hearing aid. REM consists of test probe
measurements of the ear response to free field stimulus, i.e. speakers, taken
at the tympanic membrane. A secondary reference microphone is typically
placed outside the ear canal close to the ear cans( opening. The reference
microphone is used to calibrate the test probe as well as to regulate the
stimulus level as the head moves with respect to the free field speaker.
For a comprehensive REM evaluation, measurement of the real ear
response for the unaided, i.e. open canal, condition is first taken. Target
hearing aid characteristics are then calculated based on the natural ear
canal response characteristics, as well as other criteria (see Mueller, H. G.,
Hawkins, D. B., Northern, J. L., Probe Microphone Measurements: Hearing
Aid Selection and Assessment, 7992, Ch. 5). When the hearing aid is
prescribed, ordered, and received during a subsequent visit, the aid is
inserted over the probe tube and adjusted to match the prescribed target
hearing aid characteristics.
REM evaluation and REM-based prescriptive methods provide considerable
improvements over previous fitting methods which relied on the combination
of audiometric data and hearing aid 2-cc coupler specifications. Although
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-7-
REM offers insight into the in situ performance of the hearing aid, it suffers
from several fundamental problems, as described below:
1. REM test results vary considerably depending on speaker
position/orientation with respect to the ear, particularly at higher
frequencies
(see Mueller, H. G., Hawkins, D. B., Northern, J. L., Probe Microahone
nnoauurements~ Hearinc,~Aid~~e, lection and Assessment, 1992, pp. 72-74).
2. Real ear measurements are taken with a specific stimulus type,
source-ear distancelorientation, and room acoustics. The specific test
condition may not represent realistic listening scenarios encountered by
hearing aid users. In fact, using conventional REM approaches, a hearing
aid may be optimized for a specific listening condition while compromising
the performance under other conditions that may be more important to the
hearing-impaired individual.
3. Accurate REMs require careful placement of the test probe within the
ear canal of an individual. The closer the probe to the tympanic membrane,
the more accurate the results are, particularly for high frequency
measurements (see Mueller, H. G., Hawkins, D. B., Northern, J. L., Probe
Microphone Measurements' Hearing Aid Selection and Assessment, 1992,
pp. 74-79).
Present methods of probe placement are highly dependent on the operating
clinician's skill and the specific length of the canal, which is about 25 mm
for
the average adult. Today's REM methods rely on visual observation of the
probe tip. This is especially problematic when a hearing aid is placed in the
canal during the aided evaluation process. The only exception to the
conventional visual method is the acoustic response method developed by
Nicalet Corp. for use in the Aurora system (see Chan, J., Geisler, C.,
Estimation of Eardrum Acoustic Pressure and Far Canal Length from
Remote Points in the Canal, J. Acoust. Soc. Am. 87 (3), March 1990, pp.
1237-1247; and U.S. Patent #4,809,708, Method and Apparatus for Real Ear
Measurements, March 1989). However, Nicolet's acoustic response method
requires two cafibration measurements prior to placement of the probe at the
desired position within the ear canal.
4. REM test results vary considerably depending on the placement of the
reference microphone near the ear. The errors are especially significant at
CA 02235067 1998-04-16
WO 9?/23117 PCT/US96lI3126
_g_
frequencies of 6 kHz and higher (see Mueller, H. G., Hawkins, D. B.,
Northern, J. L., Probe Microphone Meas~rements~ Hearing Aid Selection
and Assessment, 1992, pp. 72-74).
5. REM instruments employ sound field speakers in a room with ambient
background noise that often exceeds 50 dB SPL across standard
audiometric frequencies. This necessitates stimulus levels of 60 dB or
higher to produce measurements having sufficient signal-to-noise ratios.
This is problematic it hearing aid performance characterization under low
level acoustic stimuli is required.
IV. The problem of correlating diagnostic, prescription formulae, and real ear
measurements.
A significant factor that contributes to the results of a hearing aid fitting
is the
problem of adequately correlating diagnostic data with fitting needs of the
hearing-impaired individual. Diagnostic measurements are typically taken in
dB HL with transducers that are calibrated in 6-cc couplers. Hearing aid
specification and performance measurements employ 2-cc couplers which
do not represent the real-ear. Fitting involves the use of one of several
prescriptive formulae, with results that are known to vary as much as 15 dB
for the same diagnostic data across standard audiometric frequencies {see
Mueller, H. G., Hawkins, D. B., Northern, J. L., Probe Microphone
Measurements' Hearing Aid Selection and Assessment, 1992, p 107). These
fitting formulae incorporate statistically based conversion factors that
simplify
the correlation of hearing aid requirements to a particular hearing
impairment. However, averaged conversion factors are known to vary
considerably with respect to objectively measured individual conversion
factors.
Several methods and protocols have been suggested to alleviate errors
associated with measurement errors and data correlation (see Sandberg, R.,
McSpaden, J., Allen, D., Real Measurement from Real Ear Equipment.
Nearing Instruments, Vol. 42, No. 3, 1991, pp. 17-18). However, many of
these protocols have not yet been widely accepted due to limitations of
conventional audiometry and Real-Ear Measurement (REM) equipment and
other factors related to efficiency of the proposed protocols in clinical
setups.
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
_g_
Hearing rehabilitation through the use of hearing aids remains the only
viable option for many hearing impaired individuals who cannot be
medically or otherwise treated. A full audiometric evaluation is a required
first step prior to fitting a hearing aid. Pure tones and one or more speech
perception tests are typically involved in the basic audiometric test battery.
Suprathreshold measurements may also be taken to establish a hearing
dynamic range profile, in addition to the frequency response profile obtained
in the threshold audiogram test . Following the audiometric evaluation, a
hearing aid is then prescribed, selected, ordered, and subsequently tried
and adjusted after being received from the manufacturer or assembled in the
clinic. The fitting or determination of the electroacoustic parameters of a
hearing aid typically involve a combination of objective measurements to
achieve a desired target characteristics based on one of many prescriptive
formulae and subjective measures based on the individual's subjective
response to speech and other sounds at various loudness levels..
Conventional audiometry methods, employing headphones, inserts, or
sound-field speakers, rely on presenting acoustic energy to the ear of the
individual in a manner which is not representative of sound delivery under
realistic listening conditions. Conventional audiometers present various
tones, speech, and noise stimuli to each ear individually and thus are not
capable of investigating the individual's binaural integration advantage, or
of
assessing the hearing function in a three-dimensional sound environment.
Another major disadvantage of conventional audiometry methods is the
inability of such methods to assess accurately and objectively, in absolute
physicaE terms such as dB SPL, the hearing function of an individual with
respect to the inside of the ear canal to correlate unaided evaluation results
to hearing aid requirements. One exception is the probe-mike-calibrated
fitting system developed by Ensoniq, which only addresses testing accuracy
(see Gauthier, E. A., Rapisadri, D. A., A Threshold is a Threshold is a
Threshold.., or is it?: Hearing Instruments, vol. 43, no. 3, 1992).
Furthermore, conventional audiometry instruments and methods are not
capable of simulating the electroacoustic performance of one or more
prescribed hearing aids and assessing their simulated function in realistic
acoustic conditions relevant to the individual's unique listening
requirements.
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-10-
The master hearing aid concept, which gained some popularity in the '70s
and '80s, involves an instrument that presents simulated hearing aids to the
hearing aid user (see Selection InstrumentationlMaster Hearing Aids in
Review, Hearing Instruments, Vol. 39, No. 3, 1988). Veroba et al (U.S Patent
#4,759,070, Patient Controlled Master Hearing Aid, Jul. 19, 1988) describe
a patient controlled hearing aid module that is inserted into the ear canal
and connected to a test module which offers multiple signal processing
options, e.g. analog circuit blocks, to the individual. Hearing aid
characteristics are determined by a tournament process of elimination, while
the hearing-impaired person is presented with real-word sounds played
back from tape decks via a set of speakers located around the hearing
impaired person's head. The system's fitting process is based on subjective
responses of the hearing-impaired who must continuously decide on an
alternative signal processing option, and supposedly eventually arrive at an
optimal fitting.
The fitting process via the Veroba system, commercially known as the
Programmable Auditory Comparator, an essentially obsolete product, does
not involve any objective measurements or calculations for selecting and
fitting of the hearing aid. In fact, the entire fitting process is based on
the
subjective response of the hearing impaired person. Clearly, most hearing
impaired individuals, on their own, cannot explore in a timely and efficient
manner the spectrum of various complex and interrelated electroacoustic
parameters of a hearing aid under various listening environments. A serious
limitation of Veroba is that it does not teach how to assess objectively the
performance of the simulated hearing aid, nor does it teach how the aided
performance is related to the individual's unaided response determined
previously during the audiometric evaluation process.
A major unsubstantiated claim in Veroba's system is the simulation of a
realistic acoustical environment via tape-deck playback and speakers
located around the head of the hearing-impaired individual. However,
recorded acoustic signals that are played back are further subjected to
acoustic modifications due to speaker characteristics, speaker position with
respect to ear/head, and acoustic characteristics of the room, i.e. wall
reflections and acoustic absorption. Without factoring in all of the specific
acoustic modifiers in the transmission channel between the tape-deck and
the individual's ear, a realistic listening condition cannot be achieved with
Veroba or any such system. Furthermore, Veroba is not capable of
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/I3126
-11-
manipulating the acoustic condition from its recorded form, e.g. by projecting
an audio source in a specific location within a three-dimensional acoustic
space with a specific acoustic boundary condition.
Another hearing aid simulator, the lTS-hearing aid simulator developed by
Breakthrough, Inc. offers computer digital audio playback of digital
recordings obtained from the output of various hearing aids (see ITS-
Nearing Aid Simulator, Product brochure, Breakthrough, lnc., 1993). Each
recording segment represents a specific acoustic input, listening scenario,
hearing aid model, and hearing aid electroacoustic setting. The recording
segments require memory space either on a hard disk or other known forms
of memory storage devices, such as compact-disk read-only-memory. This
digital-recording-based approach renders impractical the arbitrary selection
of a hearing aid, hearing aid setting, and input stimulus for a hearing-
impaired individual, when considering all the possible combinations.
