EP0054450A1 - Hearing aid devices - Google Patents

Abstract

1. Hearing aid device comprising means for taking bands of acoustic signals (C,D,R,S,T) contained respectively in various bands of useful frequencies which are in the range inaudible to the deaf, and for transposing them into acoustic signals the amplitude of which is sufficient and the frequency is in the audible range for deaf persons, characterised in that the means are arranged so as to give to the latter signals the form of image bands (c,d,r,s,t) of the above mentioned bands (C,D,R,S,T), constituting respectively images of the said bands faithfully amplified in amplitude, but condensed and reduced in frequency, the modulation of amplitude of each image band being similar to that of the corresponding band, but the frequency at each point of this image band being equal to a fraction 1/n of the frequency at the corresponding point of the said band and the rate of reduction n being a number greater than 1, preferably a whole number, which is the same for all the points of a given band, in that the different image bands (c,d,r,s,t) are arranged together stepwise in the spectrum audible to the deaf person, which arrangement corresponds to different rates of reduction n for different image bands and also in that the spacing period adopted for the taking of the amplitude envelope of each band, in order to modulate the amplitude of the corresponding image band, is between 0.8 and 1.2 ms.

Description

The invention is intended for hearing aid devices, devices intended to remedy deafness by enabling deaf people to understand human speech, and it relates more particularly, but not exclusively, among these devices, those intended to remedy insufficiency hearing impairment of severely affected deaf or "deep deaf" deaf people by making speech intelligible to them.

It is recalled that the deep deaf perceive only the sounds whose amplitude is sufficiently high and whose frequency is relatively low and generally located in the range of frequencies between 20 and 1000 Hz, range which will be designated hereinafter by the expression "spectrum audible by the deaf".

To provide these deaf people with hearing information relating to the rest of the range of frequencies corresponding to human speech, that is to say from 1000 to 7000 or 8000 Hz, it has been proposed to transform into low-pitched sounds, audible by these deaf, the high sounds corresponding to said rest of the range, which remain will be designated in the following by the expression "spectrum inaudible to deaf".

To achieve this goal, the applicant himself proposed an interesting solution in his French patent No. 1,382,916.

This solution is based on the discovery that it is not essential, for the understanding of human speech, to know all the information contained in the spectrum inaudible to deaf people: it suffices for this purpose to take the acoustic information located in certain bands of this spectrum do not cover in total only part of it, bands hereinafter called "useful bands".

In his patent above, the applicant has proposed for this purpose to transpose the portions of the received sounds, the frequencies of which are included in two useful bands distinct from the spectrum inaudible to deaf people by replacing them respectively by two low frequency single modulated sounds respectively by the envelope curves of the amplitudes of the two said portions, each of these two serious transposition sounds possibly being a pure sinusoidal sound.

Prosthesis devices based on this principle give great satisfaction and allow to solve, to a certain extent, the problem of deep deafness.

However, the applicant has found that it is possible to further improve the devices based on the above principle.

He noticed in fact that the transformation of the acoustic signal contained in a useful frequency band into a single low sound, even if this sound is modulated in amplitude according to the envelope of the amplitudes of this band, caused certain losses of information which can lead to difficulties of interpretation of the global sound message received by the deaf person.

Thus, for example, if one of the two useful bands chosen is that between 1500 to 3500 Hz, the above formula amounts to transforming the acoustic components having frequencies 2000 and 3000 Hz respectively into strictly identical information. , to know in two identical fractions and superimposed of the same grave sound of transposition.

The main object of the present invention is to remedy this loss of information.

To this end, the hearing aid devices of the above type according to the invention are characterized in that they comprise means for transposing the different sections, of the overall acoustic signal received at each instant, contained respectively in the different useful bands, in as many slices of images situated in the spectrum audible by the deaf and constituting respectively faithful images amplified in amplitude, but compressed and reduced in frequency, of said slices, the amplitude modulation of each image being similar to that of the slice corresponding, but the frequency at each point of this image being equal to a fraction 1 / n of the frequency at the corresponding point of said section and the reduction rate n being a number greater than 1, preferably an integer, which is the same for all the points of a given slice and in that the time constant adopted for the sampling of the amplitude envelope, of each slice, intended to modulate the amplitude of the corresponding image slice is between 0.8 and 1.2 ms.

