[0001] The present invention relates to a method for individually fitting a hearing instrument.
[0002] A hearing instrument usually comprises a microphone for generating an input audio
signal from ambient sound, an audio signal processing unit (which nowadays often is
digital) for processing the input audio signal into a processed output audio signal
and an output transducer for stimulation of the user's hearing according to the processed
output audio signals. Audio signal processing in the audio signal processing unit
involves applying a gain function to the input audio signal, which depends on level
and frequency of the input audio signal. Hearing instruments usually are used by persons
suffering from a hearing loss compared to normal-hearing persons, which depends on
level and frequency of the ambient sound. Usually the hearing instrument undergoes
a fitting procedure in order to individually set the gain provided by the hearing
instrument such that the hearing loss of the user is compensated as far as possible.
[0003] In the prior art various attempts have been made for the fitting of hearing aids
to the needs of an individual patient.
[0004] US 4,577,641 relates to a fitting process for a cochlear implant wherein equal loudness contour
(ELC) measurements are conducted after the device has been implanted in order to determine
the individual optimized gain function of the hearing instrument. The ELC measurements
are carried out at the most comfortable loudness level.
[0005] DE 32 05 685 A1 relates to a hearing instrument with an electroacoustic output transducer, wherein
a sound generator is integrated within the hearing instrument for performing hearing
threshold measurements as a function of frequency.
[0006] DE 199 14 992 A1 relates to the integration of a sound generator for audiometric measurement within
a partially or fully implantable hearing instrument for e.g. direct mechanical stimulation
of the inner ear.
[0007] According to
EP 0 535 425 B1 loudness curves as a function of the sound input level are measured for various frequencies.
From these loudness curves contours of equal loudness as a function of frequency are
plotted for various loudness values. However, the loudness curves are obtained without
the hearing aid being used.
[0008] DE 100 41 726 C1 relates to an implanted hearing instrument with an electromechanical transducer,
wherein the quality of the coupling between the transducer and the user's ear is evaluated
by measuring the mechanical impedance after implantation of the transducer.
[0009] EP 0 661 905 B1 relates to a fitting model for hearing aids in order to take into account various
psycho-acoustic effects, i.e. in order to take into account the fact that loudness
curves are measured with sinus tones or low-band noise while practical ambient sound,
in particular speech, is perceived by the user in a much more complex manner than
sinus tones or narrow-band noise.
[0010] Fitting procedures for hearing instruments with electroacoustic output transducers
are suggested in
US 6,574,342 B1 and
US 6,201,875 B1 which propose to perform measurements of the contours of equal loudness with the
hearing instrument worn by the user and with the stimuli created by the hearing instrument
itself. In these measurements, for a plurality of frequencies f or frequency bands,
the input level of the hearing instrument is varied such that the loudness perceived
by the user is kept constant. This procedure is repeated for different values/categories
of loudness ranging from very soft to extremely loud. The thus measured contours of
equal loudness are used to determine the individual gain function which is finally
implemented in the hearing instrument to compensate for the user's individual hearing
loss.
[0011] While using the fitting procedures suggested in
US 6,574,342 B1 and
US 6,201,875 B1 the hearing device can be well fitted to the individual hearing loss experienced
by the patient, these procedures are disadvantageous in that a large number of measurements
has to be taken. In particular, in the fitting procedures described in
US 6,574,342 B1 and
US 6,201,875 B1 measurements are taken at 12 different frequencies for 7 different loudness levels.
That means that 84 individual reading points are investigated, which necessitates
a lengthy and troublesome procedure both for the patient and the physician or audiologist.
[0012] It is an object of the invention to provide for a simple and nevertheless accurate
method for individually fitting a hearing instrument comprising an output transducer
for stimulation of the human auditory system.
[0013] This object is achieved by a method as defined in claims 1 and 21, respectively.
[0014] The solution according to claim 1 is beneficial in that, by measuring the perceived
loudness at the intermediate loudness level for a larger number of frequencies or
frequency bands and with a finer frequency resolution than at said low and high loudness
levels and calculating the individual gain function to be implemented in the audio
signal processing unit in order to achieve the pre-defined target loudness function
using the transducer input audio signal levels taken during such measurements, the
number of measurement points to be investigated can be substantially reduced.
[0015] The present invention takes advantage of the finding that the fine frequency dependency
of the overall transfer function, i.e. that part of the frequency dependency which
varies strongly/steeply within short frequency intervals (in other words, the short
scale variations), is relatively similar for even significantly different perceived
loudness levels of the signal. It has been found that thus, instead of individually
determining the transfer functions for every single loudness category by taking measurements
at a high frequency resolution for each loudness category of interest, it is usually
sufficient to determine the transfer function only for a single intermediate loudness
level at a higher frequency resolution, whereas for the low and high loudness levels
measurement with low frequency resolution at only a few frequencies/frequency bands,
i.e. at only a few measurement points is sufficient, since the actual short scale
variations with frequency for those low and high loudness levels usually will be similar
to that having been measured for the intermediate loudness level, which thus can be
estimated by transferring that having been measured for the intermediate loudness
level to the low and high loudness levels. Thus, having determined the frequency dependency
of the overall transfer function by conducting measurements with a high frequency
resolution at one loudness level, such as the most comfortable level (MCL), it is
thus possible to quite precisely interpolate, preferably by linear interpolation,
the frequency dependency also for loudness levels in which measurements were taken
only for a few frequencies
[0016] It is noted that the finding that the coupling transfer function from the output
transducer to the user's hearing usually depends strongly on frequency but less on
loudness level, while the hearing loss of the patient usually depends strongly on
loudness level but less on frequency, enables separation of these two components by
calculations so that common fitting software tools, which usually need these two components
as separate input, can be used.
