CN102625220B - Method for determining hearing compensation gain of hearing-aid device - Google Patents

Abstract

The invention relates to a method for determining the hearing compensation gain of a hearing aid device, comprising the following steps: 1) testing the hearing of the subject to determine the frequency CF at which the gain compensation needs to be performed; 2) based on the loudness of normal hearing and impaired hearing The sensory model, aiming at the damage value HLohcdB(CF) of the outer ear hair cells at the hearing loss frequency CF, calculates the compensation gain value IGohcdB(CF) of the outer ear hair cell damage; 3) for the ear inner hair cells at the hearing loss frequency CF The damage value HLihcdB(CF), calculates the compensation gain value IGihcdB(CF) of hair cell damage in the ear; 4) When the hearing loss frequency CF is lower than 1000 Hz, the compensation gain at the frequency CF should be attenuated correspondingly according to the frequency CF, The attenuation value is INTdB(CF); when the hearing loss frequency CF is greater than or equal to 1000 Hz, the attenuation value at the frequency CF is 0; 5) Integrating steps 2) to 4), the compensation gain IGdB(CF at the frequency CF of the hearing loss person is obtained ). The invention can be widely used in hearing aids such as hearing aids.

Description

A kind of method of definite hearing-aid device hearing compensation gain

Technical field

The present invention relates to a kind of method of definite hearing-aid device parameter, a kind of method particularly gaining about definite hearing-aid device hearing compensation based on ear external hair cell and inner hair cell damage.

Background technology

Hearing loss (deaf) patient can carry out sound amplification processing to recover or partly to recover perception of sound by wearing hearing-aid device.Due to the complexity of people's ear, for different frequency place hearing loss in various degree, when different sound import, to give the compensation of amplification in various degree.In order to provide accurately and effectively compensating gain, the formula of the suitable compensating gain that people should give various hearing loss situations according to correlative study experience and achievement under various sound import conditions calls hearing aid and tests and join formula.Existing hearing aid test join formula be mainly based on the threshold of audibility calculate compensating gain, but the hearing loss information of threshold of audibility reflection is also insufficient, mainly because external hair cell damage and inner hair cell that hearing loss is ear damage coefficient result, and external hair cell damage and inner hair cell damage role in the process of hearing loss are also different, that is: external hair cell damage can cause the threshold of audibility to rise, audible frequencies resolution weakens with sense of hearing non-linear gain and declines, and the general threshold of audibility that only can cause of inner hair cell damage rises, therefore according to the result of threshold of audibility calculating compensating gain, not only very accurate.

Existing hearing compensation gain need to realize by multichannel amplification and compression algorithm in hearing-aid device, existing hearing aid is tested and is joined formula and generally only provide 50dB SPL(Sound Pressure Level), the compensating gain of the long-time average speech frequency spectrum under 65dB SPL and the 85dB SPL sound intensity, then by number of active lanes and passage infall frequency are set, and then compression threshold and the compression ratio of definite each passage, so that multichannel interpolator arithmetic is realized the compensation under above-mentioned three input sound intensity conditions, and the compensation while automatically calculating other input sound intensity according to above-mentioned rule, but adopt said method to have two problems: 1, can not select rapidly and accurately optimum number of active lanes and passage infall frequency, the compensating gain calculating when 2, other is different from the Speech input of long-time average speech frequency spectrum and desirable compensating gain may have deviation.

Summary of the invention

For the problems referred to above, the object of this invention is to provide a kind of can be fast, accurately determine the method for hearing-aid device hearing compensation gain.