• - les différentes tranches-images s'échelonnent jointivement dans le spectre audible par le sourd, ce qui correspond à des taux de réduction n différents pour ces différentes tranches-images,
• - les bandes utiles de fréquence sont au nombre de deux, comprises respectivement entre environ 1500 Hz et environ 240Q Hz et entre environ 4800 Hz et environ 6000 Hz, et les taux de réduction appliqués à ces deux bandes sont respectivement de 3 et de 6,
• - les bandes utiles de fréquence sont au nombre de trois, comprises respectivement entre environ 1200 Hz et environ 2000 Hz, entre environ 3000 Hz et environ 4500 Hz et entre environ 6000 Hz et environ 8000 Hz, et les taux de réduction appliqués à ces trois bandes sont respectivement de 2, 3 et 4,
• - le dispositif comprend des moyens pour amplifier d'un taux relativement élevé toutes les portions acoustiques, des sons normalement reçus, dont la fréquence est inférieure à environ 500 Hz, et pour réduire ce taux d'amplification progressivement jusqu'à 1000 Hz et au-delà, la constante de temps adoptée pour l'amplification étant comprise entre 0,8 et 1,2 ms.

In the preferred embodiments, use is also made of one and / or the other of the following arrangements:

• - the different image slices are staggered in the spectrum audible by the deaf, which corresponds to different reduction rates n for these different image slices,
• - the useful frequency bands are at number of two, respectively between approximately 1500 Hz and approximately 240Q Hz and between approximately 4800 Hz and approximately 6000 Hz, and the reduction rates applied to these two bands are respectively 3 and 6,
• - there are three useful frequency bands, respectively between approximately 1200 Hz and approximately 2000 Hz, between approximately 3000 Hz and approximately 4500 Hz and between approximately 6000 Hz and approximately 8000 Hz, and the reduction rates applied to these three bands are respectively 2, 3 and 4,
• - The device comprises means for amplifying at a relatively high rate all the acoustic portions, normally received sounds, the frequency of which is less than about 500 Hz, and for reducing this amplification rate gradually up to 1000 Hz and at beyond, the time constant adopted for the amplification being between 0.8 and 1.2 ms.

The invention includes, apart from these main provisions, certain other provisions which are preferably used at the same time and which will be more explicitly discussed below.

In what follows, two preferred embodiments of the invention will be described with reference to the accompanying drawing, of course in a non-limiting manner.

Figure 1, of this drawing, is a gra phique to explain an application of the invention to a hearing aid for the deaf.

FIG. 2 is a graph relating to a variant application of the invention concerning the treatment of less deep deafness than the previous one.

On each of these graphs, the frequencies f of the received sounds to be processed are plotted on the abscissa and their amplitudes on the ordinate.

• - des moyens pour recevoir intégralement les portions, convenablement amplifiées, des signaux acoustiques, comprises dans la bande des fréquences inférieures à 500-Hz, portions audibles par les sourds, même profonds, ainsi qu'il a été schématisé par le tronçon de courbe A,
• - et des moyens pour recevoir avec un taux d'amplification progressivement décroissant vers les fréquences élevées les portions, des signaux acoustiques, comprises dans la bande des fréquences allant de 500 à 1000 Hz et au-delà, comme schématisé par le tronçon de courbe B.

In each case, the prosthesis device concerned is made to understand:

• means for integrally receiving the portions, suitably amplified, of the acoustic signals, included in the frequency band below 500-Hz, portions audible by the deaf, even deep, as has been shown diagrammatically by the section of curve A ,
• - And means for receiving, with an amplification rate progressively decreasing towards the high frequencies, the portions, acoustic signals, included in the frequency band going from 500 to 1000 Hz and beyond, as shown diagrammatically by the section of curve B .

This partial erasure of the region of the audible spectrum between 500 and 1000 Hz makes it possible to partially free this region to inject therein the artificial signals carrying useful information which will be discussed below.

In the case of FIG. 1, these signals are of two types, developed respectively from the information sound mations contained respectively in two useful bands C and D of the spectrum inaudible to the deaf.

These two bands are here those comprised respectively between 1500 and 2400 Hz (band C) and between 4800 and 6000 Hz (band D).

From the slices of acoustic signals C ₁ and D ₁ contained respectively in these two bands C and D, two "image slices" c ₁ and d _{1 are produced,} respectively contained in the band c between 500 and 800 Hz and in the band d between 800 and 1000 Hz.

The first image slice c ₁ corresponds to a faithful image of the slice CI but amplified in amplitude and "compressed in frequency" with a reduction rate of 3.