[0017] Preferably, in such measurements the intermediate loudness level is the most comfortable
level, which may be from 60 to 70 phon and which is the input sound pressure level
at which intelligibility of the stimulus by the user is best and to which the user
could comfortably listen over an extended period of time. The low loudness level preferably
is the hearing threshold, which is the input sound pressure level at which the stimulus
becomes detectable by the user, and the high loudness level preferably is the uncomfortable
level (UCL), which is the input sound pressure level at which loudness becomes uncomfortable
to the user and the sensation could not be tolerated for an extended period of time.
[0018] Whereas for the intermediate loudness level the transducer input audio signal level
should preferably be measured at at least 5 different frequencies or frequency bands,
respectively, for the low and/or high loudness level measurements at 3 to 5 frequencies
or frequency bands can be sufficient in the practice of the invention. In the range
of 0.75 to 3 kHz which is the range into which usually the resonance of an electromechanical
output transducer falls so that the variation of the overall transfer function due
to individual spread of the output transducer resonance can be eliminated, each contour
of equal loudness preferably is measured at at least 5 different frequencies or frequency
bands, respectively.
[0019] Preferably, in order to obtain a good frequency resolution, the transducer input
audio signal level for the intermediate loudness level is measured for at least 8
frequencies or frequency bands. An even finer frequency resolution can be obtained
by increasing the number of frequencies at which the loudness perception is measured,
such as by measuring the transducer input audio signal level for the intermediate
loudness level for at least 15 frequencies or frequency bands.
[0020] In a preferred embodiment of the invention the transducer input audio signal level
is measured for each loudness level for frequencies or frequency bands in a range
of from 100 to 10,000 Hz. Preferably, the frequencies or frequency bands are spaced
in equal distances in the range of from 100 to 10,000 Hz.
[0021] In a particularly preferred embodiment of the invention the frequency dependence
of the values of the transducer input audio signal level as measured for the intermediate
loudness level is used to interpolate between the values of the transducer input audio
signal level to be applied to the transducer input as measured for the low and the
high loudness level. In this manner the present method enables to obtain functions
representing the frequency dependency of the loudness perception at a high frequency
resolution also for loudness categories in which readings are taken at a substantially
lower frequency resolution, i.e. by taking measurements only for a few frequencies.
[0022] Preferably, the measurements for the intermediate loudness level are conducted as
an equal loudness contour measurement, that is as a measurement wherein subsequently
for each frequency or frequency band a transducer input audio signal level is selected
which causes a constant level of loudness perception for the user.
[0023] Whereas for taking the measurements for the intermediate loudness level preferably
narrow band noise is used, the measurements for the low and high loudness levels preferably
are conducted with pure sinus tones.Instead of measuring ― according to the solution
defined in claim 1 - the transducer input audio signal level which has to be applied
to the transducer input in order to achieve a certain intermediate perceived loudness
level, according to the solution defined in claim 21 a predetermined level of the
processed output audio signal at a number of frequencies or frequency bands is present
to the user and then the loudness level perceived by the user at the respective frequency
or frequency band is measured. Using the thus obtained relationship between the level
of the processed output audio signal, which is used as input signal for the transducer,
and the perceived loudness level the overall transfer function can be determined.
[0024] In a preferred embodiment of the present invention transducer input audio signal
levels are measured for the low and high loudness levels as an equal loudness contour
measurement, wherein subsequently for each frequency or frequency band the transducer
input audio signal level is selected such that the same loudness level is perceived
by the user. Then a preliminary individual gain function is calculated by taking into
account the measured transducer input audio signal levels for the low and high loudness
levels, so as to achieve a pre-defined target loudness function which at least in
a range of medium input sound pressure levels corresponds to the standard loudness
function of a normal hearing person. In a further step, the contour of equal loudness
is estimated for the intermediate loudness level from the preliminary individual gain
function, and the individual gain function is calculated by correcting the preliminary
individual gain function by taking into account the difference between the contour
of equal loudness measured for the intermediate loudness and the estimated contour
of equal loudness.
[0025] In this embodiment of the invention thus a two-stage fitting procedure is employed,
wherein in-situ ELC measurements are made in the second stage, whereby a particularly
simple but nevertheless accurate fitting procedure is provided which does not require
knowledge of the transfer function of the components of the hearing instrument; in
particular, it does not require knowledge the transfer function of the electromechanical
or electroacoustical output transducer and the coupling to the user's anatomy. In
particular, with the help of the ELC measurements in the second stage of the fitting
procedure the hearing instrument can be compensated for the usually unknown and significantly
spreading resonance of the electromechanical output transducer in the range of 0.75
to 3 kHz.