For achieving the above object, the present invention takes following technical scheme: a kind of method of definite hearing-aid device hearing compensation gain, comprises the following steps: 1) testee's the sense of hearing is tested, determined the frequency CF that need to carry out gain compensation place; 2) loudness based on normal hearing and the damage sense of hearing is experienced model, for the impairment value HLohcdB (CF) of the ear external hair cell at acoustic trauma frequency CF place, calculates the compensating gain value IGohcdB (CF) that ear external hair cell damages; 3) for the ear inner hair cell impairment value HLihcdB (CF) at acoustic trauma frequency CF place, calculate the compensating gain value IGihcdB (CF) of ear inner hair cell damage; 4) when hearing impairment frequency CF is during lower than 1000Hz, the compensating gain at frequency CF place will carry out corresponding decay according to frequency CF, and pad value is INTdB (CF); When hearing impairment frequency CF is more than or equal to 1000Hz, frequency CF place pad value is 0; 5) comprehensive step 2)～4), obtain hearing loss person and at the compensating gain IGdB at frequency CF place (CF) be:

Step 2) in the compensating gain value IGohcdB (CF) of ear external hair cell damage be:

IGohcdB(CF)＝m[GdBN(CF)-GdBI(CF)]

In formula, m is coefficient, and the computing formula of m is:

$m=\left[\frac{0.5}{\mathrm{GdBN}\left(\mathrm{CF}\right)-(-13)}\right]\mathrm{GdBI}\left(\mathrm{CF}\right)+[1-\frac{0.5\·\mathrm{GdBN}\left(\mathrm{Cf}\right)}{\mathrm{GdBN}\left(\mathrm{CF}\right)-(-13)}]$

In formula ,-13dB is the initiatively minimum value of gain of the sense of hearing of setting, the active gain that GdBI (CF) is damage ear, the active gain that GdBN (CF) is normal ear.

The compensating gain value IGihcdB (CF) of step 3) middle ear inner hair cell damage:

IGihcdB(CF)＝0.5·max[HLihcdB(CF),40]

In formula, max represents to get the maximum of HLihcdB (CF) and 40 in the two.

In step 4), the pad value of the compensating gain at different frequency CF place is respectively: frequency CF is that the pad value INTdB (CF) that 125HZ is corresponding is-15dB, frequency CF is that the pad value INTdB (CF) that 250HZ is corresponding is-10dB, frequency CF is that the pad value INTdB (CF) that 500HZ is corresponding is-5dB, frequency CF is that the pad value INTdB (CF) that 1000HZ is corresponding is 0dB, and other pad value INTdB (CF) lower than other frequency CF at 1000HZ frequency place passes through interpolation calculation.

The present invention is owing to taking above technical scheme, it has the following advantages: 1, the present invention experiences model with the loudness of existing normal hearing and the damage sense of hearing and does basis, according to hearing loss person's ear external hair cell and inner hair cell damage, to the different qualities of hearing impairment, give corresponding compensation, and take the hearing impairment of ear external hair cell and inner hair cell, as basis, show that hearing aid is tested and join formula, and obtain final compensating gain, therefore for hearing loss, can give to compensate more accurately and effectively, having solved existing hearing aid tests and joins the not accurate enough problem of compensating gain causing when formula only compensates according to the threshold of audibility.2, the present invention has comprised passage processing and compression amplification processing procedure in the process of ear external hair cell damage compensating gain of calculating different frequency place, this process is simulated physiology course completely, thereby can in hearing-aid device, reach better desirable Compensation Objectives when inputting any sound, therefore need to just can not calculate fast the needed compensating gain of hearing loss person according to existing multichannel amplification and compression algorithm.The present invention can be widely used in the determining of hearing compensation gain of hearing-aid device.

Accompanying drawing explanation

Fig. 1 is calculation process block diagram of the present invention;

Fig. 2 (a) is the function relation figure of maximum initiatively gain G dBm (CF) of the present invention and frequency CF, specifically as shown in Equation (5), wherein abscissa represents frequency CF, and unit is Hz, ordinate represents maximum initiatively gain G dBm (CF), and unit is dB;

Fig. 2 (b) is the active gain G dB (CF) and passive exciter response E in the present invention _pFthe function relation figure of dB (CF), concrete as formula (4), shown in formula (5) and formula (6), wherein abscissa is passive exciter response E _pFdB (CF), unit is dB; Ordinate is active gain G dB (CF), unit is dB, the active gain of the corresponding normal ear of uppermost 4 heavy lines in figure, called after GdBN (CF), wherein alphabetical N represents normal ear, the active gain of the corresponding acoustic trauma ear of fine line below, called after GdBI (CF), wherein alphabetical I represents to damage ear, the corresponding initiatively minimum value of gain of lowermost dotted line, for-13dB;