Furthermore, the effects of hearing aid vent sizes, and associated occlusion
effect, insertion depth, and individual external ears, cannot be simulated
with
the proposed hearing aid simulator because it relies on conventional
transducers, i.e. headphones and insert earphones.
For similar reasons, many other commercially available master hearing aid
systems, do not have the ability to simulate accurately a hearing aid in a
realistic listening environment. Furthermore, these systems do not include
objective measurement methods for evaluating simulated aided versus
unaided conditions. Far these and other reasons, virtually ail dispensed
hearing aids today are fitted without the use of master hearing aid or hearing
aid simulator instruments.
State-of-the-art REM equipment allows for in-the-ear-canal acoustic
response measurements. The acoustic stimuli are typically generated by the
REM equipment itself and delivered via a speaker, typically positioned at
0°
azimuth, or with two speakers positioned at 45° azimuth, with the
respect to
_ the transverse plane of the head. The response measurements, i.e. free-field
to real-ear transfer function, are essentially one-dimensional since they only
provide a single transfer function per ear in a particular speaker-ear
relationship, and are thus not capable of establishing a multi-dimensional
profile of the real-ear response. Another disadvantage of conventional REM
equipment and methods is the lack of real speech stimuli presentation
because most REM equipment only offer pure-tone, pure-tone sweep,
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-12-
speech-noise and other speech-like stimuli. These stimuli do not explore
responses to particular speech segments that may be important to the
hearing-impaired individual during unaided and aided conditions.
Recent developments relating to electroacoustic hearing aid measures
involve the testing of hearing aids in more realistic conditions. Real speech
.
signals instead of pure tones and speech-like noise signals were employed
in a recommended test protocol; and spectrogram plots indicating temporal,
i.e. time, analysis of the acoustic energy in dB SPL versus frequency was
compared for hearing aid input versus output (see Jamieson, D., Consumer
Based Electroacoustic Hearing Aid Measures, JSLPA Suppl. 1, Jan. 1993).
The limitations of the proposed protocol include: limited acoustic reality due
to the specified sound delivery method via a speaker to a hearing aid in an
enclosed chamber; and limited value of the spectrogram plots which do not
directly indicate the relationship of the plot to audibility and loudness
discomfort.
Other recent developments involve three-dimensional sound presentation
via headphone transducers (see Wightman, F. L., Kistler, D. J., Headphone
Simulation of Free-Field Listening. I: Stimulus Synthesis, JASA. vol. 85, no.
2, 1989, pp. 858-867; and Wightman, F. L., Kistler, D. J., Headphone
Simulation of Free-Field Listening. II: Psychophysical Validation, JASA. vol.
85, no. 2, 1989, pp. 868-878). These three-dimensional effects are achieved
by recreating the in-the-ear-canal acoustic response to free-field signals via
headphones or speakers (see U.S. Patent No. 4,118,599, Stereophonic
Sound Reproduction System, Oct. 3, 1978; U.S. Patent No. 4,219,696,
Sound Image Localization Control System, Aug. 26, 1980; U.S. Patent No.
5, i 73,944, Head Related Transfer Function Pseudo-Stereophony, Dec. 22,
1992; U.S. Patent No. 4,139,728, Signal Processing Circuit, Feb. 13, 1979;
and U.S. Patent No. 4,774,515, Altitude Indicator, Sep. 27, 1988). This
involves digital filtering of source signals based on head-related-transfer-
function (HRTF). The HRTF, essentially real-ear unaided response (REUR)
in three-dimensional space, is a frequency dependent amplitude and time
delay measurement that results from head shadowing, pinna, concha, and
ear canals. The HRTF enables externalization of localized sound with
headphones. Source signals that are processed with HRTF provide the
listener with free-field listening experience according to the controls of the
signal processing parameters.
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-13
Present research and development efforts in three-dimensional audio is
mainly focused on commercial musical recordings, playback enhancement,
and human-machine interface enhancement (see Bagault, D. R., Call Sign
Intelligibility Improvement Using a Spatial Auditory Display, Ames Research
Center, NASA Technical Memorandum 104014, April 1993; and Begauit, D.,
- Wenzel, E., Headphone Localization of Speech, Human Factors, 25 (2), pp.
361-376, 1993) and virtual reality systems (see The Beachtron-Three
dimensional audio for PC-compatibles, reference manual, Crystal River
Engineering, Inc., Revision D, Nov., 1993). The object of these three
dimensional audio systems has been limited to simulating situational
awareness in an approximate virtual acoustic environment since non-
individuafized HRTF set is typically employed.
The application of three-dimensional audio in objective in-the-ear-canal
assessment of hearing in the unaided, simulated aided, and aided
conditions would be a significant and extremely helpful departure from
known audiometric techniques.
SUMMARY OF THE INVENTION
The invention provides a virtual electroacoustic audiometer (VEA), which is a
system used in the assessment of human hearing function in the unaided,
simulated aided, and aided conditions. A pair of intro-canal prostheses (ICP)
are placed in the two ear canals of an individual to deliver acoustic stimuli.
A
probe measurement system, partially inserted in the ICP, measures the in-
the-ear-canal response conditions near the tympanic membrane during all
hearing evaluation, thus providing a common reference point for correlating
responses in the unaided, simulated aided, and aided evaluation conditions.
A unique modular hearing aid defined in accordance with the results of such
hearing assessment is also provided that includes highly configurable
electroacoustic and electronic signal processing elements.
During unaided evaluation, the system performs audiometric tests, such as
pure tone thresholds, uncomfortable loudness levels (UCL), speech
reception threshold, and speech discrimination. These peripheral hearing
tests, as well as other central auditory processing (CAP) tests, evaluate the
hearing function of the human in response to acoustic stimuli measured near
the tympanic membrane in absolute sound pressure level (SPL) terms,
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/I3126
-14-
unlike conventional stimuli which are presented in relative hearing level (HL)
terms.
Another significant feature of the VEA is its ability to synthesize, or
create,
acoustic signals that are representative of signals received in real listening
environments in a three-dimensional space. This is achieved by
incorporating the various filtering effects of room acoustics, atmospheric
absorption, spreading loss, interaural delay, and spectral shaping of external
ear, and other body effects. For example, a listening condition representing a
teacher-talker in classroom is digitally synthesized and acoustically
delivered via the ICP to a child to assess hislher unaided and aided listening
ability in a classroom environment. Spatialized competing signals
representing school children noise is optionally presented in addition to the
spatiaiized primary speech signal, i.e. the teacher, to assess further the
1 b child's speech discrimination ability in the presence of background noise.
The unaided evaluation method involves both ears in the listening
experience similar to the way humans normally hear sounds, with each ear
receiving a portion of the acoustic energy according to the relationship
between each ear and the various virtual audio sources. in contrast,
conventional audiometry methods present intracranial acoustic stimuli to
each ear individually, for example, speech to one ear, and competing noise
in the opposite ear.
The simulated aided assessment of the VEA system is accomplished by
incorporating the electroacoustic performance of a desired hearing aid into
the unaided digital synthesis of acoustic signals. The simulated hearing aid
eiectroacoustic parameters include microphone and receiver transfer
functions, and amplifier and fitter characteristics.
Specific or generalized acoustic models are digitally presented to the input
of the simulated hearing aid process. Specific acoustic models represent
listening scenarios that are important to the individual under evaluation and
that may be selected and manipulated by the operating clinician, for
example a teacher-talker source model in a classroom environment model
with a specific source-ear relationship. A typical goal in such a specific
scenario is to maximize speech intelligibility by optimizing the
electroacoustic characteristics of the simulated hearing aid. Generalized
acoustic conditions represent listening scenarios that are associated with
CA 02235067 1998-04-16
WO 97/23117 PCTlUS96/13126
-15-
normative response data. An example of a generalized model is an
audiofogic word list, such as W-22, having a specific spatialized background
noise. Test scores are compared with general model normative data stored
in the system's memory.
The VEA system also simulates other hearing aid effects that can not be
simulated by the digital synthesis process due to the unique effects of the
individual ear. These include the occlusion effect, venting size, and
oscillatory feedback potential. The occlusion effect is a phenomenon that
results in changes to the perceived characteristics of the individual's own
voice when the ear canal is occluded with a hearing aid.
!n addition, the VEA system offers a method of measuring various
individualized acoustic transfer functions in a three-dimensional space,
which are incorporated during the various synthesis processes to create
virtual acoustic conditions for an individual.
