Method for simulating auditory patient perception of acoustic signal after cochlear implantation

FIELD: medicine.

SUBSTANCE: method involves applying analog-to-digital input signal transformation expressed as word, dividing transformed signal spectrum into odd and even frequency bands, summing odd bands, carrying out digital-to-analog transformation of resulting summed signal and training its perception by preliminarily getting familiar with the word shown for listening and following testing. Spectrum division is based on tonotopic frequency distribution law over cochlea axis. Frequency bands having odd numbers are arranged in equal distances along basilar membrane length in agreement with normal tonotopic frequency distribution law over cochlea axis. At least three odd spectrum bands are summed up. Training is carried out by multiple repetition of the word shown for listening until unambiguous correlation to the known word meaning given in preliminary acquaintance takes place. The same words are to be shown in testing and training.

EFFECT: partially retained speech spectrum.

The invention relates to medicine, namely to audiology and audiology, and may find application in the "Cochlear implant" to simulate auditory perception in implanted patients and illustrate the process of the development of speech perception implanted patients people with normal hearing.

In the last decade in Russia began the implementation of the method of auditory rehabilitation of deaf patients cochlear implant (CI). For implantation are created in several countries the system KEY, and therefore the main question of the program KEY is to develop ways of auditory rehabilitation of patients after cochlear implantation.

It is known that in the process of aural rehabilitation of children after cochlear implantation significant role of parents. This is due to the necessity of their participation in the rehabilitation of their children, and for their effective operation it is necessary that they should properly understand the problem. Unfortunately, in most cases, their parents cochlear implant is very limited. Moreover, referring to the funds invested in the operation, they cannot understand why their child cannot immediately speak and understand addressed to him the words.

In the case when the configuration of the speech processor of the cochlear implant is carried out at 2-3 children at a time, ro is Italy, of course, you can see the differences between children and do not understand the reason for different achievements in different children. Knowing about the success of some of the previously implanted patients, they cannot understand why they have it is otherwise. A clear physiological explanation of the features of auditory perception in implanted patients and resulting problems requires to start the presentation with the basics of acoustics and physiology that is hardly possible.

How to show parents the problems facing them implanted children while developing their perception of speech processed audio processor KEY? To address this question it is necessary to model the speech signal, which is perceived implanted patients, and to present it to his family and all those who are interested in the problem KEY.

To build a real program of rehabilitation of specific contingent deaf patients, namely patients, who underwent cochlear implantation, it is absolutely necessary a clear picture about what they hear. It is known that their auditory perception is significantly different from the perception of a person with normal hearing and the perception of hearing impaired patients with a hearing aid.

You can give an example.

From literature it is known that people with normal hearing mo is et to determine 600 gradations frequency, in patients with sensorineural hearing loss the number of gradations is 10 times less implanted in the same patient is able to distinguish only a small number of frequency bands, which is limited by the number entered in the cochlea electrodes. At present it does not exceed 20 pages.

Thus, the difference in perception, only on the spectral parameter, huge, and for this reason, the modeling of the speech signal, as close as possible to the one who perceives the implanted patient, it is very important.

There is a method of processing a speech signal, where the model of the sound image perceived by the implanted patient, by creating a limited frequency representation of the speech signal and evaluate the intelligibility of the processed this way of speech in people with normal hearing [1].

The work was used 8-channel cochlear implant system "Combi 40" of the Austrian firm "Med El". The width of the spectrum of the speech signal processed by the processor, is in the range from 300 to 5500 Hz. CPU implant "Combi-40" divides this range into 8 strips. In each of them measured energy, and in accordance with its magnitude is formed by a pulse applied to the respective electrodes. The program incorporated into the speech processor, allows you to change the boundary (and Central) cha is toty these bands depending on the applicable law processing. The program of the implant are 4 of the law division of the spectrum: logarithmic, linear-logarithmic, tonotopically and linearly falling.