More precisely, this image c _l has an amplitude modulation similar to that of the slice C ₁ that it reproduces, but at each of the points m of this image, the amplitude is higher and the frequency f is equal to a third. from frequency F to the corresponding point M of said section C ₁ .

In short, the image c ₁ corresponds to the slice C ₁ amplified in amplitude and divided by three in frequency.

It results from the choice of an integer (here 3) adopted for the division rate, that the acoustic transposition carried out corresponds to a change of octave preserving the "tonal" quality of the transposed acoustic information: the "notes" remain the same, but are changed octave.

As for the amplitude modulation of this image c ₁ , it is carried out from the envelope of the amplitude of the signal slice C _{1 and} this envelope is taken with a time constant of less than 1 , 2 ms: in other words, the vibrational quality of the por tions of original acoustic signals, the periodicity of which is less than 1.2 ms, that is to say the laryngeal vibrations of these portions, which are characteristic of the height of the human voice.

This observation is of great interest for the intellectual reconstruction of the human voice by the deaf person equipped with the prosthesis considered.

More generally, the time constant in question is given a value between 0.8 and 1.2 ms.

The constant time is similarly chosen for the various amplifications, for the same reason as above.

Exactly under the same conditions as for the image c ₁ , the image d ₁ is formed from the signal slice D ₁ by means of an amplification of amplitude and a frequency division of the latter, but here by the factor 6 .

• - les deux images c₁ et d₁ s'étendent chacune sur une bande de fréquences relativement large au lieu d'être constituées par des sons graves de fréquence unique : il en résulte un transfert d'information beaucoup plus riche et complet que dans l'hypothèse de ces sons graves uniques,
• - les bandes de fréquences c et d dans lesquelles sont contenues respectivement les deux images c₁ et d₁ se succèdent jointivement dans la gamme des fréquences, de sorte qu'elles occupent ensemble la totalité de la zone comprise entre 500 et 1000 Hz qui a été partiellement libérée de la manière prévue plus haut pour recevoir les sons transposés, et ce sans se gêner mutuellement par un chevauchement partiel desdites bandes.

That said, we can make the following two observations:

• - the two images c ₁ and d ₁ each extend over a relatively wide frequency band instead of being constituted by low frequency sounds of single frequency: this results in a much richer and more complete transfer of information than in l hypothesis of these unique bass sounds,
• the frequency bands c and d in which the two images c ₁ and d ₁ are respectively contained successively in the frequency range, so that they occupy together the whole of the zone between 500 and 1000 Hz which has been partially released in the manner provided above to receive the transposed sounds, and this without interfering with each other by a partial overlapping of said bands.

The amplified sound message thus reconstructed is reveals to be very rich in information and particularly suitable for understanding human speech.

The graph of FIG. 1 - just like that of FIG. 2 - is only a diagram making it possible to explain the frequency divisions carried out according to the invention: the bands c and d are not hatched on this diagram only to distinguish them from other bands; it is also for the sake of distinction that they have been given a higher amplitude than that of the other bands. In addition, strictly speaking, the amplitudes of the image signals c ₁ and d should not be counted from the x-axis, but from the amplitudes of the signals (not shown) which are directly received in the corresponding range. between 500 and 1000 Hz, the two amplitudes considered overlapping.

One can proceed in any desirable way to ensure the transposition of each signal slice C ₁ or D ₁ into its reduced frequency image c ₁ or d _l .

This is how we can consider in each of the slices C ₁ and D ₁ a certain number of successive and contiguous frequency sub-bands selected using appropriate filters, then transform the signal fractions contained in each of the sub -bands thus selected into an alternating electrical signal, after which the frequency of this signal is divided by the reduction rate n desired for the section considered by removing, in said alternating signal, n-1 alternations out of n, and finally amplifying the reduced frequency signal thus obtained so as not only to compensate for the loss of energy due to the above suppression, but also to reach the zone audible by the deaf.

According to a variant, each fraction of a slice corresponding to one of the defined sub-bands is transformed above in a coded train of pulses the number of which is linked to the frequency of said fraction, then a certain number of these pulses are suppressed (namely n-1 over n, if one wishes to ensure a division by the factor n) and finally an amplification is carried out as in the previous case of a signal whose reduced frequency is linked to the number of remaining pulses.

It is also possible to use, for each frequency division, a logic circuit, a microprocessor or any other desirable device.

The embodiment shown schematically in FIG. 2 is based on the same principles as the previous schematized in FIG. 1.

It applies to the treatment of deaf people who are less seriously affected than in the previous case, who are able to hear sounds up to a frequency of at least 2000 Hz.