BRIEF DESCRIPTION OF THE DRAWINGS
Fig. 1 is a block level schematic diagram showing the major components of
the VEA system, including dual ICP prostheses inserted in the ear canal of
an individual; a probe microphone system; and a computer system including
a digital audio synthesizer module, a digital audiometer module, and a
virtual acoustic space measurement module according to the invention;
Fig. 2 is a block level schematic diagram of a digital audio synthesizer
module according to the invention;
Fig. 3 is a block level schematic diagram of a digital audiometer module
according to the invention;
Fig. 4 is a block level schematic diagram of a virtual acoustic space
measurement module according to the invention;
Fig. 5 is a block level schematic diagram of a virtual acoustic space
measurement system according to the invention;
Fig. 6 is a perspective view of an adjustable chair used for positioning a
patient's head during virtual acoustic space testing;
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/I3126
-16-
Fig. 7 is a schematic diagram showing speaker arrangement in a virtual
acoustic space measurement system, including transverse plane speakers,
and sagittal plane speakers according to the invention;
Fig. 8 is a schematic diagram showing an example of transfer function
interpolation at a point i3 from transfer functions, measured at points my and
m2 in a two-dimensional transverse plane according to the invention;
Fig. 9 is a schematic diagram showing an example of realization of a realistic
listening scenario for unaided hearing evaluation conditions, and in
particular showing a teacher-talker/child-listener scenario including direct
acoustic paths PR y and P~ i and early reflection paths PR2 and PL2 to
the right and left ears of the child-listener according to the invention;
20
Fig. 10 is a block level schematic diagram showing an example of realization
of a realistic listening scenario for unaided hearing evaluation conditions,
and in particular showing a process representation of a teacher-talker/child-
listener scenario during unaided evaluation according to the invention;
Fig. 11 is a partially sectioned, perspective view showing an intra-canal
prosthesis (ICP) for an ICP-ITE representing hearing aids for shallow ear
canal placement according to the invention;
Fig. 12 is a partially sectioned, perspective view showing an intra-canal
prosthesis (ICP) for an ICP-1TC representing hearing aids for deep ear canal
placement according to the invention;
Fig. 13 is a perspective view showing an intra-canal prosthesis (ICP) face-
plate end, including face-plate probe tube holders and probe tube
placement according to the invention;
Fig. 14 is a partially sectioned, side view showing an ICP core module for a
two-part ICP configuration according to the invention;
Fig. 15 is a partially sectioned, side view showing adjustable vent inserts
and an ICP-ITE sleeve for an ICP-ITE configuration according to the
invention;
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-17-
Fig. 16 is a partially sectioned, side view showing an ICP-ITC sleeve for a
two-part ICP configuration according to the invention;
Fig. 17 is a partially sectioned, side view showing a complete two-part ICP-
ITC assembly according to the invention;
Fig. 18 is a partially sectioned, side view showing an ICP having a
programmable vent according to the invention;
Fig. 19 is a partially sectioned, side view showing a hearing aid and direct
acoustic coupling method to an iCP, including direct acoustic coupling via a
magnetic attraction method according to the invention;
Fig. 20 is a partially sectioned, side view showing a hearing aid and direct
acoustic coupling method to an ICP, including direct acoustic coupling via
an acoustic coupler method according to the invention;
Fig. 21 is a partially sectioned, side view showing a hearing aid and direct
acoustic coupling method to an ICP, including a programming and acoustic
coupling interface according to the invention;
Fig. 22 is a partially sectioned, side view showing a hearing aid and acoustic
coupling to an ICP via an acoustic coupler tip according to the invention;
Fig 23 is a block level schematic diagram showing an example of a fitting
process provided by the virtual electroacoustic audiometer system according
to the invention;
Fig. 24 is a graphic computer generated display showing a reference
measurements module according to the invention;
Fig. 25 is a graphic computer generated display showing an unaided
evaluation module according to the invention;
38 Fig. 26 is a graphic computer generated display showing a predicted aided
module according to the invention;
Fig. 27 is a graphic computer generated display showing a simulated aided
evaluation module according to the invention;
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/I3126
-18-
Fig. 28 is a graphic computer generated display showing an aided
evaluation module according to the invention;
Fig. 29 is a line graph plotting the variability of measured SPL versus
distance of probe tip from tympanic membrane for 5 kHz and 15 kHz tones
for an individual according to the invention;
Fig. 30 is a bar graph plotting the measured SPL for 5 kHz and 15 kHz
during probe advancing at 6 mm from tympanic membrane according to the
invention;
Fig. 31 is a bar graph plotting the measured SPL for 5 kHz and 15 kHz
during probe advancing at 5 mm from tympanic membrane according to the
invention;
Fig. 32 is a bar graph plotting the measured SPL for 5 kHz and 15 kHz
during probe advancing at 4 mm from tympanic membrane according to the
invention;
Fig. 33 is a block level schematic diagram showing an example of a teacher-
talker/child-listener scenario using predicted aided evaluation for the right
ear according to the invention;
Fig. 34 is a block level schematic diagram showing an example of a teacher-
talker/child-listener scenario using simulated aided evaluation for the right
ear according to the invention;
Fig. 35 is a block level schematic diagram showing a simulated hearing aid
with directional microphone according to the invention;
Fig. 36 is a block level schematic diagram showing an example of the
realization of realistic listening scenarios for aided hearing evaluation
conditions according to the invention; and
Fig. 37 is a block level schematic diagram showing an example if the
prediction and simulation of oscillatory feedback of a simulated hearing aid.
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
_ig_
DETAILED DESCRIPTION OF THE INVENTION
For purposes of the description herein, the following definitions shall be
consistently applied:
Windaw: Refers to a graphical area displayed on a computer screen, that
represents a collection of controls, objects, entry fields, and plots, that
are
grouped together according to a logical functional manner.
iconized: Refers to an active window that is shown as an icon. Its display
is disabled but may be enabled by clicking on the icon on the computer
screen.
The virtual electroacoustic audiometer (VEA) described herein is a unitary
instrument that is used in the hearing assessment in the unaided, simulated
aided, and aided conditions. The VEA also offers new methods for hearing
aid fitting and analysis using a combination of digital synthesis of realistic
acoustic stimuli and in-the-ear-canal response measurements throughout
the assessment and fitting processes.
Fig. 1 shows the main components of the preferred embodiment of the VEA
system 15. A pair of intra-canal prostheses (tCP) 22 is inserted in the ear
canal 21 of an individual for delivering acoustic stimuli 25 in a manner
similar to that of a hearing aid. Each ICP contains a receiver, i.e. a
speaker,
for transmitting acoustic signals to the tympanic membrane 26. The 1CP also
contains a probe tube 24 for measuring the acoustic response that results
from the unique interaction of the receiver-produced acoustic stimuli and the
ear-canal characteristics of the individual. A probe microphone system
consisting of a probe tube 24 and probe microphone 23 measures acoustic
signals from the ear canal 21 and provides electrical signals representative
of the acoustic signals. A response keyboard 27 is provided to register a
response from the test subject 20 during various hearing evaluation tests.
Each ICP receiver 22 is electrically connected to a digital audiometer
module 19 that provides an interface to various audiometric transducers
including the 1CP receiver 22 and probe measurement system 23. The digital
audiometer module is connected to a digital audio synthesizer module 18
and a virtual acoustic space measurement module 14 via various inter-
module cables. The virtual acoustic space measurement module includes an
CA 02235067 1998-04-16
WO 97/23117 PCTlUS96/13126
-20-
output terminal 16 for connection to a plurality of test speakers. These
modules may be contained at or within a standard personal computer (PC)
11 that also contains standard computer accessories such as memory
storage devices 17, a display monitor 10, a keyboard 12, and a mouse 13. .
Memory storage devices are collectively referred to as system memory 17.
Block diagrams of the digital audio synthesizer, digital audiometer and
virtual acoustic space measurement modules are shown in Figs. 2, 3, and 4.
In the exemplary embodiment of the invention, the digital audio synthesizer,
digital audiometer, and virtual acoustic space measurement modules are
connected to the personal computer system via the Industry Standard
Architecture (1SA)-bus interface 34 and ISA-bus 39 of the personal computer
(see, for example Fig. 2). Digital data representing audio sources are
retrieved from the system memory via the bus interface 34, and are digitally
processed by a digital signal processor 33 within the digital audio
synthesizer module 18. The digitally processed data are then converted to
analog form using an digital-to-analog converter 35 that typically operates at
conversion rate of 44.1 kHz, or at another rate depending on the desired
signal bandwidth required.
The digital audio synthesizer module also receives analog signals
representing audio signals via its input connector 31 from external audio
sources such as tape or CD players (not shown). Received analog signals
are converted to digital signals by the analog-to-digital converter 32 for
signal processing via digital signal processor 33.
Multiple digital audio synthesizer modules (not shown) may be used to
enhance the system's digital signal processing capability. This is
particularly
useful for parallel real-time binaural signal synthesis. Multiple digital
audio
synthesizer modules are cascaded by connecting the output 38 of one digital
audio synthesizer module to the auxiliary input 30, or input 31 of another
digital audio synthesizer module. The internal and auxiliary signals are
combined within the module at a summing node 36 prior to output. In the
preferred embodiment of the invention, two digital audio synthesizer
modules are used. Each module employs a Motorola DSP56001 digital
signal processor clocked at 40 MHz.
CA 02235067 1998-04-16
WO 97/23117 PCTIUS96/13126
-21-
The analog output 38 from the digital audio synthesizer module 18 is routed
to the mixer 45 of the digital audiometer module 19 (Fig. 3) via a connector
42. Analog audio signals received at the digital audiometer module are
. mixed via mixer circuit 45, amplified via an audio amplifier circuit 46, and
impedance matched and routed to various audiometric transducers via an
_ audiometric transducer interface circuit 49. Outputs to audiometric
transducers include ICPs 50 (discussed above, and in further detail below),
bone vibrators 51 (not shown), a headphone 52 (not shown), arid other
conventional methods of delivering sounds to the ear of an individual.
Amplified signals from the audio amplifier 46 are also sent to the digital
audio synthesizer module input 31 from an audio buffer circuit 47 output
connection 48. The mixer circuit 45 also includes connections for receiving
audio signals from ICP microphones 55, an operating clinician microphone
56 (not shown), and a patient microphone 57 (not shown), via a microphone
amplifier 58.
External line-level signals received at input connectors 53 are also amplified
via an amplifier 54 and sent to mixer circuit 45. A response keypad interface
circuit 60 is employed to interface the system to the response keypad via a
connector 59 to register an individual's response to acoustic stimuli during
various audiometric evaluation processes. The operating clinician
microphone, connected to the digital audiometer module, allows the
operating clinician to communicate with the patient via the ICP pair. The
patient microphone allows the patient to communicate back to the operating
clinician during certain audiometric tests that require verbal responses from
the patient. The patient microphone is also used in occlusion effect
measurements, as are described in more detail below.
The digital audiometer module also includes a PC-BUS connection 43 and
PC-BUS interface circuit 44 that fink the digital audiometer module to the
VEA to coordinate module operation at the system level.
The VEA also includes a virtual acoustic space measurement system (Fig. 5)
that is used to evaluate the individual's acoustic transfer function set. A
block diagram of the virtual acoustic space measurement module 14 is
shown in Fig. 4. The virtual acoustic space measurement module receives
electrical signals, representing various acoustic signals, from the digital
audio synthesizer module output connectors 38 via a set of input connectors
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-22-
64. fnput signal level adjustment and routing is accomplished via a mixer
circuit 65, an audio amplifier circuit 66, and a speaker routing and interface
circuit 71. The output of the virtual acoustic space measurement module is
thence coupled to various test speakers in a speaker array 16.