Fig. 3 is the audiogram of four kinds of common hearing losses in the specific embodiment of the invention,

the expression threshold of audibility,

represent the damage of ear external hair cell, represent the damage of ear inner hair cell, wherein Fig. 3 (a) is the hearing loss of moderate flat type MF; Fig. 3 (b) is the slow hearing loss that rises type MG of moderate; Fig. 3 (c) is the skyrocket hearing loss of type MS of moderate; Fig. 3 (d) is severe flat type SF hearing loss;

Fig. 4 is that in Fig. 3, four kinds of common hearing losses adopt hearing aid of the present invention to test to join the compensating gain of formula and test the contrast schematic diagram of joining formula with existing NAL-NL2, abscissa is frequency f, unit is Hz, ordinate is compensating gain IGdB(f), unit is dB, wherein hearing aid of the present invention is tested compensating gain employing "-" expression of joining formula, existing NAL-NL2 tests compensating gain employing "---" expression of joining formula, Fig. 4 (a) is that MF type hearing loss is when being input as long-time speech frequency spectrum, adopt hearing aid of the present invention to test to join formula and NAL-NL2 to test and join the compensating gain contrast schematic diagram that formula gives, wherein go up, in, lower three curves respectively input sound intensity of correspondence are 50dB SPL, situation when 65dB SPL and 80dB SPL, Fig. 4 (b) is that MG type hearing loss hearing aid of the present invention when being input as long-time speech frequency spectrum is tested and joined formula and NAL-NL2 and test and join the compensating gain contrast schematic diagram that formula gives, Fig. 4 (c) be MS type hearing loss when being input as long-time speech frequency spectrum, adopt hearing aid of the present invention to test to join formula and NAL-NL2 to test and join the compensating gain contrast schematic diagram that formula gives, Fig. 4 (d) be SF type hearing loss when being input as long-time speech frequency spectrum, adopt hearing aid of the present invention to test to join formula and NAL-NL2 to test and join the compensating gain contrast schematic diagram that formula gives,

Fig. 5 is specific embodiment of the present invention, in Fig. 5, (a)～(c) middle abscissa is time t, and unit is s, and ordinate represents sound amplitude, in Fig. 5, (d)～(f) middle abscissa is frequency f, unit is Hz, and ordinate represents sound spectrum, and unit is dB SPL, the abscissa of (g) in Fig. 5～(h) is frequency f, unit is Hz, and ordinate represents compensating gain IGdB (f), and unit is dB; Fig. 5 (a) is that one section of duration is that 2s, the sound intensity are the time domain waveform schematic diagram of not testing the steady-state noise of joining formula manipulation through hearing aid of 65dB SPL, and the frequency spectrum of this noise is identical with long-time average speech frequency spectrum; Fig. 5 (b) is that hearing aid of the present invention is tested and joined formula to the recompense time domain waveform schematic diagram of the steady-state noise that obtains after gain of MF type hearing loss; Fig. 5 (c) is that hearing aid of the present invention is tested and joined formula to the recompense time domain waveform schematic diagram of the steady-state noise that obtains after gain of MS type hearing loss; Fig. 5 (d) is the spectrum diagram of not testing the steady-state noise of joining formula manipulation through hearing aid; Fig. 5 (e) is that hearing aid of the present invention is tested and joined formula to the recompense spectrum diagram of the steady-state noise that obtains after gain of MF type hearing loss; Fig. 5 (f) is that hearing aid of the present invention is tested and joined formula to the recompense spectrum diagram of the steady-state noise that obtains after gain of MS type hearing loss; Black line in Fig. 5 (g) is the differential intention of the frequency spectrum in Fig. 5 (e) and Fig. 5 (d), this difference is that algorithm is realized the compensating gain to MF, ash white line corresponding diagram 4(a) in middle solid line, this ash white line is inputted when sound intensity is 65dB SPL and is tested and join the compensating gain that formula provides MF; Black line in Fig. 5 (h) is the differential intention of the frequency spectrum in Fig. 5 (f) and Fig. 5 (d), this difference is that algorithm is realized the compensating gain to MS, ash white line corresponding diagram 4(c) in middle solid line, this ash white line is inputted when sound intensity is 65dB SPL and is tested and join the compensating gain that formula provides MS;