In this paper was applied a computer program for spectral processing of the speech signal. It allows you to cut from the acoustic signal different parts of the spectrum. As test signals were used frequency band width of 50 Hz, which were located around the Central frequencies of bandpass filters sound processor. The remainder of the spectrum was removed. Thus, after processing the speech signal in this informative program was only 8 frequency bands with a width of 50 Hz. Therefore, the spectrum of speech (300-5500 Hz)processed by the speech processor of the implant, there was only 8×50=400 Hz. Thus, for the assessment of speech perception were used only 400/5200=1/13 part of the spectrum.

After processing the speech signal in accordance with the 4 laws of separation of the spectrum was carried out measuring the intelligibility of speech. This was used speech table Grinberg-Zinder to 30 words each. The speech signal with the IBM PC is supplied to the amplifier "Amfiton" and then on the phone TDH-39. Tone controls of the amp set to the middle position, providing a uniform amplification of all frequencies of the speech spectrum. The subject in biret only the level of comfort loudness of the speech signal. Before registration, conduct a short training session (5-10 minutes) for the development of the new sound of the Russian language.

Measurement of speech intelligibility conducted in 6 people with normal hearing when a monaural presentation of speech stimuli at the right ear. All subjects showed 100% speech intelligibility in all four variants of the processing of the speech signal. Based on the fact that after processing in the speech signal could only 1/13, i.e. 7.7% of the spectrum, we can conclude that the redundancy in the speech from the point of view of the frequency representation exceeds 92,3%.

Based on the results of such a significant redundancy of speech (from the point of view of the frequency representation), it becomes clear why patients implanted with such a low frequency selectivity successfully learn a new sound image of the words of the Russian language.

It should be noted that in the beginning of the study all subjects were lost from such an unusual voice signal from the first training dozens of words (regardless of the law division of the spectrum) they are discerned only 2-3. But quickly grew accustomed to a changed pattern of speech, and further testing is not caused significant problems. To achieve 100% speech intelligibility was enough only for a short workout.

This representation of the speech signal mo is but be regarded as a model of perception only in the first approximation, but to demonstrate it unusable.

Thus, on the basis of the results obtained in this way, it is possible to draw Parallels between speech perception, processed comb filter, people with normal hearing and speech perception implanted patients and then to simulate the auditory perception of patients after cochlear implantation.

However, it should be noted that in the described method of processing a speech signal, the result is not sufficiently conspicuous, as is not difficult to learn the perception of the processed speech. In turn, the results of this method show that perception depleted spectral content of the speech can be learned. Therefore, to create a working model, you must select the appropriate parameters comb filter and methods of presentation of the speech material.

There is also known a method of processing signals, namely, that the input signal after analog-to-digital conversion share with bandpass filters on multiple frequency bands, after rectification and smoothing of these bandpass signals in each of the bands measure the energy and in accordance with its value form the amplitude of the sine waves which are then added and after digital to analog conversion on the deposits received by the output device [2].

The purpose of this conversion is to assess the perception of speech processed by analogy with the program the speech processor of the cochlear implant.

The most significant disadvantage of this method is that during this conversion of the speech signal is excluded spectral representation of the speech signal, i.e. completely missing the real spectrum of the speech signal.

The closest solution to the claimed proposal is a method of processing a speech signal to simulate auditory perception in implanted patients by analog-to-digital conversion of the input signal, expressed in word form, the separation of the spectrum of the converted signal on the frequency, odd and even, the lines, the summation of odd bands, digital to analog conversion of the received total signal and training in perception through the preliminary examination with the word presented for listening and testing [3].

The disadvantage of this method is that during this conversion of the speech signal the allocation of the bands reflects insufficient spectral representation of the speech signal, determined by the arrangement of the electrodes implanted in the cochlea of the patient, the training does not correspond to that which is implanted in patients, and testirovanie is a new speech material.

The technical result of the proposal is to demonstrate to people with normal hearing peculiarities of the perception of words patients after cochlear implantation to assess the challenges facing the implanted patients, and to help surdopedagogs who are engaged in rehabilitation of these patients.