[0026] The reason for performing in this embodiment initial coarse audiogram measurements
is to determine preliminary fitting parameters in order to enable accurate measurements
of the contour of equal loudness in the second step, with the hearing instrument already
being operated in a manner taking into account those preliminary fitting parameters
(e.g. hearing threshold, MCL, UCL). Thereby it becomes possible to perform the contour
of equal loudness measurements with the hearing instrument already being operated
in a manner more or less close to the finally fitted hearing instrument.
[0027] Generally, the measurement of the contour of equal loudness should have a finer frequency
resolution than the initial audiogram measurements, thereby serving as a correction
of the initial, relatively coarse audiogram measurement. To this end, the number of
frequencies or frequency bands at which the contour of equal loudness is measured
preferably is larger than the number of frequencies or frequency bands at which the
initial audiogram measurement is performed.
[0028] When estimating the contour of equal loudness from the initial loudness measurements,
this can be done by linear interpolation between the measured frequencies or frequency
bands, respectively.
[0029] Preferably, the initial audiogram measurements are performed with pure sinus tones,
while the measurement of the contour of equal loudness is performed with narrow-band
noise.
[0030] The target loudness function at least in the range of medium input sound pressure
levels preferably corresponds to the standard loudness function of a normal hearing
person. For medium input sound pressure levels an individual gain function thus can
be determined by adding the difference between the standard loudness function of a
normal hearing person and the initially determined individual loudness function being
derived from the audiogram data to the preset standard gain function.
[0031] For low and high input sound pressure levels the gain in the target loudness function
may be progressively reduced compared to the gain for medium input sound pressure
levels, i.e. for low and high sound input pressure levels the gain may be smaller
than the sum of the difference between the standard loudness function of a normal
hearing person and the determined individual loudness function and the preset standard
gain function. For low input sound pressure levels, the gain in the target loudness
function may be progressively reduced towards low input sound pressure levels, while
for high input sound pressure levels the gain in the target loudness function may
be progressively reduced towards high input sound pressure levels. Above a given high
input sound pressure level the gain may be reduced below zero in order to provide
for a maximum power output limitation, so that the hearing instrument saturates at
very high input sound pressure levels.
[0032] While the transducer input audio signal used in the measurements can be generated
by providing corresponding sound to the microphone, preferably the stimulus is generated
by the audio signal processing unit itself. To this end, the audio signal processing
unit can be provided with a sound generator. Such measurements, which are more accurate
and reproducible than earphone measurements, are made possible by the fact that usually
the transfer function of ambient sound to the audio signal processing unit via the
microphone is known.
[0033] According to a preferred embodiment, an electromechanical output transducer is used
which is directly connected, via an artificial incus, with the stapes or the footplate
of the stapes or with the round window or an artificial window of the cochlear wall.
Such hearing instruments also are known as DACS (Direct Acoustic Cochlear Stimulator).
[0034] However, the fitting method according to the invention also can be used for hearing
instruments with electroacoustic output transducer or for cochlea implants.
[0035] In the following, examples of the invention will be described in more detail by reference
to the attached drawings.
- Fig. 1
- is a schematic view of a hearing instrument according to the invention;
- Fig. 2
- is an example of the spread of the electromechanical output transducer transfer function;
- Fig. 3
- shows an example of the loudness curve of a normal-hearing person and an measured
individual loudness curve of a hearing impaired person using a hearing instrument
operated at a preset standard gain function;
- Fig. 4
- shows an example of a preliminary individual gain function of the hearing instrument
derived from the measured individual loudness curves;
- Fig. 5
- shows an example of hearing instrument output curves estimated from the measurements
of the individual loudness curves as a function of frequency for several input levels;
- Fig. 6
- shows an example of an equal loudness contour of a normal hearing person at 65 phon,
wherein the input level necessary to achieve this loudness is given as a function
of frequency, together with arrows indicating the difference between a measured equal
loudness contour of the person using the hearing instrument operated at the preliminary
individual gain function to an equal loudness contour estimated by frequency interpolation
from the initial loudness measurements;
- Fig. 7
- shows how the obtained differences of the individually measured equal loudness contour
curve and the estimated equal loudness contour curve, i.e. the arrows of Fig. 6, are
used to correct the 75 dB input level curve;
- Fig. 8
- shows the hearing instrument output curves of Fig. 5 after having been corrected according
to the equal loudness contour measurements;
- Fig. 9
- is a diagram illustrating a signal conversion as performed in a hearing instrument;
- Fig. 10
- is a graphical representation of the points of measurement;
- Fig. 11A
- shows a measuring curve obtained at the most comfortable level (MCL);
- Fig. 11B
- shows the results obtained by the measurements taken at the most uncomfortable level
(UCL), wherein the data curve is obtained by interpolation between the measuring data
using the curve shown in Fig. 11A; and
- Fig. 11B
- shows the results obtained by the measurements taken at the hearing threshold (THR),
wherein the data curve is obtained by interpolation between the measuring data using
the curve shown in Fig. 11A.