Fig. 6 is the embodiment of specific implementation of the present invention, in Fig. 6, (a)～(c) middle abscissa is time t, and unit is s, and ordinate is sound amplitude, in Fig. 6, (d)～(f) middle abscissa is frequency f, unit is Hz, and ordinate is sound spectrum, and unit is dB SPL, the abscissa of (g) in Fig. 6～(h) is to represent frequency f, unit is Hz, and ordinate represents compensating gain IGdB (f), and unit is dB; Fig. 6 (a) be one section of duration be 2s, the sound intensity be 65dB SPL not through testing the time domain waveform schematic diagram of the real speech of joining formula manipulation; Fig. 6 (b) is that hearing aid of the present invention is tested and joined formula to the recompense time domain waveform schematic diagram of the voice that obtain after gain of MF type hearing loss; Fig. 6 (c) is that hearing aid of the present invention is tested and joined formula to the recompense time domain waveform schematic diagram of the voice that obtain after gain of MS type hearing loss; Fig. 6 (d) is the spectrum diagram of not testing the voice of joining formula manipulation through hearing aid; Fig. 6 (e) is that hearing aid of the present invention is tested and joined formula to the recompense spectrum diagram of the voice that obtain after gain of MF type hearing loss; Fig. 6 (f) is that hearing aid of the present invention is tested and joined formula to the recompense spectrum diagram of the voice that obtain after gain of MS type hearing loss; Black line in Fig. 6 (g) is the differential intention of the frequency spectrum in Fig. 6 (e) and Fig. 6 (d), this difference is that algorithm is realized the compensating gain to MF, and grey white line is according to hearing aid of the present invention, to test and join the compensating gain that formula calculates MF type hearing loss as inputting take the frequency spectrum of Fig. 6 (d); Black line in Fig. 6 (h) is the poor of frequency spectrum in Fig. 6 (f) and Fig. 6 (d), this difference is that algorithm is realized the compensating gain to MF, and grey white line is according to hearing aid of the present invention, to test and join the compensating gain that formula calculates MS type hearing loss as inputting take the frequency spectrum of Fig. 6 (d).

Embodiment

Below in conjunction with drawings and Examples, the present invention is described in detail.

As shown in Figure 1, the method that the present invention is based on definite hearing-aid device hearing compensation gain of ear external hair cell and inner hair cell damage comprises the following steps:

1, testee's the sense of hearing is tested, is determined hearing loss person is at which frequency CF(Center Frequency) locate to exist hearing loss, determine the frequency CF that need to carry out gain compensation.

For hearing loss person, when wearing hearing aids, medically conventionally need to detect its ear threshold of audibility, the method detecting is generally to choose some conventional Frequency point testee's ear is carried out to sense of hearing diagnosis, conventional Frequency point is 125Hz, 250Hz, 500Hz, 1000Hz, 2000Hz, 4000Hz, 6000Hz and 8000Hz, by diagnosing the intensity of sound of determining that each frequency testee of place can receive, obtain the audiogram of testee's ear, the audiogram of the audiogram of testee's ear and normal ear is compared, determine that testee exists loss in the hearing of which CF of frequency place, at this frequency CF place, its hearing loss is carried out to suitable compensation.

The loudness of the normal hearing 2, obtaining based on current research and the damage sense of hearing is experienced model, for the impairment value HLohcdB (CF) of the ear external hair cell at acoustic trauma frequency CF place, it is poor that the sense of hearing of calculating hearing normal ear and hearing impairment ear gains in the active at this frequency CF place, obtains the compensating gain value IGohcdB (CF) of ear external hair cell damage.

As shown in Figure 1, the sound spectrum of supposing to be input under free found field condition testee's ear is X (f), sound spectrum X (f) is successively after external ear and middle ear filtering OME (f), and the sound spectrum that arrives ear cochlea is Y (f):

Y(f)＝X(f)OME(f) （1）

In formula, OME (f) is the english abbreviation of external ear and middle ear frequency response.