This result is achieved by the fact that in the simulation of the auditory perception of the acoustic signal by the patient after cochlear implantation by an analog-to-digital conversion of the input signal, expressed in word form, the separation of the spectrum of the converted signal on the frequency, odd and even, the lines, the summation of odd bands, digital to analog conversion of the received total signal and training in perception through the preliminary examination with the word presented for listening and testing, according to the invention the separation of the spectrum produced by tonotopically law distribution of the frequency axis of the cochlea, while the frequency bands with odd numbers have equal distance from each other along the length of the basilar membranes in accordance with normal tonotopically the frequency distribution along the axis of the cochlea, summarize at least three odd band of the spectrum, and the training is conducted by multiple repetition of words is, required to listen to unambiguously correlate it with the known meaning of the word, obtained by pre-reading, while for presentation during testing, use the same words as in the training.

The presence of distinctive features, namely the division of the spectrum tonotopically law distribution of the frequency axis of the cochlea, the location of the frequency bands with odd numbers on equal distance from each other along the length of the basilar membrane in accordance with normal tonotopically the frequency distribution along the axis of the cochlea, the summation of at least three odd bands of the spectrum, learning through repeated repetition of words presented for listening to unambiguously correlate it with the known meaning of the word, obtained by the preliminary read, use for claim when testing the same words as in the training, demonstrates compliance with the proposed technical solution criterion the patentability of "novelty."

Due to the proposed method had the opportunity to demonstrate to the relatives of the implanted patients, the complexity of the perception of a new sounding speech directly to ensure that the development of this sound pattern and on their own experience to assess the problems facing implanter the data of the patients.

This method is based on surveys conducted in Saint-Petersburg scientific research Institute of ENT with the participation of six subjects people with normal hearing.

The conducted research was based on the conclusion about the origin of the feeling of the band during stimulation of the auditory nerve electrical impulses [4] and on information about training opportunities the perception of spectral deprivileging speech signal [1].

The method is as follows.

The speech signal is fed through an analog-to-digital Converter on the IBM PC, spend comb filtering processed signal is divided into even and odd bands. So the bottom border of the first odd band equal to 200 Hz, and the upper bound of the last fringes of equal 6250 Hz. Odd strips have a width of 50 Hz and in accordance with normal tonotopical snails they are at equal distances from each other (10,4 mm) along the length of the basilar membrane.

These calculations were carried out in accordance with the formula Hartmann W. the dependence of the characteristic frequency of a neuron from the coordinates of a point on the basilar membrane [5].

The signals of three odd lines summarize and further through a digital-to-analog Converter output on the output device.

As the characteristics of the bandpass filters we used the bandwidth of the filter at level I=-20 dB, i.e. the width is preset frequency bands on the intensity level of 20 dB below the maximum value of the intensity in this band.

The bandwidth of all filters was equal to 170 to 200 Hz, i.e. used a computer program provided the steepness of the decline fronts bandpass filter 0,3-0,35 dB/Hz, as in the direction of high and low frequencies.

The drawing shows the spectrum of the speech signal processed comb filter and used in our work, where the x-axis is frequency in Hz, the y-axis is the intensity level in dB of attenuation.

As a result of the research it was found that at the first listening of the two lists, tables, containing 60 words, subjects identified only 11,7±4.3 words, i.e. less than 20%. The most frequent assessment of the quality of the sound was as follows: "it cannot be that these words are. And if this word, it is impossible to understand."

It should be noted that the individual scores of the subjects were located in a wide range - from 4 to 17 words. Since all subjects - persons with normal hearing, we can say that from the point of view of the state of auditory function during the survey they were in the same position.

However, as you can see by the results, they are in varying degrees master the perception of processed words. The detected range of results is a bright illustration of the different abilities of different people get used to the new sound picture of speech. That is, in case of equal initial conditions etc which are psychophysical features of "same" in auditory function of the examinees.