[0036] To facilitate understanding of the fitting procedure, signal conversion as performed
in a hearing instrument will be explained below by reference to Fig. 9. As it is schematically
shown in Fig. 9 a sound level L
0 is applied to a microphone M which is arranged in an environment U. Microphone M
converts the sound signal into an electric signal level L
1, which by means of an audio signal processing unit E is converted into an electric
signal level L
2 to be applied as input signal to an output transducer TD. Output transducer TD, which
can be an electroacoustic transducer (i.e. a speaker/receiver), an electrode for direct
electric stimulation of the cochlea or an electromechanical output transducer for
direct mechanical stimulation of the middle ear or the inner ear, and which in case
that the output transducer TD is an electrode or an electromechanical output transducer
has to be implanted, is coupled to the hearing apparatus EAR of the patient. In dependency
of the type of output transducer used, the coupling , for example, may be acoustically
via the tympanic membrane, or mechanically via the stapes or the oval window, so as
to generate within the hearing apparatus EAR a stimulus which is perceived by the
patient as loudness sensation L
3.
[0037] Conversion of the original sound level L
0 into the loudness sensation L
3 perceived by the patient involves a number of transfer functions which also are indicated
in Fig. 9. In particular, conversion of the original sound level L
0 into the electric signal level L
1 is governed by a transfer function T
01 which basically is dependent on the frequency of the signal presented to the microphone
and thus can be assumed to be known. The transfer function T
12 which describes conversion of electric signal level L
1 to processed electric signal level L
2 to be applied as input signal to output transducer TD can be adjusted by means of
audio signal processing unit E. For audio signal measurements, instead of using an
airborne sound signal to be picked up by microphone M, there can be provided a signal
processor SG which feeds a known sound level L
2 to output transducer TD. The transfer function T
23 which associates a certain loudness perception to a certain input signal level L
2 of the transducer TD generally is not known and depends on the individual circumstances
of the patient. In particular, transfer function T
23 combines a coupling portion T
C which accounts for the transducer resonance and the coupling of the transducer to
the anatomic structures of the patient as well as a hearing loss portion T
HL which represents the individual hearing loss experienced by the patient. In order
to be able to determine the overall transfer function T
03 by which conversion of a sound event into a hearing impression can be described and
which is composed of the above partial transfer function T
01, T
12 and T
23, transfer function T
23 has to be determined in the course of the fitting procedure.
[0038] In view of the above the fitting procedure aims at adjusting the audio signal processing
unit E (and hence transfer function T
12) such that the overall transfer function T
03 (and hence association of a certain loudness perception L
3 to a certain input signal level L
2) assumes a certain shape, which often, at least for intermediate loudness levels,
approximates the overall transfer function T
03 that is realized in normal healthy hearing. For low and high loudness levels often
an overall transfer function T
03 is preferred which differs from that achieved in normal hearing. Thus, whereas for
low loudness levels often a so-called "Soft Squelch" function is implemented by which
the gain function is progressively reduced towards low input sound levels, for high
loudness levels many patients prefer a limitation of the loudness level, i.e. a compression
of the gain function.
[0039] In the course of the fitting procedure first the initially unknown transfer function
T
23 is determined with the aid of audiologic measurements. In a second step, using the
known microphone transfer function T
01 and the required transfer function T
12 of the audio signal processing unit E as a function of loudness level and frequency,
a desired overall transfer function T
03 can be calculated and implemented in the audio signal processing unit E.
[0040] Fig. 1 is a schematic view of an example of a hearing instrument according to the
invention comprising an external part 10 and an implantable part 12 which are connected
via a percutaneous plug 14. The external part 10 comprises a housing 16 to be worn
somewhere at the user's body, for example, behind the ear. The housing 16 forms a
control unit 18 which comprises at least one microphone 20 for converting ambient
sound into an input audio signal, a battery 22, a data memory 24 and a digital audio
signal processing unit 26. The digital audio signal processing unit 26 is for processing
the audio input signal provided by the microphone 20 into a processed output audio
signal by applying a gain function, which depends on frequency and audio signal input
level, to the input audio signal provided by the microphone 20. The gain function,
together with other operating parameters and the operating program for the digital
audio signal processing unit 26, may be stored in the memory 24. The digital audio
signal processing unit also may comprise a sound generator 28. In an alternative embodiment
the sound generator 28 may be provided separate from the digital audio signal processing
unit 26.
[0041] The control unit 18 is connected to the percutaneous plug 14 via a tube 30 which
houses wires for providing the output audio signal from the digital audio signal processing
unit 26 to an electromechanical output transducer 32 and for supplying the electromechanical
output transducer 32 with power from the battery 22. To this end, the output transducer
32 is electrically connected to the percutaneous plug 14 via a tube 34. The implantable
part 12 consists of the output transducer 32, the tube 34 and the implantable part
of the plug 14. The implantable part 12 is implanted into the skull of the user, with
the output transducer 32 comprising a bone plate 36 which is fixed at the user's skull.