The virtual acoustic space measurement module also includes a PC-BUS
connection 68 and PC-BUS interface circuit 67 that link the virtual acoustic
space measurement module to the VEA to coordinate module operation at
the system level. Such coordination includes processing information
indicative of patient head position connected to the module from a patient
head positioning sensor via a connector 70 and a positioning sensor
interface circuit 69.
An adjustable chair 78 is preferably used to ensure proper ear positioning
within the measurement space, as shown in Fig 6. A vertical adjustment
lever 79 adjusts the vertical position of the individual on the chair. A back
adjustment knob 81 adjusts a chair back support 80. The head support 82 is
adjustable to support the head of the individual seated on the chair. An ear
position reference arm 84 provides a target reference by pointing a set of
ear canal opening pointers 83 to the individual's ear canal openings. The
ear position reference arm 84 is preferably removable from the ear area via
a reference arm vertical adjustment knob 85 to minimize acoustic reflections
into the ear area during transfer function measurements.
An infrared tracking method (not shown) may also be used to position and
maintain the head in the proper position with respect to the speaker array 16,
Fig. 5; 89-94, Fig. 7). A light- reflective target object (not shown) placed
just
below the ear lobe of the individual, may be used to reflect the infrared
light
from the incident infrared light emitter. Proper ear placement is indicated by
reflected light which is detected by the positioning sensor interface 69 (Fig.
4).
The virtual acoustic space measurement system generates various sets of
transfer functions that are used during the hearing evaluation process.
Generally, a transfer function of a linear system defines a complex function
H(jw) having magnitude and phase characteristics that are dependent on
frequency (w). Once a transfer function H(jw) is determined, a system's
response to an arbitrary input signal can be predicted or synthesized.
CA 02235067 1998-04-16
WO 97/23117 PCTlUS96/13126
-23-
The transfer function set in the virtual acoustic space measurement system is
obtained from a set of acoustic sources, such as speakers, positioned in a
three-dimensional space. The preferred speaker setup is an array of six
. speakers 89-94 positioned at an equal distance (d) from a patient head
reference point 88, as shown in Figs. 5 and 7. The head reference point 88 is
defined as the point bisecting the line joining the centers of the openings of
the ear canal 21.
Four of the speakers, i.e. #1 (89}, #2 (90), #3 (91), and #4 (92) are located
in
i 0 the transverse plane 95 containing the head reference point 88. Speakers 1
through 4 are positioned at azimuth angles 0°, 45°, 315°,
and 270°,
respectively, as shown in Fig. 7 at A. Three of the speakers, i.e. #1 (89), #5
(93), and #6 {94} are located in the sagittal plane 96 containing the head
reference point 88. Speakers #1, #5, and #6 are positioned at altitude
angles of 0°, 45°, and -45°, respectively, as shown in
Fig. 7 at B.
A set of transfer functions for the six-speaker configuration shown in Fig. 7
allows six pairs, i.e. right and left ear measurements, of frontal
measurements where the head is facing speaker #1. An additional six pairs
of back measurements are preferably taken where the head is facing
opposite (not shown) to speaker #1. Accordingly, a complete transfer
function set consists of 12 pairs of measurements that represent finite points
in a sphere of a radius (d). Of the twelve paired measurements, eight paired
measurements are in the transverse plane and six paired measurements are
in the sagittal plane. Two paired measurements are common to both planes.
Paired measurements contain not only individual transfer functions for each
ear, but also contain the interaural phase relationship with respect to each
speaker.
A transfer function measurement set with a pair of probes placed near the
tympanic membrane in the unoccluded ear canal is referred to herein as the
unaided transfer function H~a (pn. lw), where pn is the location of speaker n
defined by polar coordinates d, 8, and a., where d is the distance between
the speaker and the head reference point as shown in Fig. 7 at A.; B is the
azimuth angle of sound incidence with the respect to transverse plane as
shown in Fig. 7 at A.; and a is the altitude angle with respect to the
sagittal
plane as shown in Fig. 7 at B. Hua (p~, jw ) represents the acoustic transfer
function that results from sound propagation from a speaker #n to the
tympanic membrane when various acoustic factors are considered, including
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-24-
atmospheric propagation losses, effects of head, torso, neck, pinna, concha,
ear canal, tympanic membrane, and middle ear impedance.
Transfer function measurements with a probe tube placed on the face-plate
of the ICP may also be made. These measurements are referred to herein as
Hfp(p~, jw), which represent the transfer function from a speaker #n to a
face-plate (fp) of the ICP (discussed in more detail below), at a location
representative of the microphone position on a face-plate of a simulated
hearing aid.
Generally a transfer function H(p(~,e,a), jw) at an arbitrary point pd,8,a in
space at coordinates d, B , and ,a can be interpolated from the set of
measured transfer functions as shown in Fig. 8. For example, it is known that
the sound pressure from an audio source is inversely proportional to
distance in normal atmospheric conditions. Furthermore, a transfer function
of a point in space can be approximated by the weighted average of the two
nearest measured transfer functions. Fig. 8. shows an example of an
approximate transfer function H(i3 ,jw) interpolated in the transverse plane
at point i3 from transfer functions H (i j , jw) and H (i2 , jw), which are
also
interpolated from transfer functions H(m ~ , jw) and H (m2 , jw) measured
with speakers #1 (89) and #2 (90).
Thus;
H(i3, jw)= jH(m7, jw)fH(m2,lw)I~~2~Lat(.%w)I 11l
where Lat (jw) is the atmospheric loss transfer function due to atmospheric
absorption and spreading roll-off of sound.
Similarly, interpolation can be used to approximate any transfer function at
an arbitrary point in a three-dimensionaE space from the weighted average of
the nearest set of measured transfer functions. The accuracy of interpolated
functions can be improved if additional measurements are made with
additional speakers and/or speaker-head orientations. The preferred
embodiment of the invention employs a practical compromise between the
number of speakers, e.g. six in the embodiment of the invention described
herein, and individual orientations, e.g, two: a front and a back orientation.
Furthermore, non-linear weighting for transfer function interpolation may be
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13I26
-25-
more appropriate if determined from statistical data obtained from transfer
function measurements of large number of individuals.
Other transfer functions measured by the VEA system include:
(1 ) the Hicp-rec (lw) transfer function, which represents the ICP receiver to
in-the-ear-canal electroacoustic transfer function, as measured by a probe
when the 1CP is positioned in the ear canal of the individual;
(2) the Hicp-mic (lw) transfer function, representing the electroacoustic
transfer function from an 1CP speaker to the microphone of the hearing aid
used during the hearing aid evaluation; and
(3) the Hicp-fb(jw) transfer function, representing the acoustic leakage, i.e.
acoustic feedback, from the receiver of the 1CP measured at face-plate of the
ICP.
The transfer functions Hua (Pn~ jw)~ Hfp (Prn lW)~ Hicp-rec (lw)~ Hicp-mic
(lW),
and Hicp-fb(lw) are employed in various combinations to digitally synthesize
acoustic signals, representing unaided, simulated aided, or aided listening
conditions, with realism that is.not possible with conventional evaluation and
fitting methods.
In Fig. 9, for example, a teacher-talker i01 and a child-listener 102 acoustic
environment 100 is created as follows: direct acoustic paths pR i and pL ~ ,
and reflection paths pR2 and p~ , for right and left ears of the child-
listener
102 are represented by transfer functions interpolated from previously
measured transfer functions of the child.
The acoustic realization of the environment of Fig. 9 is shown in Fig. 10, in
which a digital audio file 107 that represents teacher-talker speech is
retrieved from a system memory 106 and digitally processed by digital signal
processor 114. The digital signal processor performs signal processes HUa
(PRy , Ice) 108, HUa (PL7 . l~) 110, HLa (pR2 ~ lw) 109 and Hua (PL2 . lw)
111, which represent the paths pRt, pL1 ~ PR2 ~ and pL2 , respectively.
Right and left ear path processes are summed at summing nodes 112 and
113 and are further processed with inverse transfer functions, 1 /Hicp-rec-Rt
(jw) (116) and 1/HicP-rec-Lt (jw) (104), for right and left ICP receivers
119/120, respectively.
CA 02235067 1998-04-16
WO 97/23117 PCT/LTS96/13126
-26-
The inverse transfier functions are provided to cancel the acoustic transfier
function that occurs between the ICP receiver and the residual volume of the
ear canal as the sound is delivered. The processed right and left digital
signals are then converted to analog signals via a digital-to-analog converter
115 and routed to right and left ICPs via an audiometric interface circuit
117.
The process of projecting a virtual audio image to a listener at a particular
point in a three-dimensional space, such as teacher-talker speech to a child-
listener, is referred to as spatiaiization.
Alternatively, five-voice signals from the operating clinician via the
operating
clinician microphone can be used, instead of digital audio data, for
spatialization and delivery to the listener wearing the ICP pair. The virtual
position and volume of the spatialized audio source are under the control of
the virtual audiometer system ofi the present invention, as is explained in
more detail below.
Transfer function measurements of linear time-invariant systems, such as the
transfer functions Hua (pn~ l~'~, Hfp ~Pn~ lw>> Hicp-rec ~lw~~ Hicp-mic (lw~~
and
Hicp-fb~lw~r typically employs discrete or swept pure tone acoustic stimulus.
Other stimuli include speech-noise, white-noise, and other speech-like noise
signals. Pseudo-random noise sequences and other signals have also been
used to reduce the time required to compute the transfer function.
Computational methods include Fast Fourier Transform (FFT), Maximum-
Length Sequence (MSL), and Time-Delay Spectrometry (TDS) (see Rife. D.,
Vanderkooy, J., Transfer-Function Measurement vzrii'h Maximum-Length
Sequences, J. Audio Engineering Soc., Vol. 37, No. 6, June 1989, pp. 418-
442). The advantages of MSL and TDS measurement include reduction of
room reflection effects on the transfier function. One important component of
measured transfier functions used in the present invention is the direct path
transfer function.