On the basis of analogy with cochlear implantation with a variety of characteristics of implanted patients differences in the test results allow the audiologist to answer a common question parents: Why different results rehabilitation of their children?

Upon presentation of words with a prior acquaintance of the subjects about what word they will hear, it was found that the treated words become recognizable. All subjects were identified audible as they already know the word, and many wondered, saying: "that's Exactly it."

Thus, in the remaining after computer processing of three strips of width 50 Hz contains such spectral characteristics of speech, which is enough to correlate with the word with a full range.

It should be noted that the subjects differed also by the ability to correlate know the word and its sound image obtained after processing. Part of the subjects did not immediately agree that they hear a word that they said. Some recognized the similarity with the first presentation.

To the subjects admitted that they hear the same word that they said that different subjects had to listen to the processed word from 1 to 4 times. That is, in this case again the differences between the subjects.

In ex the contrast of the low at first sight (the average for the group - 11,7), with repeated listening - after training, subjects identified 35,8±5,6 words. Thus the intelligibility of speech has increased more than three times. Such a large (three-fold) (p<0.01) increase in the intelligibility of speech can be explained by two factors, namely, first, some words during simultaneous presentation memorable, and, secondly, there is training on the new perception of the sound pattern of speech.

In our opinion dominates the second factor, because, first, it is difficult to assume that once the listening subject remembered the test words. Second, all subjects, except one, themselves say that they do not remember a single word and in their new understanding of words only role played by the training.

It should be emphasized that, despite the considerable scatter of individual results, between the first and the second presentation of the words found a strong correlation (r=0.89). Because at the first listening test subjects were based only on auditory perception, high correlation allows us to conclude that the determining factor in the development of new sound pattern of speech is not memorization, and learning.

But in any case, it is entitled to assume that it is in this way (learning and remembering new images) and the development of play is embracing speech implanted postlingually patients also learn to perceive significantly distorted speech, Ostrava new picture to the former. Typing "library" of images in the process of developing new patterns of speech, they are getting better and better identify new unknown words. After accumulation of auditory experience they say they hear as before. That is, here again, you can draw Parallels between our study and perception postlingually implanted patients. It is obvious that in this and in another case, the subjects in some way set up (rebuild) their perception on the new pattern of speech.

Differences between the first and final measurements of speech intelligibility in subjects people are used to them a vivid demonstration of the different abilities of people to understand the spectral converted speech and on the basis of analogy with cochlear implants allow the audiologist to explain to parents the differences in the results of rehabilitation of their children.

It should be noted that at the end of the study, all subjects expressed surprise at the results.

As can be seen from the obtained results, when using only a 150*100/6000=2,5% of the spectrum of speech, subjects, people with normal hearing after short-term training begin to identify more than half of the test words. This is a surprising result with such poor is petrolina representation of the speech signal, and proceeding from him, and Parallels with the KEY, you can understand why implanted patients can understand significantly distorted speech.

It should be noted the effectiveness of such demonstrations for parents of implanted children. At the first listening processed words parents recognize a few words and some are even afraid that, as I hear their children. But after training and re-listening they are very surprised with the achieved result, and convinced of the correctness of the operation selection KEY. This demonstration is also very useful for surdopedagogs who are engaged in the rehabilitation of cardiac patients. Working previously with deaf patients, they unwittingly bring their experience with them on a whole new contingent of patients and do not take into account peculiarities of the perception of cardiac patients [4].

Hearing about such demonstrations, they are asked to show them the program and after the examination and develop a better understanding of the problem and several differently assess their own classes with implanted patients. When discussing some questions after the demonstration surdopedagogs start to ask more practical questions.

We are aware that the proposed program is not an absolute model of auditory perception in implanted patients, n is, based on our own data and global experience, we find that our ongoing demonstration program useful and quite illustrative.

The proposed method is carried out in the Department of pathophysiology of the ear of the St. Petersburg scientific-research Institute of ear, throat, nose and speech.

