HRP20000641A2 - Multichannel implantable inner ear stimulator - Google Patents

Description

Overview of related patent applications

This application claims priority from the U.S. Provisional patent application no. 60/080,268 filed Apr. 1, 1998.

Background of the invention

The present invention relates to a system and procedure for electro-stimulation of the inner ear. More specifically, the present invention relates to an implantable device for electrical stimulation of the 8th nerve. In more detail, the present invention relates to an implantable device for electrical stimulation of the 8th nerve to create the sensation of hearing.

It is well known that the brain and nerve impulses have an electrical character. It is also known that electrical stimuli applied to receptor centers such as nerves cause a response dependent on the electrical characteristics of the stimulus. Many devices use these characteristics to compensate for the deficient functioning of the body's sensory organs.

In normal hearing, hair cells are a critical link in the auditory chain. They support two functions together with the brain: (1) they establish background neural activity that is perceived as silence ("active silence" as described below); and (2) when sound enters the ear, they produce a potential that varies and modulates that background neural activity in response to the sound. The resulting nerve activity is a constant, plus the derivative of atmospheric pressure. This derivation or the intensity of the pressure change carries information about the sound. Important for the present invention is the realization that the intensity or frequency or density of the resulting nerve activity can be viewed as a carrier frequency modulated by sound.

In completely deaf patients, the root cause of deafness is loss of hair cell function. In 30% of deaf people, loss of nerve fibers from the spiral ganglion to non-functioning hair cells is a contributing cause of deafness. This may be caused by the inactivity of the nerve fibers from the hair cells to the spiral ganglion. Therefore, in order to restore hearing to a person with partial or total hearing loss, it is necessary to replace these functions beyond the point of loss of function, which means at a higher connection with the brain.

In the case of the ear and associated auditory functions, many devices have been developed to electrically stimulate the auditory nerve of the human body, which is known as the 8th (eighth) cranial nerve. In any case, these devices work on principles derived from an inappropriate extrapolation of certain observations made by Beckes in the thirties of this century. Beckesv's considerations were related to the basilar membrane of the ear, which extends along the entire length of the cochlear apparatus (cochlea of the ear). These observations revealed that the basilar eardrum vibrates in response to the vibrations of sound entering the ear. Beckesv observed, which was also confirmed by others, that sound vibrations caused the membrane to vibrate with a standing wave, whereby the maximum amplitude of the standing wave appeared at some point of the membrane depending on the frequency of the incoming sound vibrations. The activity of individual hair cells in these places was also particularly emphasized in the places of maximum amplitude of the standing wave. High frequencies resulted in maximum amplitude at the entrance to the cochlea. As the frequency decreases, the location of this maximum amplitude moves towards the very end of the cochlea.

While this mechanical activity is correct and the individual activity of the hair cells is enhanced at those places of maximum amplitude, the others inadequately extrapolated the above observations to reach the conclusion that hearing was affected by the response of individual nerve fibers along the cochlear apparatus depending on the frequency. Thus, a theory was developed, known as the Place Theory of hearing (hereinafter referred to as the Location Theory of Hearing), that nerve fibers in the cochlea transmit different frequencies to the brain depending on their location in the cochlea. It is strange that the absolute length of the canal of the cochlear apparatus, which varies from 5 mm in the chicken to over 100 mm in the whale, does not seem to play an important role in the frequency range of the cochlea, i.e. the whale has only a slightly larger frequency range than the chicken, although according to the Location according to the theory of hearing, with a 20 times longer cochlea, the range of frequencies in a whale should be 20 times greater than in a chicken.

The locational theory of hearing requires that the nerves in the cochlear apparatus work in a different way than all other nerves in the body. The subject invention is based on a model of hearing that is completely different from the Locational Theory of Hearing. This invention, contrary to the Locational Theory of Hearing, is based on the application of the principle of signal processing to the function of the nerve fibers of the 8th nerve ending in the vestibule of the ear and the cochlear apparatus, very similar to the operation of modern communication receivers that use digital signal processing to reduce noise and data processing.