In the preferred embodiment of the invention, the VEA's probe microphones
are calibrated at the head reference point when the VEA is first installed in
its
clinical setup. These calibration data, stored in the system memory, are
subsequently used during transfier function measurements to correct for the
unique frequency response characteristics of each probe microphone used
and the unique characteristics of room acoustics.
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-27-
Fig. 11 is a partially sectioned, perspective view showing an intra-canal
prosthesis (ICP} for an ICP-1TE representing hearing aids for shallow ear
canal placement; Fig. 12 is a partially sectioned, perspective view showing
an ICP for an ICP-iTC representing hearing aids for deep ear canal
placement; Fig. 13 is a perspective view showing an lCP face-plate end,
. inciuding face-plate probe tube holders and probe tube placement; Fig. 14 is
a partially sectioned, side view showing an 1CP core module for a two-part
ICP configuration; Fig. 15 is a partially sectioned, side view showing
adjustable vent inserts for an ICP-ITE; Fig. 16 is a partially sectioned, side
view showing an ICP-ITC sleeve for a two-part !CP configuration; Fig. 17 is a
partially sectioned, side view showing a complete two-part ICP-lTC
assembly; Fig. 18 is a partially sectioned, side view showing an ICP having a
programmable vent; Fig. 19 is a partially sectioned, side view showing a
hearing aid and direct acoustic coupling method to an ICP, including direct
acoustic coupling via a magnetic attraction method; Fig. 20 is a partially
sectioned, side view showing a hearing aid and direct acoustic coupling
method to an 1CP, including direct acoustic coupling via an acoustic coupler
method; Fig. 21 is a partially sectioned, side view showing a hearing aid and
direct acoustic coupling method to an ICP, including a programming and
acoustic coupling interface; and Fig. 22 is a partially sectioned, side view
showing a hearing aid and acoustic coupling to an ICP via an acoustic
coupler tip, all according to the invention.
In the foregoing figures, those elements of the invention that are common to
the various embodiments have a common numeric designator. For example,
the ICP of Figs. 11 and 12 each have a receiver 136, while the housing 129
in the embodiment of Fig. 11 is different from the housing 152 of the
embodiment of Fig. 12.
The infra-canal-prosthesis (ICP), shown in Figs. 11-22, consists mainly of a
receiver 136, a receiver port 199, a probe tube 133 inserted in probe tube
canal 134, vent inserts 128 inserted in vent canal 130, a probe microphone
131, a face plate 122, and a housing made of a flexible material, such as an
acrylic. The ICP is generally designed to represent physical and
electroacoustic characteristics of a desired type of hearing aid with the
exception of the signal processing and generation, which is performed by
the audio synthesizer board of, the computerized virtual electroacoustic
audiometer system. Figs. 11 and 12 show 1TE and ITC ICPs that represent
hearing aids having shallow and deep canal placement, respectively.
CA 02235067 1998-04-16
WO 97!23117 PCTlL1S96/13126
_~8_
The receiver 136 used in the preferred embodiment of the present invention
(manufactured by the Knowles Corp. of Itasca, Illinois) was chosen for its
acoustic characteristics, which are similar to receivers used in commercially
available hearing aids, as well as its very low noise output characteristics.
ICP receiver variations from simulated hearing aid receivers are stored in the
VEA system memory as a correction transfer function used during various
simulation processes. The probe tube 133, preferably made of a silicone
rubber material and having a diameter of approximately 1 mm, is inserted in
the probe tube canal 134 of the ICP as shown in Figs. 11-22.
A vent canal 130 is preferably provided far pressure equalization in the ICP-
ITC versions that have deep canal insertion depths (Figs. 12 and 17), and to
accommodate vent inserts for the ICP-ITE version having shallow canal
insertion depths (Figs. 11 and 15). In the ICP-ITE versions, a vent canal
allows the insertion of various vent inserts into the vent canal to achieve
desired in situ acoustic characteristics. For example, a vent insert of
relatively large diameter may be used to reduce the occlusion effect that
results from increased perceived volume of the individual's own voice. On
the other hand, a smaller vent insert may be used to eliminate acoustic
leakage from the receiver via the vent insert. A miniature connector socket
138 and connector plug 123 electrically connects the ICP to the VEA system
via attached connector cable 125.
The VEA system, in conjunction with the probe microphone system, permits
measurements of the occlusion effects versus ICPs and vent types, as is
explained later. The ICP also contains two probe tube holders 124 and a
placement handle 126 for placement of the probe tube, as shown in Figs. 11,
12, and 17. Fig. 13 shows a more detailed illustration of a face plate 122,
including the face plate tube holders 124. In the figure, a ICP/ITC sleeve
156, and a hearing aid microphone position 132 are also shown. This
configuration is used when measuring acoustic leakage feedback and face-
plate transfer functions.
The ICP housing (129, Fig. 11; 152, Fig. 12) is preferably made of a soft
flexible material with acoustic baffling effects to provide comfort and
acoustic
sealing. Several versions of the 1CP can accommodate a variety of ear
canal sizes. For example, a small housing version is more suitable for
pediatric populations, while a larger version is suitable for adults who have
CA 02235067 1998-04-16
WO 97!23117 PCT/US96/13126
-29-
large ear canals. The ICP, shown in Figs. 11 and 12 is preferably disposable
to avoid contamination from individuals who have infected ear canals.
An alternate embodiment of the invention provides a two-part ICP
configuration, as shown in Figs. 14-17. A core part 169 (Fig. 14) is inserted
in a variety of disposable sleeves 177, as shown in Figs. 15 and 16. This
option provides an economical alternative to the configuration shown in Figs.
11-13 because only the sleeve component is disposable. The core part 169
is encapsulated in a protective material, preferably having semi-flexible
properties. A decvupling capacitor 167 may be used to filter extraneous
electromagnetic signals that cause audible noise.
The sleeve part shown in Figs. 15 and 16 is typically made of flexible
material, such as a soft acrylic, such that the ICP fits comfortably into a
variety of ear shapes and sizes. Fig. 16 shows a sleeve suitable for deep
canal insertions, representing lTC and C1C hearing aid types. Also shown
in Fig. 16 is an acoustic baffle system 186 that provides an acoustic sea!
while the 1CP is inserted in the ear canal.
Fig. 15 shows an ICP sleeve for shallow canal insertions representing ITE
hearing. aid types. The ICP core is inserted in the sleeve cavity 179 of any
ICP, including those shown in Figs. 15 and 16. The specific size of the ICP
sleeve selected by the operating clinician depends upon the test performed,
individual canal size, and hearing aid simulation requirements. An example
of the combined parts of a core ICP and an ICP sleeve are shown in Fig. 17,
which represents an ICP-ITC assembly.
Fig. 18 shows a variation of the vent mechanism where the size of the vent is
electronically controlled and adjusted (see Zdeblick, K., A Revolutionary
Actuator For Microstructures, Sensors Magazine, eb. i 993). This is
accomplished by employing programmable micro-valve 193 (such as the
NO-300 manufactured by Redwood Microsystems of Redwood City,
California) which contains a silicon diaphragm i94 which is to regulate the
size of the vent attached to the vent canal 197 via the micro-valve port 195.
Typical vent size range is between .032 and 1.5 mm, according to the
voltage level supplied from the virtual electroacoustic audiometer module in
response to operating clinician test selections.
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-30-
The ICP is also used in a novel way to test a new type of hearing aids
adapted to interface to the iCP, as shown in Figs. 19-22. Unlike
conventional hearing aid and aided hearing evaluation methods that
typically employ remotely positioned speakers to deliver acoustic signals
into the hearing aid microphone, the ICP of the present invention presents
acoustic signals directly to the microphone 211 of the hearing aid 214. The
acoustic coupling of the present invention spans a minimal distance typically
less than 15 mm.
Figs. 19 and 21 show an embodiment of the invention in which acoustic
coupling is accomplished via a magnetic attraction method. In such method,
the iCP receiver 136 is coupled to the hearing aid microphone 211 via
magnetic attraction between a magnet disk 206 on the receiver end of the
ICP and another magnet disk 209 near the hearing aid microphone port 210,
and which is part of the face-plate 218 of a hearing aid 214, as shown in Fig.
19. A sealing ring 205 provides acoustic sealing to minimize leakage in the
coupling. Also provided are a hearing aid battery holder 227 , a hearing aid
volume control 219, a hearing aid circuit 212, and a hearing aid vent canal
217, all representing conventional components of a hearing aid device.
Additionally, the embodiment of the invention shown in Fig. 21 provides a
programmable hearing aid circuit 253 that allows dynamic ITE testing via
control signals routed from the VEA over a programming cable 257. Fig. 21
shows an electrically programmable hearing aid with a programming cable
257 connecting the hearing aid circuit to the VEA of the present invention.
These hearing aids contain circuits that are programmable or ad~ustabie,
typically via electrical signals. The shown programming interface at the face-
plate is via the battery holder which is adapted to route programming
electrical signals to the hearing aid circuit. The programming signals and
interface methads are typically unique to the hearing aid model as provided
by the specification of the hearing aid circuit used. These programming
signals and interface methods are known to persons skilled in the art of
hearing aid design. Other programmable hearing aids currently
commercially available employ ultrasonic or infra-red signals with the
appropriate signal interface circuits within the hearing aid.
An 'alternative acoustic coupling method couples the ICP receiver 136 to the
hearing aid microphone 211 via a acoustic coupler 243, as shown in Fig. 20.
The extended microphone port 242, unique to the present invention, also
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-31-
acts as a handle to facilitate insertion and removal of hearing aid 214 during
its normal use.
. Another embodiment of the invention, shown in Fig. 22, employs an
acoustic coupler 290 adapted for insertion into a microphone port 299 of the
hearing aid 214. The microphone port 299 is recessed to accommodate an
acoustic coupler tip 291.
Another acoustic coupling method (not shown) employees a suction-cup ring
to couple the ICP receiver to existing conventional hearing aids that are not
equipped with special interface parts.
One major advantage of the direct acoustic coupling of the present invention
is to improve the signal-to-noise ratio at the microphone of the hearing aid
i 5 white the aid is being adjusted or evaluated. This is primarily
accomplished
by acoustically isolating the microphone of the hearing aid from ambient
room noise via its coupling to the ICP.