The sound spectrum Y (f) at cochlea place is through the passive broadband filter W of the sense of hearing _pFafter (CF, f), the passive exciter response of output is E _pF(CF) be:

$E_{\mathrm{PF}}\left(\mathrm{CF}\right)=\underset{f}{\mathrm{\Σ}}{W}_{\mathrm{PF}}(\mathrm{CF},f)Y\left(f\right)---\left(2\right)$

In formula, W _pF(CF, f) is that frequency is the frequency response of the sense of hearing Passive filter device at CF place, W _pF(CF, f) is:

$\left\{\begin{array}{c}{W}_{\mathrm{PF}}(\mathrm{CF},f\≤\mathrm{CF})=(1+\frac{\mathrm{CF}-f}{\mathrm{CF}}{t}_L\left(\mathrm{CF}\right))\exp (-\frac{\mathrm{CF}-f}{\mathrm{CF}}{t}_L\left(\mathrm{CF}\right))\\ W_{\mathrm{PF}}(\mathrm{CF},f>\mathrm{CF})=(1+\frac{f-\mathrm{CF}}{\mathrm{CF}}{t}_U\left(\mathrm{CF}\right))\exp (-\frac{f-\mathrm{CF}}{\mathrm{CF}}{t}_U\left(\mathrm{CF}\right))\\ t_L\left(\mathrm{CF}\right)=\frac{\mathrm{CF}}{0.108\mathrm{CF}+2.33}\\ t_U\left(\mathrm{CF}\right)=15.6\end{array}\right.---\left(3\right)$

In formula, t _land t (CF) _u(CF) be that control frequency is the coefficient of the passive auditory filter shape at CF place.

According to E _pF(CF) the active gain G dB (CF) that can calculated rate be CF place hearing impairment ear is:

$\left\{\begin{array}{c}\mathrm{GdB}\left(\mathrm{CF}\right)=\mathrm{GdBm}\left(\mathrm{CF}\right)\left\{\begin{array}{c}1-\frac{1}{1+\exp [-0.005(E_{\mathrm{PF}}\mathrm{dB}\left(\mathrm{CF}\right)-(100-\mathrm{GdBm}\left(\mathrm{CF}\right)))]}\\ +\frac{1}{1+\exp \left[0.05(100-\mathrm{GdBm}\left(\mathrm{CF}\right))\right]}\end{array}\right\}\\ \mathrm{if}{E}_{\mathrm{PF}}\mathrm{dB}\left(\mathrm{CF}\right)>30,\mathrm{thenGdB}\left(\mathrm{CF}\right)=\mathrm{GdB}\left(\mathrm{CF}\right)-0.003{(E_{\mathrm{PF}}\mathrm{dB}\left(\mathrm{CF}\right)-30)}^2\end{array}\right.---\left(4\right)$

In formula, E _pFdB (CF)=10lg (E _pF(CF)), GdBm (CF) is that frequency is the maximum initiatively yield value of CF people from place ear, works as E _pFdB (CF)=0 o'clock, GdB (CF)=GdBm (CF).The formula (4) is here applicable to the calculating of the active gain of normal ear and damage ear simultaneously.For normal ear, replace respectively GdBm (CF) and the GdB (CF) in formula (4) with GdBmN (CF) and GdBN (CF); For damage ear, replace respectively GdBm (CF) and the GdB (CF) in formula (4) with GdBmI (CF) and GdBI (CF).

Normal ear at the maximum initiatively gain G dBmN (CF) at frequency CF place is:

$\mathrm{GdBmN}\left(\mathrm{CF}\right)=\frac{\mathrm{CF}}{0.019\mathrm{CF}+1.10}---\left(5\right)$

Can be that 125Hz, 250Hz, 500Hz, 1000Hz, 2000Hz, 4000Hz and 8000Hz calculate respectively the maximum initiatively yield value GdBmN (CF) of normal ear at different frequency place in substitution formula (5) respectively by frequency CF, result of calculation as shown in Figure 2 (a) shows, circle in figure represents the maximum initiatively yield value that frequency CF calculates according to formula (5) while being 125Hz, 250Hz, 500Hz, 1000Hz, 2000Hz, 4000Hz and 8000Hz, and the solid line in figure is the curve that each frequency is coupled together.