[0042] The output transducer comprises a housing 38 comprising a drive 40 for driving a
rod 42 for reciprocating movement. The free end of the rod 42 is provided with an
artificial incus 44 which is to be mechanically connected to the cochlea of the user.
The fixation of the artificial incus 44 at the user's cochlea can be achieved by surgical
techniques which are known as stapedotomy or stapedectomy. Conventionally, these techniques
are used for connecting the artificial incus of a middle ear prosthesis to the patient's
stapes (stapedotomy) or footplate of the stapes (stapedectomy) when treating otosclerosis.
[0043] The drive 40 may be an electromagnetic drive (an example of which is described in
US 6,315,710 B1) or a piezoelectric drive (an example of which is described in
US 6,554,762 B2).
[0044] The hearing instrument of Fig. 1 is particularly suited for patients who cannot be
effectively treated with electroacoustic hearing aids alone and therefore would require
surgery anyhow. The hearing instrument used in the present invention completely bypasses
the middle ear and thus does not require a functional middle ear.
[0045] Fitting of such hearing instrument to the individual user, i.e. determination of
the most appropriate gain function, is critical for several reasons. First, the transducer
resonance may spread significantly from device to device (this is shown by two examples
in Fig. 2). Second, the coupling of the output transducer to the cochlea is not known
and may spread significantly from case to case. Third, the output of the output transducer
is not available in acoustic form.
[0046] Therefore it is necessary to perform a fitting procedure in which the gain function
which the hearing instrument applies to the sound signal is adapted to the individual
circumstances and requirements of the user of the hearing instrument.
[0047] In such fitting procedure audiogram measurements are made wherein loudness perception
of a stimulus by the user is tested when using the hearing instrument. In particular,
upon having pre-defined a desired target loudness function, measurements are taken
of the transducer input audio signal level which has to be applied to the transducer
input in order to achieve a certain intermediate perceived loudness level, which preferably
is the most comfortable level (MCL). These measurements are repeated for a number
of frequencies or frequency bands. Most preferably the measurements for the intermediate
loudness level are conducted with a frequency resolution which corresponds to the
frequency resolution of the hearing instrument.
[0048] As it is indicated in Fig. 10, which illustrates the points where measurements are
taken in terms of stimulus frequency f and perceived loudness L
3, the measurements of the perceived loudness level preferably are conducted as an
equal loudness contour measurement, wherein a transducer input audio signal level
L
2 is selected such that the same loudness level L
3 is perceived by the user.
[0049] Fig. 11A illustrates an exemplary chart of results obtained by the measurements taken
at the most comfortable level (MCL) and indicates for each frequency tested the respective
transducer input audio signal level L
2 that is required to obtain that the constant loudness level L
3. As can be seen from Fig. 11 A, the transfer function T
23 which describes the relationship between the transducer input audio signal level
L
2 and the perceived loudness level L
3 is not a linear function but usually shows large variations over the tested frequency
range.
[0050] Since, however, such variation of the perceived loudness level L
3 in the course of the frequency range was found to be largely independent from the
loudness level being tested, in the method suggested herein the measurements for low
and high loudness levels such as the hearing threshold (THR) and the uncomfortable
level (UCL), respectively are restricted to only a few measurement points, for example
to only three frequencies, as it is indicated in Fig. 10.
[0051] As it is shown in Figs. 11B and Fig. 11C, using the frequency dependency obtained
at high frequency resolution for the intermediate perceived loudness level shown in
Fig. 11, curves for the low and high loudness levels are obtained by interpolation
between the measurement values taken for the low and high loudness levels at the lower
frequency resolution.
[0052] In a practical example, the present fitting procedure further may be designed as
a two-stage fitting procedure, wherein at the first stage the hearing instrument is
used in order to perform audiogram measurements at a few frequencies, whereby some
points of the individual loudness curve versus sound input level are obtained between
which the individual loudness curve is interpolated. From the individual loudness
curve a preliminary individual gain function is calculated which may be used for operating
the hearing instrument at the second stage (in particular if the stimulus is provided
by an earphone to the microphone 20) wherein at least one contour of equal loudness
is measured with a finer frequency resolution than that of the loudness curve of the
first stage. The measured contour of equal loudness then is used for correcting the
individual preliminary gain function, in particular in between the frequencies already
measured in the first stage, in order to consider, for example, the relatively sharp
resonance of the output transducer. Thereby from the individual preliminary gain function
obtained in the first stage a corrected individual gain function is determined in
the second stage, which then is finally used for operating the hearing instrument.
[0053] The audiogram measurements may be performed such that for each frequency at least
two points of the loudness curve are determined, usually the hearing threshold (denoted
by A in Fig. 3) and at least one of MCL (denoted by B) and UCL (denoted by C). Such
loudness curve as shown in Fig. 3 should be determined at least for four different
frequencies spread over the most relevant part of the audible frequencies. The loudness
measurements of the first stage are performed with pure sinus tones. While in principle
it would be possible to provide the stimulus by an earphone to the microphone 20 (in
that case a standard gain function would be used for operating the hearing instrument
in the first stage, which preferably is linear with respect to sound input level),
it is preferred to generate the stimulus by the sound generator 28 within the control
unit 18.