Hearing aids of the present invention also employ a probe tube canal to
allow for probe tube insertion and subsequent in-the-ear-canal acoustic
measurements via the probe measurement system as shown in Figs. 19-22.
The conventional method of in-the-ear-canal measurements with hearing
aids involve probe placements beneath the hearing aid which subjects the
probe to pinching effects, thus affecting the accuracy of the measurement.
Furthermore, placing the probe tube beneath the hearing aid creates an
acoustic leakage path which causes oscillatory feedback. The probe tube
canal of the present invention also provides an improved method of
advancing the probe while the hearing aid is placed in the ear canal.
The sequence of these phases as outlined in Fig. 23 represents a typical
fitting process unique to the system of the present invention. The fitting
process offered by the virtual electroacoustic audiometer system in the
preferred embodiment of the present invention is implemented in five
phases: (1) reference measurements 284, (2) unaided hearing evaluation
265, (3) predicted aided evaluation 266, (4) simulated aided evaluation 267,
and (5) aided evaluation 268. However, individual phases or a components
of each phase can be administered individually, or in other sequence as
suitable for the individual under hearing evaluation. Each process phase is
implemented in a graphical module, as shown in Figs. 24-28.
CA 02235067 1998-04-16
WO 97/23117 PCT/~1596/13126
-32-
The first phase, f. e. reference measurements, is implemented by a reference
measurements module (Fig. 24) that contains a reference measurement
window {shown open in Fig. 24) and a signal model window (shown
iconized in Fig. 24). The reference measurement window allows for
measurements of various transfer functions that are used later throughout
the fitting process.
The unaided transfer function HUa (pn ~ %~~ described above, is measured
when the 3D-REUR (3 Dimensional Real-Ear Unaided Response) option is
selected. Measurements are obtained from the frontal {facing speaker #1) or
back (facing opposite speaker #1) orientations, depending on the Front/Back
option selected. Plots of right and left ear transfer functions can be
displayed
in either transverse or sagittal plane depending on the Transverse/Sagittal
option selection. Fig. 24 shows a set of 8-paired H~a (pn ~ Ice) transfer
functions in the transverse plane. The measurement is performed by
positioning the individual centrally to the speaker array (discussed above)
and placing right and left probe tubes in their respective unoccluded ear
canal.
Another novel feature of the invention is the ability to measure and quantify
the occlusion effect of the simulated hearing aid, as well as the fitted
hearing
aid. However, before the occluded measurement is taken, a reference
measurement with the ear canal unoccluded must be taken. The procedure,
briefly described here, is to request the individual to utter a vowel,
preferably
a vowel with high energy contents in its low frequency spectrum, such as
"ee." A measurement is taken with the probe positioned near the tympanic
membrane. The occlusion effect reference measurement, i.e. unoccluded, is
saved for occlusion effect measurement with the ear canal occluded using
either the ICP or the hearing aid, as is explained below. The occlusion effect
reference measurement is performed when the occlusion reference option is
selected.
The face-plate transfer function Hfp (pn, jw) (plots not shown) is measured
when the Face-Plate Response option is selected. The ICP is placed in the
ear and the probe tube tip is placed in the microphone position 132 of the
face-plate as shown in Fig. 13.
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-33-
The ICP-receiver to real ear transfer fiunction, Hicp-rec (JW~ is measured
when ICP Calibrate option is selected. This requires the probe tube to be
inserted in the probe tube canal of the ICP, and the tip of the tube near the
tympanic membrane.
To facilitate the proper placement of the probe in the ear canal during
various response and calibration measurements, a novel method is
employed to optimize such probe placement within the ear canal, and
specifically to minimize the efifects of standing waves present in the ear
canal
due to wave reflections from the tympanic membrane. The frequency
dependent standing wave patterns are well characterized and known to
persons skilled in the art of acoustics and particularly real ear acoustic
measurements. The new method of the invention involves acoustic
presentation of a dual tone, one at a low frequency in the range of 1 kHz to 5
kHz, and a second at a range ofi 15 kHz to 20 kHz. The acoustic response to
tone signals delivered either via a speaker or the ICP receiver, depending
on measurement, is continuously measured by microphone probe system
and displayed on the monitor, as shown in Figs. 30-32.
A plat of the acoustic response in an ear of an individual for each tone,
shown in Fig. 29, indicates a characteristic rise in the low frequency
response and a notch in the high firequency response as the probe is
advanced closer to the tympanic membrane. This notch occurs at
approximately 5 mm from the tympanic membrane for the 15 kHz tone.
Monitoring of the relative response characteristics during probe insertion
provides a visual and computer-assisted method to indicate proper probe
positioning as shown in the spectrum plots of Figs. 30-32. The end of this
procedure is generally indicated when a significant notch, typically
exceeding 15 dB as shown in Fig. 31, followed by a significant rise in the
high frequency, i.e. second tone, response.
The low frequency, i.e. second tone, response shows only a small increase,
within 3 dB, as the probe is inserted closer to the tympanic membrane.
Although probe tip to tympanic membrane distance approximation is
possible with this procedure, the object of this procedure is to position the
probe such that minimal standing waves are present at frequencies of
interest during transfer function measurements. For example, if unaided
response measurements up to 6 kHz are desired, advancing the probe until
detecting a notch in 15 kHz response ensures measurement errors not to
CA 02235067 1998-06-19
-34-
exceed 2.5 dB at 6 kHz. Improved accuracy can be achieved by selecting a
higher
frequency for the second tone, although this increases the chance of advancing
the probe
too far, resulting in touching the surface of the tympanic membrane, an
occurrence that is
generally safe but that may cause discomfort.
Other combinations of tones, including a single, triple, composite, and other
signals can
also be used to implement the above procedure of continuously measuring the
response to
various acoustic stimuli and detecting an appropriate stopping point during
probe
advancement, with little regard to probe distance to the tympanic membrane.
The
appropriate probe position is referred to hereafter as the probe reference
point.
The second phase, unaided evaluation, is implemented by an unaided evaluation
module,
shown in Fig. 25, which consists of an unaided analysis window, shown open in
the figure;
a spatialization window, also shown open; a signal model window, shown
iconized; and an
audiometric evaluation window, also shown iconized.
The unaided analysis window allows for various in-the-ear-canal measurements
and
displays for hearing evaluation in the unaided condition while the ICP is
inserted in the ear
canal. Measurements and plots include Audiogram spectrum, Distortion, Time
Analysis,
Spectrogram, and 2-CC curves. Acoustic stimuli, measurement methods, and
associated
plots for these tests are known to persons skilled in the arts of audiology
and signal
analysis. However, the Audibility Spectrogram is a new feature that is unique
to the
present invention as described below.
The Audibility Spectrogram 405 is a spectral plot showing the audibility of a
signal with
respect to the hearing profile of the individual and the critical audibility
features of an
acoustic signal. The audibility spectrogram is essentially a three-dimensional
matrix
represented in a two-dimensional plot that indicates signal dynamics (time)
and Critical
Audibility Regions (CAR) versus frequency, as shown in Figs. 25 and 26. CARS
400-402,
410-412, shown as the outer contours, are specific to each signal segment that
is selected
from the signal model window. CARS of a speech segment are defined by the
critical
sound features, such as the energy of significant formats in vowels, the
energy of
fundamental frequency of voicing, the energy of aperiodic frequency sounds,
CA 02235067 1998-06-19
-35-
The Audibility Spectrogram plots are derived by combining spectrograms of
analyzed signals and defined CARS, and probe measured spectrograms computed
and compared with the measured hearing profile of the individual at the CARS.
Probe measured spectrogram values that fall below the threshold of hearing for
the
individual are assigned to Below Threshold (B-Thresh) values which define the
outer contour region 403, 413-415, within the CAR; while measured spectrogram
values that exceed the threshold of hearing within CAR are assigned Above
Threshold (A-Thresh) values which define a region 404, 416, 418, 420 within
the
Below Threshold region; and measured spectrograms values that exceed the
uncomfortable loudness level (UCL) of the individual are assigned Above-
UnComfortable Loudness level (A-UCL) values which define the inner-most
contour regions 417, 419, 421.
The resulting plot, which may be color-coded, is typically contour shaped for
speech signals. However, any type of acoustic signal can be assigned CARS and
a
corresponding audibility spectrogram based on the individual's measured
hearing
profile. The objective of the Audibility Spectrogram plot is to provide a
quick
graphical means of indicating the audibility of dynamically received acoustic
signals
by taking in consideration the individual's hearing profile and the critical
audibility
features of a signal model. This plot is particularly important in hearing aid
fitting
optimization processes during predicted aided, simulated aided , and aided
evaluation.
The spatialization window permits selection of signal presentation mode,
either in
Spatialized or Intracranial modes. Spatialized mode presents selected sources
and
background signals to be delivered to both ears via inserted ICPs according to
the
selected spatial relationship of head, sources, background, and boundaries, as
shown in Fig. 25. Spatial relationships include the distance between the audio
source and the head reference point (c~, azimuth angle (q), and altitude angle
(a).
Various individual and calibration transfer functions are employed to
synthesize
audio signals with realistic listening effects. Signal sources and
corresponding
levels are selected from the Signal Model window (not shown). Intracranial
mode, on the other hand, offers the conventional sound
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-36-
presentation method where selected signals and corresponding levels are
delivered without spatialization to one or both ears.
The Signal Model window permits the selection of source and background
signals and corresponding level. Source selection may be of pure tone type,
speech, music, or any signal of audiological significance. Background
signals are typically competing speech, environmental noise, and other
signals of audiological significance. The level of signals selected in the
spatialized mode is preferably in dB SPL calibrated to 1 meter from the
source in free field. The measured in-the-ear-canal acoustic response is
preferably displayed in dB SPL as measured by the probe microphone
system.
In the intracranial mode, source and background signals are routed to right,
1 ~ left, or booth ears as in conventional audiometry. The level of signals
selected in the intracranial mode is preferably in dB SPL. The H~cp-rec ~lw~
transfer function measurement via the ICP calibration procedure described
above permits level selection in dB SPL. Furthermore, measurements via
the probe microphone system can be made as needed to ensure that the
probe and the ICP remain properly positioned in the ear canal.