Nerves terminating in the vestibulum and/or cochlea that transmit sound sensation are nonspecific and can fire in sequence or as sustained background neural activity with a single impulse that, when modulated, produces a sound sensation of modulation for a given time period. Accordingly, according to the principles guiding this invention, the nerve fibers of the 8th nerve act in a manner identical to that throughout the body. More precisely, the signal sent by the nerves is non-specific, but the number of nerves that trigger the signal and the frequency of the triggering transmits information to the brain, which information the brain translates into sound. The number of nerve fibers firing simultaneously, or at such a rate of repetition as to appear simultaneous, is a function of constant sound intensity, and variations in this nerve activity are perceived as sound.

The model of hearing on which the present invention is based observes that many nerve fibers of the cochlea have other functions besides conducting sound. It was found that a very regular spatial arrangement of sensory elements in the cochlea creates a predisposition for its work on spatial principles, but still not in accordance with the Locational Theory of Hearing. It has been observed that stimulation of many of these fibers does not produce the sensation of sound. The brain uses the cochlea as a mechanism to regulate variations in sound pressure as a result of sound vibrations and thus serves as a means of regulating loudness (volume).

In this way, some of the outer hair cells of the cochlea sense the movement of the basilar membrane, transmit this information to the brain, which then sends feedback signals to the many hair cells in the cochlea to regulate the stiffness of the basilar membrane and thus regulate the mechanical impedance at the entrance from the vestibule to the cochlear apparatus. . This allows for automatic loudness regulation (in the mechanical realm) and a possible means of frequency response regulation to improve clarity. Changing the mechanical characteristics of the basilar membrane changes the mechanical transmission of energy to the hair cells, thus affecting sensitivity and frequency response. The cochlea can also contribute to the sound localization process.

It has been established that the audio signals of speech and music have most of their energy concentrated in the lower frequency ranges. In order to achieve an improvement in the signal-to-noise ratio, pre-amplification (amplification of the signal before transmission) of high frequencies and a corresponding reduction at the point of detection in the brain should be considered. In accordance with this opinion, Beksev published in 1960 that vibration patterns of the cochlear septum in cadavers at different frequencies showed a preamplification of high frequencies in the first 10 mm of the distance from the stapes (middle ear). Year In 1974, Rhode published a graph of the input-output relationship, in decibels, for the malleus (middle ear) and basilar membrane (Figure 21A). The graph shows an increase of 6 dB per octave (or 20 dB per decade) of frequencies between 200 Hz and 8 kHz. Likewise, Figure 21B shows that a wide range of frequencies stimulates the hair cells in that area. These observations support the concept of premagnification. Observations also suggest that the outer hair cells of the cochlea have the function of providing information to the brain to regulate volume, dynamic range, and have an effect on frequency response, rather than transmitting the sensation of specific frequencies to the brain.

In addition, it is largely unknown that the neural activity that produces sound consists of the sum of neural activity in response to external sound or stimuli that modulate ongoing background neural activity. This constant background nervous activity was described by R. Lorente De No god. 1976 as follows; "In the absence of peripheral stimulation, the acoustic nuclei are an area of continuous activity maintained by the arrival of nerve impulses spontaneously initiated in the cochlear apparatus. The activity is necessarily accompanied by the circulation of impulses in the chains of neurons.

Considering that people perceive spontaneous activity in the cochlea and acoustic centers as silence, it must be concluded that the spontaneous activity serves to determine the background states of different subsections of the acoustic nuclei, to which deviations caused by sound refer. In other words, what we hear is the result of these deviations from the basic states or the base signal of the acoustic nuclei, which are caused by external sound sources." He calls this background state "active silence" to which the perception of sound refers.

Although others have studied this activity, no one has recognized it as a carrier frequency (carrier) which is the sum of non-specific nerve activity and modulated by external stimuli. Recognition of this principle is an important element of the subject invention. This realization is consistent with the sampled data theorem developed by Hartley of Bell Labs and Nyquist et al. 1928, where "active silence" is taken as the carrier frequency.

It is not necessary that the nerve activity is a sequence of activity of a single nerve fiber, but that the nerve activity is at such a high frequency, that it is outside the range of audible sound, double or multiple simultaneous nerve activity can be established. Active silence can be compared to the molecular activity of a gas at a given pressure (silence), and the modulation of that activity to pressure variations due to sound.