When the ear external hair cell impairment value at a certain frequency CF place is HLohcdB (CF), the maximum initiatively gain G dBmI (CF) of damage ear is:

GdBmI(CF)＝GdBmN(CF)-HLohcdB(CF) （6）

As shown in Fig. 2 (b), when by GdBmI (CF) in the GdBmN (CF) in formula (5) and formula (6), the GdBm (CF) in replacement formula (4) can calculate input spectrum and is X (f) respectively, the sense of hearing of hearing normal ear and hearing impairment ear is gain G dBN (CF) and GdBI (CF) initiatively, therefore obtains hearing aid and tests and join the compensating gain IGohcdB (CF) that formula damages for external hair cell:

IGohcdB(CF)＝m[GdBN(CF)-GdBI(CF)] （7）

In formula, m is coefficient, and initiatively gain G dBI (CF) is linear with the sense of hearing of damage ear for m, and the slope of this linear relationship and intercept change along with the variation of normal ear sense of hearing active gain G dBN (CF), and the computing formula of m is:

$m=\left[\frac{0.5}{\mathrm{GdBN}\left(\mathrm{CF}\right)-(-13)}\right]\mathrm{GdBI}\left(\mathrm{CF}\right)+[1-\frac{0.5\·\mathrm{GdBN}\left(\mathrm{Cf}\right)}{\mathrm{GdBN}\left(\mathrm{CF}\right)-(-13)}]---\left(8\right)$

In formula ,-13dB is normal ear and the initiatively minimum value (as shown in the dotted line in (b) figure in Fig. 2) of gain of the damage ear sense of hearing of setting.

3,, for the ear inner hair cell impairment value HLihcdB (CF) at acoustic trauma frequency CF place, the compensating gain IGihcdB (CF) that calculates the damage of ear inner hair cell is:

IGihcdB(CF)＝0.5·max[HLihcdB(CF),40] （9）

In formula, max represents to get the maximum of HLihcdB (CF) and 40 in the two, because when HLihcdB (CF) is more than or equal to 40dB, ear inner hair cell may damage completely, does not therefore need to carry out larger compensation.After the damage of ear inner hair cell, its part sound perception function can not be restored because sound amplifies, there is no need to allow compensating gain equal degree of injury, therefore rule of thumb in the present invention, in formula (9), be multiplied by the coefficient of 0.5, owing to there is no at present laboratory facilities, can verify the optimum value of this coefficient, therefore this coefficient may be for further adjustments according to concrete degree of impairment, such as being adjusted into 0.4 or 0.6.

4, when hearing impairment frequency CF is during lower than 1000Hz, the compensating gain of frequency CF will carry out corresponding decay according to frequency CF, to reduce up masking effect, (masking effect is psychoacoustic classical effect, be the sound that the sound of low frequency is easy to cover high frequency, from low frequency to high frequency, be called " up "), each frequency CF place pad value is INTdB (CF), as shown in table 1:

Table 1

Frequency CF, Hz	125	250	500	1000
					Decay INTdB (CF), dB	-15	-10	-5	0

Decay at other frequency place (as 200Hz) can obtain by interpolation calculation according to above-mentioned table 1, and when hearing impairment frequency CF is more than or equal to 1000Hz, frequency CF place pad value INTdB (CF) is 0;

5, comprehensive calculation procedure above, the compensating gain value IGdB (CF) that to obtain hearing loss person be CF place in frequency is:

IGdB(CF)＝IGohcdB(CF)+IGihcdB(CF)+INTdB(CF) （10）

The value that above-mentioned compensating gain IGdB (CF) is frequency CF place, and input spectrum X (f) is the value at frequency f place, the frequency CF that the two is corresponding and f may be different, therefore need that IGdB (CF) is carried out to interpolation and obtain the value IGdB (f) at frequency f place, then be transformed into linear amplitude: IG (f)=10 ^{(IGdB (f)/10)}, just can obtain thus damaging ear and obtain the input X after compensating gain _iG(f), as shown in formula below:

X _IG(f)=X(f)·IG(f) （11）

Hearing aid of the present invention is tested the algorithm implementation procedure of joining formula and is comprised the following steps:

1) by Short Time Fourier Transform, divide frame to calculate the time-domain signal that is input to testee's ear, obtain the corresponding input spectrum X of each frame time-domain signal (f), because the signal being input in practical application in testee's ear is time-domain signal, and hearing aid of the present invention to test the input signal of joining formula be spectrum signal, while therefore using, need to carry out the conversion between time-domain signal and frequency-region signal.

2) according to formula (10), calculate the compensating gain IG (f) that each frame frequency spectrum X (f) should obtain.

3) by compensating gain IG (f), obtain the frequency spectrum X of the input that damages ear compensating gain _iG(f), and the compensating gain frequency spectrum obtaining is carried out to inverse Fourier transform in conjunction with the phase-frequency response before treatment of this frame, obtain time-domain signal.

4) all time-domain signals are carried out to overlap-add procedure after and be input to damage ear in, its hearing impairment is compensated.

Below by specific embodiment, further verify that the present invention determines the correctness of hearing-aid device hearing compensation gain, concrete proof procedure is as follows:

As shown in Fig. 3 (a)～Fig. 3 (d), specific embodiments of the invention have been enumerated four kinds of typical hearing loss situations, and each hearing loss comprises respectively threshold of audibility HLdB (CF) (circle solid line), external hair cell damage HLohcdB (CF) (square dotted line) and inner hair cell damage HLihcdB (CF) (triangle dotted line), as shown in Fig. 4 (a)～Fig. 4 (d), the present invention has provided the SPL at 50dB, when the long-time average speech frequency spectrum of 65dB SPL and the 80dB SPL sound intensity is inputted, through hearing aid of the present invention, test the gain compensation of joining formula and calculating (as shown in the solid line in figure, corresponding 50dB SPL successively from top to bottom, situation when 65dB SPL and the 80dB SPL input sound intensity), the compensating gain calculating according to Australia's national acoustical laboratory NAL-NL2 formula (as shown in phantom in FIG., corresponding 50dB SPL successively from top to bottom, situation when 65dB SPL and the 80dB SPL input sound intensity), by relatively finding out through hearing aid of the present invention, test and join the result that result that formula calculates calculates to NAL-NL2 formula and have similar trend, due to the extensive use clinically of compensating gain of NAL-NL2 formula and proved validity, therefore explanation determines that the gain of hearing-aid device hearing compensation is appropriate.

As shown in Fig. 5 (a)～Fig. 5 (h), provided while thering is noise in time domain (sound intensity is 65dB SPL) the conduct input of same frequency spectrum with long-time average speech frequency spectrum, for two in Fig. 3 typical hearing loss situations (MF and MS), the result that adopts hearing aid of the present invention to test to join formula and algorithm thereof to realize.What Fig. 5 (a) showed is the noise time domain waveform of input normal human ear; Fig. 5 (d) is the long-time frequency spectrum (whole noise time domain waveform is carried out to Fourier transform, be different from each frame frequency spectrum that Short Time Fourier Transform obtains, but because this noise is steady-state noise, each frame frequency spectrum should be identical with long-time frequency spectrum) of this noise.What Fig. 5 (b) showed is for this hearing loss situation of MF, the noise time domain waveform of the input damage people ear after algorithm process of the present invention; Fig. 5 (e) is the long-time frequency spectrum of this noise; Fig. 5 (c) is for this hearing loss situation of MS, the noise time domain waveform of the input damage people ear after algorithm process of the present invention; Fig. 5 (f) is the long-time frequency spectrum of this noise.Black line in Fig. 5 (g) is the poor of Fig. 5 (e) frequency spectrum and Fig. 5 (d) frequency spectrum, and grey white line has provided and directly utilized Fig. 5 (d) input spectrum and MF hearing loss situation to join by of the present invention testing the compensating gain that formula calculates; Black line in Fig. 5 (h) is the poor of Fig. 5 (f) frequency spectrum and Fig. 5 (d) frequency spectrum, and grey white line has provided and directly utilized Fig. 5 (d) input spectrum and MS hearing loss situation to join by of the present invention testing the compensating gain that formula calculates.Can find out, the grey white line in Fig. 5 (g) and Fig. 5 (h) and the major part of black line coincide together, and have proved thus the correctness of algorithm.