[0054] For determining the preliminary individual gain function, at each test frequency
the difference between the measured individual loudness curve and the standard loudness
curve of the normal hearing person, i.e. the difference in input level necessary for
obtaining the same loudness perception, is considered. This is indicated by arrows
D1 and D2 in Fig. 3. To this end, the individual loudness curve is interpolated linearly
between the measured test input levels. Each input level difference D1 and D2 corresponds
to the necessary additional gain at the respective input level of the standard loudness
curve (which is labeled S in Fig. 3). The result is shown in Fig. 4 wherein the additional
gain relative to the standard gain function necessary for approaching the loudness
curve of a normal-hearing person is shown for a given frequency as a function of the
input level.
[0055] For low input levels the gain may be progressively reduced towards low input levels
regarding the values obtained from Fig. 3 in order to implement a function which is
known as "soft squelch" and which serves to reduce or eliminate microphone noise otherwise
occurring at very low input levels. At high input levels the gain may be progressively
reduced towards high input levels relative to the gain determined from Fig. 3 in order
to implement a "maximum power output" (MPO) function which serves to avoid uncomfortably
high loudness values so that the UCL should not be exceeded.
[0056] From this first stage measurements for each test frequency an individual preliminary
gain function is obtained by adding the standard gain function used during the audiogram
measurements to the gain curve shown in Fig. 4.
[0057] The obtained data could be represented in an alternative manner as shown in Fig.
5, wherein the transducer output level is plotted as function of frequency for various
input levels. Between the test frequencies f
0 to f
3 the values have been interpolated linearly. The transducer output level shown in
Fig. 5 corresponds to the preliminary individual gain of the hearing instrument plus
the input level.
[0058] At the second stage of the fitting procedure, the hearing instrument is operated
with the preliminary individual gain function determined at the first stage in order
to measure at least one contour of equal loudness (however, use of preliminary individual
gain function for operating the hearing instrument is not necessary if the stimulus
is generated by the sound generator 28). Preferably, the contour of equal loudness
is measured at the MCL, for example, at 65 phon. In contrast to the audiogram measurements
of the first stage, narrow-band noise other than pure sinus tones is used as the stimulus.
Analogously to the first stage measurements, the stimulus preferably is provided by
the internal sound generator 28 of the control unit 18. In order to determine the
contour of equal loudness, for a number of test frequencies the input level of the
stimulus is varied until the desired loudness is perceived by the user.
[0059] The test frequencies for the ELC measurements are selected such that the frequency
resolution is improved regarding the audiogram measurements of the first stage. In
particular, between two of the test frequencies of the first stage at least one test
frequency of ELC measurement should be located. In the example shown in Fig. 6, two
additional test frequencies of the second stage are located between each pair of adjacent
test frequencies of the first stage. Generally, the number of test frequencies of
the ELC measurements is higher than the number of test frequencies of the first stage
loudness measurements. In the example shown in Fig. 6, twenty test frequencies are
used between 0.125 and 8 kHz. The solid line in Fig. 6 shows an example of an ELC
of a normal hearing person. The arrows in Fig. 6 represent the difference between
the measured ELC and the ELC for the same loudness as estimated from the loudness
measurements of the first stage by linear interpolation between the test frequencies
of the first stage loudness measurements. In other words, the arrows of Fig. 6 essentially
show the deviation of the actually measured ELC from the linear interpolation. However,
even if the test frequencies of the first and second stage measurements coincide,
there may be some deviation, since the first stage loudness measurements were performed
with pure sinus tones, while the second stage ELC measurements were performed with
narrow-band noise, which different stimuli may cause different loudness perception
even for the same input level.
[0060] It is obvious from Fig. 6 that in the region around 2 kHz the most pronounced deviation
from the linear interpolation is observed, which is due to the relatively sharp resonance
of the output transducer 32 in that frequency range. Preferably, the ELC measurements
include at least five test frequencies between 0.75 and 3 kHz in order to be able
to compensate the resonance of the output transducer 32 accurately.
[0061] Fig. 7 is similar to Fig. 5, with the arrows of Fig. 6 having been added to the 75
dB input level curve (the transducer output level is input level times gain provided
by the hearing instrument).
[0062] Fig. 8 is a representation similar to Fig. 7, wherein the transducer output level
curves have been corrected according to the ELC measurement arrows, with the regions
between the test frequencies having been interpolated. Since the output transducer
resonance is expected to be linear regarding input level, the corrections obtained
from this single ELC measurement can be extrapolated linearly to ELC at other loudness
values, so that measurement of ELC for one single loudness is sufficient. This is
how the corrected curves at input levels other than 75 dB of Fig. 8 are obtained.
[0063] From the transducer output level curves of Fig. 8 the corrected individual gain function
can be determined, since the gain function is the ratio of the transducer output level
to the input level.