The use of the proposed method for processing speech signal allows to obtain the following benefits :

1. The results of the study demonstrate the possibility of simulating auditory perception of cardiac patients using the developed program.

2. The results obtained indicate the existence of Parallels between perception of spectral deprivileging speech signal subjects with normal hearing and speech perception implanted patients.

3. The participation of parents of implanted children in experiments on the perception of the processed speech is for them a vivid demonstration of what he hears and learns how perception of speech processed by the processor of the cochlear implant for their child.

From the above it follows that the technical result of the invention is achieved by a new set of essential features, hence the claimed method meets the criteria of patentability "inventive step".

The essence of the method is illustrated with the leader.

After explaining the nature of the survey and conducting training in the full spectrum of the speech signal began the study of the intelligibility of speech processed by our program. At the first examination of the tested P. (mother implanted girls) correctly identified 4 of 60 words (6%) and in her words sounding speech was terrible. During training, i.e. when listening with prior acquaintance about what she might hear, she was very surprised that recognize know the words. With the re-examination tested is identified 24 words (40%) and was surprised with the results. Moreover, she noted that he did not remember the test words, namely mastered new pattern of speech. Such striking result gives the opportunity to subjects with normal hearing to understand the problem of the development of speech perception implanted patients. After the first turn of the implant and for setting CPU her daughter the mother is sympathetic to her problems and even gave some explanations.

The results of the study and made in the course of the experiment observations indicate Parallels between perception of spectral converted speech signal subjects with normal hearing and speech perception implanted patients and give grounds to assert that offer aemy method of speech processing can be used to demonstrate normallesmesini process of development of speech perception by patients after a surgery cochlear implantation.

According to the described scheme examined in 6 subjects. The study found that after training there was a significant improvement of speech intelligibility (p<0.01) for presentation of the processed method proposed speech signal, on the average, from 20 to 60%.

It follows from the above that the proposed method provides the technical result, straightforward, involves the use of the developed materials and standard equipment that demonstrates compliance with the proposed technical solution to the patentability criterion of "industrial applicability".

SOURCES of INFORMATION

1. Lantsov A.A., Petrov S.M., Poods VI TO the question of modeling the auditory perception of cochlear implant users, " Vestn. torinla. - 1999. - N-6. - C.21-23.

2. Loizou, S., Dorman M. and Z. Tu On the number of channels needed to understand speech // J. Acoust. Soc. Amer. - 1999. - Vol.106, N 4. - P.2097-2103.

3. Petrov S.M. Perception of spectral deprivileging speech signal // human Physiology. - 2003. - 1. - P.72-74.

4. D.K. Eddington, W.H. Dobelle, D.E. Blackman e.a. Auditory prosthesis with multiple channel intracochlear stimulation // Ann-Otol-Rhinol-Laryngol. - 1978. - Vol.87, N 6. - Part 2. Suppl.53. - P.1-39.

5. Hartmann W. Pitch, periodicity and auditory organisation // J. Acoust. Soc. Amer. - 1996. - Vol.100, N 6. - P.3491-3502.

The simulation of the auditory perception of the acoustic signal by the patient after cochlear implantation by an analog-to-digital what about the conversion of the input signal, expressed in word form, the separation of the spectrum of the converted signal on the frequency, odd and even, the lines, the summation of odd bands, digital to analog conversion of the received total signal and training in perception through the preliminary examination with the word presented for listening and testing, characterized in that the separation of the spectrum is carried out on tonotopically law distribution of the frequency axis of the cochlea, while the frequency bands with odd numbers have equal distance from each other along the length of the basilar membrane in accordance with normal tonotopically the frequency distribution along the axis of the cochlea, summarize at least three odd band spectrum, and the training is conducted by the repeated recurrence of words presented for listening to unambiguously correlate it with the known meaning of the word, obtained by pre-reading, while for presentation during testing, use the same words as in the training.