As shown in Fig. 6 (a)～Fig. 6 (h), when specific embodiments of the invention have provided one section of actual speech (sound intensity is 65dB SPL) as input, for with hearing loss situation (MF and MS) identical in Fig. 5, through hearing aid of the present invention, test and join the result that formula and algorithm thereof are realized.Be different from the steady-state noise in Fig. 5, actual speech is time varying signal, and frequency spectrum is along with the time changes and changes.By contrast can find out, utilize Short Time Fourier Transform to carry out algorithm realization, the result obtaining is consistent with the compensation result that utilizes long-time Fourier transform to calculate, proving again the correctness of algorithm of the present invention.

The various embodiments described above are only for illustrating the present invention, and wherein the implementation step of the inventive method can change to some extent, and every equivalents of carrying out on the basis of technical solution of the present invention and improvement, all should not get rid of outside protection scope of the present invention.

Claims (3)

1. a method for definite hearing-aid device hearing compensation gain, comprises the following steps:

1) testee's the sense of hearing is tested, determined the frequency CF that need to carry out gain compensation place;

2) loudness based on normal hearing and the damage sense of hearing is experienced model, for the impairment value HLohcdB (CF) of the ear external hair cell at acoustic trauma frequency CF place, calculates the compensating gain value IGohcdB (CF) that ear external hair cell damages:

IGohcdB(CF)＝m[GdBN(CF)-GdBI(CF)]

In formula, m is coefficient, and the computing formula of m is:

$m=\left[\frac{0.5}{\mathrm{GdBN}\left(\mathrm{CF}\right)-(-13)}\right]\mathrm{GdBI}\left(\mathrm{CF}\right)+[1-\frac{0.5\·\mathrm{GdBN}\left(\mathrm{Cf}\right)}{\mathrm{GdBN}\left(\mathrm{CF}\right)-(-13)}]$

In formula ,-13dB is the initiatively minimum value of gain of the sense of hearing of setting, the active gain that GdBI (CF) is damage ear, the active gain that GdBN (CF) is normal ear;

3) for the ear inner hair cell impairment value HLihcdB (CF) at acoustic trauma frequency CF place, calculate the compensating gain value IGihcdB (CF) of ear inner hair cell damage;

4) when hearing impairment frequency CF is during lower than 1000Hz, the compensating gain at frequency CF place will carry out corresponding decay according to frequency CF, and pad value is INTdB (CF); When hearing impairment frequency CF is more than or equal to 1000Hz, frequency CF place pad value is 0;

5) comprehensive step 2)～4), obtain hearing loss person and at the compensating gain IGdB at frequency CF place (CF) be:

IGdB(CF)＝IGohcdB(CF)+IGihcdB(CF)+INTdB(CF)。

2. the method for a kind of definite hearing-aid device hearing compensation gain as claimed in claim 1, is characterized in that: the compensating gain value IGihcdB (CF) of step 3) middle ear inner hair cell damage:

IGihcdB(CF)＝0.5·max[HLihcdB(CF),40]

In formula, max represents to get the maximum of HLihcdB (CF) and 40 in the two.

3. the method for a kind of definite hearing-aid device hearing compensation gain as claimed in claim 1 or 2, it is characterized in that: in step 4), the pad value of the compensating gain at different frequency CF place is respectively: frequency CF is that the pad value INTdB (CF) that 125HZ is corresponding is-15dB, frequency CF is that the pad value INTdB (CF) that 250HZ is corresponding is-10dB, frequency CF is that the pad value INTdB (CF) that 500HZ is corresponding is-5dB, frequency CF is that the pad value INTdB (CF) that 1000HZ is corresponding is 0dB, other pad value INTdB (CF) lower than other frequency CF at 1000HZ frequency place passes through interpolation calculation.

Images