[0064] Finally, the hearing instrument is operated with the corrected individual gain function
obtained by the above-described fitting procedure.
1. A method for individually fitting a hearing instrument (10, 12) to a user, comprising
at least one microphone (20) for generating an input audio signal from ambient sound,
an audio signal processing unit (26) for processing the input audio signal into a
processed output audio signal, and a transducer for stimulation of the human auditory
system according to the processed output audio signal as input to said transducer,
the method comprising:
(a) providing the user with the hearing instrument and starting operation of the hearing
instrument;
(b) pre-defining a desired target loudness function, wherein loudness perception of
a stimulus by the user when using the hearing instrument is defined as function of
frequency and input sound pressure level at the microphone;
(c) measuring for a given measurement parameter set of perceived loudness levels and
frequencies or frequency bands the respective transducer input audio signal level
to be applied to the transducer input in order to achieve the respective perceived
loudness level at the respective frequency or frequency band, said measurement parameter
set comprising at least a low loudness level, an intermediate loudness level and a
high loudness level, and said intermediate loudness level being measured for a larger
number of frequencies or frequency bands and with a finer frequency resolution than
said low and high loudness levels;
(d) calculating an individual gain function to be implemented in the audio signal
processing unit in order to achieve the pre-defined target loudness function of step
(b) by taking into account the measured transducer input audio signal levels of step
(c);
(e) operating the hearing instrument with the individual gain function of step (d).
2. The method of claim 1, wherein in step (c) the intermediate loudness level is the
most comfortable level, which is the input sound pressure level at which intelligibility
of the stimulus by the user is best and to which the user could comfortably listen
over an extended period of time,
3. The method of claim 1 or 2, wherein in step (c) the low loudness level is the hearing
threshold, which is the input sound pressure level at which the stimulus becomes detectable
by the user.
4. The method of one of the preceding claims, wherein in step (c) the high loudness level
is the uncomfortable level, which is the input sound pressure level at which loudness
becomes uncomfortable to the user and the sensation could not be tolerated for an
extended period of time.
5. The method of one of the preceding claims, wherein the transducer input audio signal
level is measured in step (c) for the intermediate loudness level for at least 8 frequencies
or frequency bands.
6. The method of claim 5, wherein the transducer input audio signal level is measured
in step (c) for the intermediate loudness level for at least 15 frequencies or frequency
bands.
7. The method of one of the preceding claims, wherein the transducer input audio signal
level is measured in step (c) for the intermediate loudness level at at least 5 different
frequencies or frequency bands, respectively, in the range from 0.75 to 3 kHz.
8. The method of one of the preceding claims, wherein the transducer input audio signal
level is measured in step (c) for the low and/or high loudness level for 3 to 5 frequencies
or frequency bands.
9. The method of claim 8, wherein the transducer input audio signal level is measured
in step (c) for each loudness level except for said intermediate loudness level for
3 to 5 frequencies or frequency bands.
10. The method of one of the preceding claims, wherein the transducer input audio signal
level is measured in step (c) for each loudness level for frequencies or frequency
bands in a range from 100 to 10,000 Hz.
11. The method of claim 10, wherein the transducer input audio signal level is measured
in step (c) for each loudness level for frequencies or frequency bands which are spaced
in equal distances in a range from 100 to 10,000 Hz.
12. The method of one of the preceding claims, wherein the frequency dependence of the
values of the transducer input audio signal level as measured in step (c) for the
intermediate loudness level is used to interpolate between the values of the transducer
input audio signal level to be applied to the transducer input as measured in step
(c) for the low and the high loudness level.
13. The method of one of the preceding claims, wherein the measurements of step (c) for
the intermediate loudness level are conducted as an equal loudness contour measurement,
wherein subsequently for each frequency or frequency band the transducer input audio
signal level is selected such that the same loudness level is perceived by the user.
14. The method of one of the preceding claims, wherein the measurements of step (c) for
the low and high loudness levels are conducted as a series of constant frequency measurements
wherein for each frequency or frequency band the transducer input audio signal level
is selected such that first the low and then the high loudness level or first the
high and then the low loudness level is perceived by the user.
15. The method of one of the preceding claims, wherein the measurements of step (c) for
the intermediate loudness level are conducted with narrow band noise.
16. The method of one of the preceding claims, wherein the measurements of step (c) for
the low and high loudness levels are conducted with pure sinus tones.
17. The method of one of the preceding claims, wherein the values of the transducer input
audio signal measured in step (c) for each frequency or frequency band for the intermediate
loudness level are interpolated linearly.
18. The method of one of the preceding claims, wherein in step (c) said measurement parameter
set comprises only said low, intermediate and high loudness levels.
19. The method of one of the preceding claims, wherein the frequency resolution of the
measurements in step (c) for the intermediate loudness level corresponds to the frequency
resolution of the hearing instrument.