A specific selection of source and background signal type, levels, and
spatialization mode is defined as a signal model. One or more signal models
can be selected, saved, and retrieved by the system for presentation and
analysis purposes. A signal model can represent any individual or a
combination of acoustic signals/scenarios, including speech, background
noise, music, pure tone, masking noise, composite signals, and other
audiologicafly significant signals.
30. The audiometric evaluation window, shown iconized, allows for various
conventional audiometric measurements to be taken. This includes
threshold audiogram, most comfortable level (MCL}, uncomfortable loudness
level (UCL), speech reception threshold (SRT), and various other
audiometric measures known to persons skilled in the art of audiology.
However, unlike conventional audiornetry where transducers are calibrated
in various acoustic couplers and measurements are measured in relative
hearing level (HL) terms , the preferred method measures the in-the-ear-
canal response in absolute sound pressure level (SPL) terms.
CA 02235067 1998-04-16
WO 97/23117 PCTNS96/13126
-37-
Another feature of the invention relates to the modes of audiometric signal
presentation. As described above, spatiaiized or intracraniai listening modes
selected from the Spatialization window, not only affect the presentation
selected from the Signal Model window, but also the Audiometric Evaluation
window as well. For example, a standard audiological word list such as NU-
6 or W-22, commonly used in conventional speech audiometry, can be
presented in the conventional intracranial mode, or alternatively, in the
spatialized mode unique to the invention.
~ 0 The signal process of a spatialized unaided evaluation involves the
unaided
transfer function Hua (Pn ~ lW>> interpolated based on selections of the
spatialization window, and the H icp-rec ~JW~ transfer function. A signal
process implementation of a particular spacialized unaided evaluation is
shown in Fig. 10.
The third phase, the predicated aided evaluation, is implemented by the
predicated aided evaluation module. This module, shown in Fig. 26, allows
the operating clinician to select a hearing aid and predict its performance
without the involvement of the hearing-impaired individual. The module
consists of a Hearing Aid Select/Adjust window, shown open; a Predicated
Analysis window, shown open; a Signal Mode( window, shown iconized; a
Spatialization window, shown iconized;and the Audiometric Evaluation
module. The Signal Model, Spatialization, and Audiometric Evaluation
windows are essentially identical to those described in the Unaided
Evaluation phase.
The Hearing Aid Select/Adjust window permits hearing aid selection and
subsequent adjustment. The predicated results of the seiectionladjustment
are shown on the selected plots of the adjacent Predicted Analysis window.
Hearing aid selection can be automatic or manual, depending on the
hearing aid selection Automatic/Manual option selected. Automatic selection
involves selecting one or more hearing aids based on the fitting algorithm
selected, and various other criteria selected by the hearing-impaired and the
operating clinician. Conventional fitting formulae and methods, such as
POGO, Berger, and NAL-ll, are provided.
The preferred fitting method is the dynamic audibility method which employs
a rational such that Audibility Spectrogram is optimized. This corresponds to
plots that maximize the Above-Threshold (A-Thresh) contour areas while
CA 02235067 1998-04-16
WO 97/23117 PCT/US961I3126
-38-
minimizing Bellow-Threshold (B-Thresh) and Above-Uncomfortable
loudness Level (A-UCL) contour areas. Hearing aid models that best match
the selected criteria are automatically retrieved from the system memory.
Alternatively, manual selection can be made by choosing one or more
hearing aid models from the available list of models. A hearing aid model
contains al! of the necessary electroacoustic parameters that are used for
signal processing of a signal model. The results of the signal process are
used in the Predicted Analysis window for analysis and plotting purposes.
Hearing aid parameters of a selected hearing aid model are adjusted
automatically or manually depending on the hearing aid adjustment
Automatic/Manuai option and the fitting method selected.
A hearing aid control parameter set is typically unique to the hearing aid
model selected. In the example window shown in Fig. 26 with hearing aid
model DigiLink 100 selected, the control parameters are: volume control
{VC), Low Frequency Cut (LFL), compression Threshold Knee (TK),
Microphone type (M(C), Receiver type (REC), and Vent Size selection which
reflects vent size of the ICP inserted. If a different vent size is selected,
either
manually via the vent insert selection, or electronically via the programmable
micro-valve vent selection, a new H;cp-spkr (j~') transfer function is
preferably measured to improve the accuracy of the analysis.
The predicted analysis window is essentially identical to the unaided
analysis window, described above, with the exception of the signal
processing model that includes the measured face-plate transfer function
Hfp (pn, jw ) (292, 293; Fig. 33), hearing aid transfer function Hha (jw)
{294;
fig. 33), and the measured ICP receiver to real-ear H~cp-rec (jw) transfer
function for the aided ear (295; Fig. 33). The hearing aid Hha (jvv) transfer
function is typically non-linear and varies depending on the hearing aid
selected. The total hearing aid transfer function Hha-t (lw) typically
includes
transfer functions of the microphone Hm~c(jw), hearing aid circuit Hha-rec
(jw), and the receiver Hha-rec (j~)~ The transfer function Hha (jw) differs
from Hha_~ (jw) by excluding the hearing aid receiver and, instead, including
a receiver correction transfer function HRec-corr(lw), that defines the
difference between the predicted hearing aid receiver and the ICP receiver
employed. This correction transfer function HRec-corr (!w) is typically a
linear
transfer function and is supplied by the VEA system.
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-39-
The predicted aided analysis process for an aided right ear and unaided left
ear for a child-listener/teacher-talker scenario is shown in Fig. 33. The
results of the digital signal process are stored in the system memory 106 for
analysis and display.
. The analysis of the predicted data in the system memory includes audibility
analysis as described above. The plotting includes an Audibility
Spectrogram that indicates audibility contours of Below-Threshold, Above-
Threshold and Above-UCL with respect to critical audibility regions (CRAB).
Fig. 26 shows improved audibility in the predicted aided condition versus
unaided condition shown in Fig. 25, i.e. increased Above-threshold contour
areas.
Another prediction measurement unique to the present invention, is the
measurement of occlusion effect caused by the insertion of the ICP into the
ear canal that is characterized by the perceived amplification of the person's
own voice. The present invention provides a method of measuring,
subjectively and objectively, the magnitude of the occlusion effect. The
subjective method is performed by asking the individual wearing the ICP to
evaluate his own voice when speaking. If the response is objectionable to
the hearing-impaired candidate then an alternative (CP, representing a
different hearing aid, may be considered.
The objective method involves the measured response via the probe system
in the occluded ear canal and subtracting the occlusion effect reference
measurement, i.e. unoccluded ear-canal measurement, as described above.
The patient microphone 57, external to the ear canal, is typically employed to
record the individual's own voice during occlusion effect measurements to
ensure constant intensity level during both unoccluded and occluded ear
canal measurements (see Mueller, H.G., Hawkins, D.B., Northern, J. L.,
Probe Microphone Measurements- Hearing Aid Selection and Assessment,
1992, pp. 221-224). A unique feature of the present invention is to eliminate
not the only requirement of constant voice intensity, but also constant voice
spectral characteristics. This is accomplished by adjusting the calculated
occlusion effect measurement by the difference in the spectral characteristics
of the individual's own voice.
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13I26
-40-
It is known in the fiield of audiology that deep hearing aid insertion
substantially reduces the occlusion effect, particularly at low frequencies in
the range of 125 to 1000 Hz. Therefore, a smaller ICP, representing a
smaller simulated hearing aid, may be used for subsequent evaluation
phases.
The occlusion effect created by two types of ICP, i.e. ICP-ITC and ICP-ITE, is
shown in the plot of Fig. 27. This plot indicates a significant occlusion
effect
due to the ICP-ITE versus the lCP-ICP for an individual. This is expected
since the ICP-ITE creates a greater residual volume, to which the occlusion
effect is known to be directly proportional.
The advantage of ICP measurement at the probe reference point is that all
measurements taken are independent of the ICP selected or its placement in
the ear canal. However, to present accurate spatialized sounds to the
individual, the Hicp-rec (Jw~ transfer measurement is required whenever a
new ICP is selected and inserted into the ear canal of the individual.
Another measurement unique to the invention is that of acoustic feedback
caused by acoustic leakage from the ICP receiver, when simulating a
hearing aid receiver, to the fiace-plate ofi the ICP, which simulates the
fiace-
plate of the hearing aid. The transfer function Hicp_ fb(jw) (338; Fig. 37),
e.g.
amplitude and phase response, is measured at the face-plate as described
above. The opening created by the removal of the probe tube from the ICP
probe tube canal is preferably plugged during the feedback measurement to
exclude acoustic leakage due to the probe canal.
A significant application of the feedback transfer function is in the
simulation,
and thus prediction, of oscillatory feedback of the simulated hearing aid.
This undesirable oscillatory feedback manifests itself in the form of
whistling,
which interferes with the normal operation of the hearing aid. The prediction
and simulation of the oscillatory feedback ofi a simulated hearing aid having
a selected setting is accomplished by incorporating the ICP feedback
transfer function Hicp-fb(jw) 337, as shown in Fig. 37.
Oscillatory feedback can be audible to the individual wearing the ICP via the
ICP receiver. The oscillatory feedback can also be measured via the ICP
microphone system in conjunction with the VEA system. This fieature allows
the operating clinician to adjust the settings ofi the simulated hearing aid,
CA 02235067 1998-04-16
WO 97/23117 PCTJUS96/13126
-41-
particularly the gain, frequency response, and vent size, such that
oscillatory
feedback is minimized or eliminated. Similarly, the VEA system can be
employed to select automatically an alternate hearing aid or alternate
hearing aid parameter set, such that oscillatory feedback is minimized or
eliminated.