This second embodiment differs from the previous one in that the number of useful bands of the spectrum inaudible to the deaf is here equal to 3 and no longer to 2 and in that the reduced-frequency images reconstructed from the signal slices included in these three useful bands occupy here the frequency band between 600 and 2000 Hz.

• - la première, R, d'environ 1200 Hz à environ 2000 Hz,
• - la seconde, S, d'environ 3000 Hz à environ 4500 Hz,
• - la troisième,T, d'environ 6000 Hz à environ 8000 Hz.

More specifically, the three bands in question extend respectively:

• - the first, R, from around 1200 Hz to around 2000 Hz,
• - the second, S, from around 3000 Hz to around 4500 Hz,
• - the third, T, from around 6000 Hz to around 8000 Hz.

• - la première, r, la bande de fréquence comprise entre environ 600 et environ 1000 Hz,
• - la seconde, s, la bande de fréquence comprise entre environ 1000 et environ 1500 Hz,
• - et la troisième, t, la bande de fréquence comprise entre environ 1500 et environ 2000 Hz.

The three corresponding images r, s and -t are obtained from the slices of acoustic signals included in these three bands R, S and T by amplitude amplifications and frequency divisions whose dividers are respectively equal to 2, 3 and 4, these image slices therefore occupying respectively:

• - the first, r, the frequency band between approximately 600 and approximately 1000 Hz,
• - the second, s, the frequency band between approximately 1000 and approximately 1500 Hz,
• - and the third, t, the frequency band between approximately 1500 and approximately 2000 Hz.

As in the first variant, the three image bands r, s and t corresponding respectively to the three image slices follow one another in the frequency band between 600 Hz and 2000 Hz without overlapping.

To avoid any risk of trauma to the ear equipped with the prosthesis device described above during the application to it of the amplified sounds produced by this device, we make it understood to said device means for automatically removing all those, these amplified sounds that have both too much and too high a speed setting amplitude of: _m and used advantageously for this purpose the teachings of patent US patent No. 3,458,669 whose applicant is co-inventor.

As a result of what and whatever the embodiment adopted, there is finally a hearing aid device whose construction, operation and advantages result sufficiently from the above.

As is obvious, and as already follows from the foregoing, the invention is in no way limited to those of its modes of application and embodiments which have been more especially envisaged; on the contrary, it embraces all its variants.

Claims (5)

1. Hearing aid device comprising means for taking slices of acoustic signals (C, D, R, S, T) contained respectively in different useful frequency bands located in the spectrum inaudible to deaf people and for transposing them into acoustic signals whose amplitude is sufficient and whose frequency is contained in the spectrum audible to the deaf, characterized in that these means are arranged so as to give these latter signals the form of image slices (c, d, r, s, t) of the above sections (C, D, R, S, T) respectively constituting faithful images amplified in amplitude, but compressed and reduced in frequency, of said sections, the amplitude modulation of each image being similar to that of the corresponding section, but the frequency at each point of this image being equal to a fraction 1 / n of the frequency at the corresponding point of said section and the reduction rate n being a number greater than 1, preferably an integer, which is the same for all the points of a given slice and in that the time constant adopted for the sampling of the amplitude envelope, of each slice, intended to modulate the amplitude of the corresponding slice-image, is included between 0.8 and 1.2 ms. 2. A hearing aid device according to claim 1, characterized in that the different image slices are staggered in the spectrum audible by the deaf, which corresponds to different reduction rates n for these different image slices. 3. Hearing aid device according to claim 2, characterized in that the useful frequency bands are two in number, respectively between approximately 1500 Hz and approximately 2400 Hz and between approximately 4800 Hz and approximately 6000 Hz, and in that the reduction rates applied to these two bands are respectively 3 and 6 (FIG. 1). 4. Hearing aid device according to claim 2, characterized in that the useful frequency bands are three in number, respectively between approximately 1200 Hz and approximately 2000 Hz, between approximately 3000 Hz and approximately 4500 Hz and between approximately 6000 Hz and around 8000 Hz, and the reduction rates applied to these three bands are respectively 2, 3 and 4 (Figure 2). 5. Hearing aid device according to any one of the preceding claims, characterized in that it comprises means for amplifying, at a relatively high rate, all the acoustic portions, normally received sounds, the frequency of which is less than about 500 Hz, and to reduce this amplification rate gradually up to 1000 Hz and beyond, the time constant adopted for amplification being between 0.8 and 1.2 ms.

Images