20. The method of one of the preceding claims, wherein
in step (c) transducer input audio signal levels are measured for the low and high
loudness levels as an equal loudness contour measurement, wherein subsequently for
each frequency or frequency band the transducer input audio signal level is selected
such that the same loudness level is perceived by the user
a preliminary individual gain function is calculated by taking into account the measured
transducer input audio signal levels for the low and high loudness levels, so as to
achieve a pre-defined target loudness function which at least in a range of medium
input sound pressure levels corresponds to the standard loudness function of a normal
hearing person;
the contour of equal loudness is estimated for the intermediate loudness level from
the preliminary individual gain function; and
the individual gain function is calculated by correcting the preliminary individual
gain function by taking into account the difference between the contour of equal loudness
measured in step (c) for the intermediate loudness and the estimated contour of equal
loudness.
21. A method for individually fitting a hearing instrument (10, 12) to a user, comprising
at least one microphone (20) for generating an input audio signal from ambient sound,
an audio signal processing unit (26) for processing the input audio signal into a
processed output audio signal, and a transducer for stimulation of the human auditory
system according to the processed output audio signal as input to said transducer,
the method comprising:
(a) providing the user with the hearing instrument and starting operation of the hearing
instrument;
(b) pre-defining a desired target loudness function, wherein loudness perception of
a stimulus by the user when using the hearing instrument is defined as function of
frequency and input sound pressure level at the microphone;
(c) measuring for a given measurement parameter set of levels of the processed output
audio signal and frequencies or frequency bands the loudness level perceived by the
user at the respective frequency or frequency band, said measurement parameter set
comprising at least a low audio signal level, an intermediate audio signal level and
a high audio signal level, and said intermediate audio signal level being measured
for a larger number of frequencies or frequency bands and with a finer frequency resolution
than said low and high audio signal levels;
(d) calculating an individual gain function to be implemented in the audio signal
processing unit in order to achieve the pre-defined target loudness function of step
(b) by taking into account the perceived loudness levels measured in step (c);
(e) operating the hearing instrument with the individual gain function of step (d).
22. The method of claim 21, wherein the perceived loudness level is measured in step (c)
for the intermediate audio signal level for at least 8 frequencies or frequency bands.
23. The method of claim 22, wherein the perceived loudness level is measured in step (c)
for the intermediate audio signal level for at least 15 frequencies or frequency bands.
24. The method of one of claims 21 to 23, wherein the perceived loudness level is measured
in step (c) for the intermediate audio signal level at at least 5 different frequencies
or frequency bands, respectively, in the range from 0.75 to 3 kHz
25. The method of one of claims 21 to 24, wherein the perceived loudness level is measured
in step (c) for the low and/or high audio signal level for 3 to 5 frequencies or frequency
bands.
26. The method of claim 24, wherein the perceived loudness level is measured in step (c)
for each loudness level except for said intermediate audio signal level for 3 to 5
frequencies or frequency bands.
27. The method of one of claims 21 to 26, wherein the perceived loudness level is measured
in step (c) for each audio signal level for frequencies or frequency bands in a range
from 100 to 10,000 Hz.
28. The method of claim 27, wherein the perceived loudness level is measured in step (c)
for each audio signal level for frequencies or frequency bands which are spaced in
equal distances in a range from 100 to 10,000 Hz.
29. The method of one of claims 21 to 28, wherein the frequency dependence of the values
of the perceived audio signal level as measured in step (c) for the intermediate audio
signal level is used to interpolate between the values of the perceived loudness level
measured in step (c) for the low and the high audio signal level.
30. The method of one of claims 21 to 29, wherein the measurements of step (c) for the
intermediate audio signal level are conducted with pure sinus tones.
31. The method of one of claims 21 to 30, wherein the measurements of step (c) for the
low and high audio signal levels are conducted with narrow band noise.
32. The method of one of claims 21 to 31, wherein the values of the perceived loudness
level measured in step (c) for each frequency or frequency band for the intermediate
audio signal level are interpolated linearly.
33. The method of one of the preceding claims, wherein the target loudness function at
least in a range of medium input sound pressure levels corresponds to the standard
loudness function of a normal hearing person.
34. The method of claim 33, wherein for low input sound pressure levels the target loudness
function is progressively reduced towards low input sound pressure levels with respect
to the standard loudness function of a normal hearing person.
35. The method of one of claims 33 and 34, wherein for high input sound pressure levels
the target loudness function is progressively reduced towards high input sound pressure
levels with respect to the standard loudness function of a normal hearing person.
36. The method of one of the preceding claims, wherein the transducer input audio signal
used in the measurements of step (c) is generated by the audio signal processing unit
(26).
37. The method of one of claims 1 to 35, wherein the transducer input audio signal used
in the measurements of step (c) is generated by providing corresponding sound to the
microphone (20).
38. The method of one of the preceding claims, wherein the output transducer is an electromechanical
output transducer (32) for direct mechanical stimulation of the middle ear or the
inner ear.
39. The method of claim 38, wherein the electromechanical output transducer (32) is directly
connected in step (a) with the stapes , footplate of stapes or the cochlea wall.
40. The method of one of claims 1 to 36, wherein the output transducer is an electroacoustic
output transducer.