The predicted aided analysis window also includes other analysis and
corresponding plots of Audiogram, Distortion, Time Analysis, Spectrogram,
2-cc Curve. These are standardized measurements and plots that are
known to persons skilled in the art of hearing sciences and technology. The
2-cc coupler curves involve conversion of measured in-the-ear-canal
response to standard 2-cc coupler curves using real-ear-to-2-cc coupler
conversion formulas. Standard signal models, such as pure tones, ace
typically involved in the 2-cc coupler measurements (see Specification of
Hearing Aid Characteristics, ANSI-S3.22-1987, American Standards
National Institute). Other evaluation methods conceived and well within the
means of the invention include the Articulation Index (AI) measures for
unaided, predicted aided, simulated aided, and aided conditions.
An objective of the predicted aided module is to predict objectively the
performance of a selected hearing aid according to the selected signal
model, selected hearing aid parameter set, and the individual's hearing
profile, without the involvement of the hearing-impaired individual.
The fourth phase, simulated aided evaluation, is implemented by the
simulated aided evaluation module, as shown in Fig. 27. This module allows
the operator to select and optimize one or more hearing aids and simulate
their audible characteristics. The module consists of a Hearing Aid
Simulation window, shown open; a Simulated Aided Analysis window,
shown open; a Signal Model window, shown iconized; a Spatialization
window, shown iconized; and the Audiometric Evaluation module, shown
iconized. The Signal Model, Spatialization, and Audiometric Evaluation
_ windows are essentially identical to those described above. The Simulation
Aided window is essentially identical to the Hearing Aid Seiect/Adjust
window of the Predicted Aided Evaluation module. Similarly, the Simulated
Aided analysis window is essentially identical to the Predicted Analysis
window.
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-42-
A major difference in the simulated aided evaluation module is the module's
ability to synthesize simulated aided conditions and to present the audible
results to the hearing-impaired individual. Another significant difference is
that analysis is performed by the module based on measured, rather than
predicted, data. The measured response is obtained via the microphone
probe measurement system with the probe tip placed at the probe reference ,
point, as discussed above.
An example of a simulated aided signal process, shown in Fig. 34, involves
the transfer function of the hearing Hha (jnr) that includes the HRec-corn
(lw)~
and the face-plate transfer function Hfp (p~, jw) for simulation of the aided
ear. The results of the process are converted to analog signals via the
digital-to-analog-converter 115 and routed to the right and left ICPs, 119
and 120 respectively, inserted in the ear canals of the individual.
If the microphone of the predicted hearing aid is of directional type, then
separate microphone transfer functions, representing its directional
properties are employed, as shown in Fig. 35. A digits! audio file 107 is
retrieved from the system memory 106 and processed with face-plate
transfer functions Hfp(p~,jw) (310; Fig. 35) and Hfp (p2, jw) (312; Fig. 35),
where p ~ and p2 represent two points in a three-dimensional space. Signs!
paths from p~ and p2 may represent direct and primary reflective paths,
respectively. Secondary reflective paths p,3 , p4 ..., pn (not shown) can be
similarly represented in the digital signal process.
The results of each face-plate transfer function step are further processed
with the corresponding microphone transfer function 318, 320 for each
signal path from points p~ , p2, ...pn. The results are summed 326 and are
processed by the hearing aid circuit transfer function Hha-cir ( lw ) 322,
HRec-corn (lw ) 324, as shown in Fig. 35. The resulting digitally processed
signal is then converted to analog signal via the digital-to-analog converter
1 y 5 and routed to the appropriate iCP within the ear canal via the
audiometric transducer interface 117.
The simulated aided analysis window includes measurements and
corresponding plots of Audiogram, Distortion, Time Analysis, Spectrogram,
Audibility Spectrogram, 2-cc Curve, Occlusion Effects, and Feedback
Analysis. These measurements are essentially identical to those described
above for the predicted analysis window. This process is based on the
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/I3126
-43-
system's ability to compute a hearing aid prescription based on a selected
fitting prescription formula/rational. The selected hearing aid can be
adjusted
and results analyzed and plotted with or without the involvement of the
- hearing-impaired individual.
An objective of the simulated aided module is to optimize, objectively and
subjectively, the performance of a selected hearing aid according to
measured in-the-ear-canal probe response as a function of the selected
signal model, hearing aid parameter set, the individual's measured hearing
profile, and subjective responses to the presented audible signal.
One feature unique to the invention is the ability to compute the
characteristics of a simulated monaural or binaural hearing aid system that
produces natural sound perception and improved sound localization ability
to the hearing impaired individual. This is accomplished by selecting a
simulated hearing aid transfer function that produces, in conjunction with the
face-plate transfer function, a combined transfer function that matches that
of
the unaided transfer function for each ear. The matching requirement
typically involves frequency and phase responses. However, the magnitude
response is expected to vary because most hearing impaired individuals
require amplification to compensate for their hearing losses.
Once the hearing aid selection and optimization processes are completed
via VEA system simulation, the characteristics of the simulated hearing aid
2~ are translated to hearing aid specifications for manufacture/assembly.
Manufacturing specifications include: hearing aid components simulated by
the VEA system, including the microphone and receiver; shape and size of
hearing aid according to the ICP selected; hearing aid circuit blocks and
circuit components; hearing aid parameter setting; and vent typelsize. An
objective of the VEA system is to provide a detailed specification to the
manufacturerlassembler to manufacture/assemble a monaural or binauraily
matched hearing aid system that closely matches the preferred simulated
hearing aid. Ordering of the actual hearing aid is performed from the Order
menu shown in Fig. 27 which provides a printout of detailed hearing aid
specification.
The final step in the process, aided evaluation, is represented by the aided
evaluation module as shown in Fig. 28. This module consists of an Aided
Evaluation window, shown open, an Aided Analysis window, shown open;
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13126
-4.4-
an Audiometric Evaluation window, shown iconized; a Signal Model
window, shown iconized; and a Spatialization window, shown iconized. The
tatter three windows are essentially identical to those in the predicted aided
evaluation and simulated aided evaluation windows. The aided evaluation
window permits electronic adjustment of manufactured hearing aid
parameters as in the case of a programmable hearing aid, shown in Fig. 21,
or displaying the suggested parameter setting in the case of a manually
adjusted hearing aids, shown in Fig. 20.
The aided analysis window is similar to the analysis window for unaided,
predicted aided, and simulated aided evaluation process steps, except that
the measurements and corresponding plots reflect the response from the
actual hearing aid inserted in the ear canal of the individual rather than
predicted or synthesized signal, i.e. simulated aided, response analysis.
Synthesized realistic acoustic signals are presented to the hearing aid by
coupling spatialized sounds directly to the microphone of the hearing aid, as
shown in Figs. 19-21. The face-plate transfer function, Hfp (pn, jw), and the
supplied lCP receiver-to-microphone transfer function Hicp-mic (lw) are
employed in the digital synthesis process, as shown in Fig. 36. A digital
audio file 107 representing an audio source at location pn in space is
retrieved from the system memory 106 for processing with the free-field to
face-plate transfer function H fp (pn, jw ) 340, 342 for right and left ears,
individually. Other parallel processes reflecting filtering of additional
audio
sources or filtering of reflective paths, shown collectively in the dashed
rectangles 341, 343, are summed with the right 112 and left 113 summing
nodes. The outcome of summing nodes is further processed to equalize the
ICP receiver to hearing aid microphone coupling effects by applying the
inverse transfer function 1 / H;cp-mic (lw) 344, 345. The acoustic signals
supplied to the microphones 350 of the hearing aids 351 represent
spatiaiized signals with characteristics selected and controlled by the VEA
system operator via the Spatialization, Signal Module, and Audiometric
Evaluation windows.
Electroacoustic testing of the hearing aid, coupled with the 1CP as described
above, may also be performed external to the ear canal, for example 2-cc
coupler measurements can be performed by connecting the receiver output
of the hearing aid to the 2-cc coupler input. The ICP, in conjunction with the
signal generation capability of the VEA, can produce various acoustic stimuli
CA 02235067 1998-04-16
WO 97/23117 PCT/US96/13I26
-45-
as input to the hearing aid during its 2-cc coupler-based hearing aid
evaluation. Similarly, 2-cc coupler measurements can be performed on the
ICP, i.e. a simulated hearing aid, by connecting the receiver output of the
iCP
to the 2-cc coupler input.
The invention not only deals effectively with today's diagnostic and fitting
problems but also provides a basis for new tools that are audiologically
signifiicant. For example, the system's ability to synthesize realistic
acoustic
conditions, both simulated aided and aided, can be used as an auditory
rehabilitative tool where a hearing impaired listening ability is improved by
interactive training. In such application, the hearing impaired person is
presented with spatialized signals that represent spoken words in noisy
background. Even though the words might be audible as determined from
the audibility measurements and methods described above, these words
might not be intelligible for the untrained hearing-impaired individual.
Depending on the verbal response, or registered response via a response
keypad, the VEA system can provide audible or visual feedback to the
hearing impaired individual that indicates the appropriateness of the
response. The object of this new test is to teach the hearing-impaired how to
improve speech perception and intelligibility beyond mere audibility.
Another test made possible by the invention determines the individual's
ability to localize a sound in a plane or in three-dimensional space. An
example is the detection ofi minimal audible angle (MAA) test whereby the
individual's ability to detect, in degrees, the minimal angular separation of
pure tones versus frequency (see Mills, A. W., On the Minimum Audible
Angle, Journal of Acous. Soc. of Am. 30:237-246, 1 S56). Furthermore, a
comparison of the individual's localization ability can be compared across
unaided, simulated aided, and aided conditions.
The invention also makes it possible to determine the individual's ability to
detect sound movements in a plane or in a three-dimensional space. For
example, a sound object can be synthesized to represent movement in a
particular geometrical and frequency pattern. The individual's impaired
ability to detect the movement can be assessed. Furthermore, a comparison
of the individual's ability to detect sound movements can be compared
across various listening conditions in the unaided, simulated aided, and
aided conditions.
CA 02235067 1998-04-16
WO 97/23117 PCTNS96/13126
-46-
Although the invention is described herein with reference to the preferred
embodiment, one skilled in the art will readily appreciate that other
applications may be substituted for those set forth herein without departing
from the spirit and scope of the present invention. Accordingly, the invention
.
should only be limited by the Claims included below.