The mechanical characteristics of the basilar membrane at the entrance to the cochlea (see Fig. 21A and 21B) are such that the modulation is greatest for high frequencies and decreases at a rate of 6 dB per octave to lower frequencies. Audio signals of speech and music have been found to have most of their energy concentrated in lower frequency areas. Amplification of the high-frequency components of the audio signal is introduced before the neural activity noise is introduced, to the point where they create a constant deviation of the background neural activity as a function of frequency. This equalization of low-frequency and high-frequency parts of the audible spectrum enables the signal to completely occupy the width of the frequency band of the neural communication link. The spectrum of noise (noise) introduced at the summation output of the nerve occupies the entire bandwidth. The noise power spectrum of the output sum is amplified at higher frequencies. At the output of the sum of nerve fibers, an inverse function is introduced, the reduction of high-frequency components, which restores the power distribution of the original signal. This reduction process also reduces high-frequency noise components and thus effectively increases the signal-to-noise ratio.

This high-frequency preamp function compensates for the inverse function at the far end of the nerve bundle in the brain. This is similar to preamplifying the high frequencies in an FM (frequency modulated) transmitter and then attenuating the high frequencies in the receiver. The result, along with the compensation of the "receiving" function of the brain by the basilar membrane, is an improved signal-to-noise ratio. A secondary characteristic of the cochlea is that, in fact, all frequencies stimulate the nerve fibers near the ear vestibule (vestibulum) with pre-amplification. High frequencies predominate at the input, and low frequencies predominate at the other end. However, the sense of frequency does not refer to which nerve fibers are stimulated, but to the change in the total activity of the nerve fiber when the sum of the total activity of the nerve fiber is observed (see Fig. 20).

The previous function of the cochlea can be compared to a system of low-frequency, mid-range and high-frequency loudspeakers. When the sounds reach the ear, the individual hears the sum of the activity of each speaker. Similarly, the brain receives signals that are the sum of the activity and signals sent from the hair cells and the associated nerve activity. In any case, it is important that the nerve activity related to each individual stimulated hair cell contributes to the summed nerve activity completely independently of the contribution made by the nerve activity of the other stimulated hair cells, but is influenced by other nerve activity. So often, when viewed in isolation, neural activity appeared to be frequency selective. However, when viewed more closely in terms of recognizing spontaneous or background hair cell activity as a carrier frequency for received sound stimuli, the subject invention becomes clear that it is the modulation of background or spontaneous neural activity that is "heard" and not the neural activity associated with individual cells hair.

it is easily true that different frequencies can increase the activity of nerve fibers in different areas of the cochlea, this does not affect the transmission of sound. The summed change in nerve activity from the vestibule and cochlea is heard as sound, not which nerve fiber is activated at any given time or when a given frequency is heard. This concept was first proposed by Rinne in 1865, but he had no formal theory to support it. Year In 1880 Rutherford gave a plausible explanation, THE TELEPHONE THEORY. However, at that time little was known about the characteristics of nerve fibers and this was nearly 50 years before Hartley and Nyquist's SAMPLED VALUE THEOREM.

Physiological characteristics of the eighth (8th) auditory nerve are equally important in the design of any system based on the above accepted theory of hearing. More precisely, five characteristics play an important role in the design of all such systems: strength-duration, flow, latency, regeneration and fatigue (loss of properties due to external effects). The strength-duration characteristics of human nerve fibers are graphically represented by the strength-duration curve shown in Figure 1A. The strength-duration curve expresses the relationship between the smallest amount of current (stimulus) applied to a nerve fiber and the shortest time in which the current (stimulus) must flow to reach the excitation threshold. In other words, the strength-duration curve is a graphic representation of the intensity threshold just sufficient for axon excitation and its relationship to the duration of the stimulus current. Indeed, nerve fibers will not respond with excitation to current densities below some minimum. The strength-duration curve is further described in Medical Physiology and Biophysics, Ruch and Fulton, 18th ed., W.B. Saunders &Co. Ruch and Fulton model this nerve behavior according to a capacitive circuit with a single resistance. The power duration in combination with the electric field gradient determines the pulse length range for the stimulus pulse caused by the "current".

Current is the sequential (sequential) activation (triggering) of nerve fibers within a group of nerve fibers or ganglia that is stimulated by a single-pulse stimulus. When stimulated with a single-pulse stimulus such as a single square wave via a gradient electric field acting on a group of nerve fibers or ganglia, the individual nerve fibers within the group will each receive a stimulus that decreases in intensity as the individual nerve fiber within the group moves away from the source electrode. This phenomenon is shown in Fig. 1B. This firing frequency of a single nerve fiber will correspond to the one shown on the strength-duration curve from Fig. 1A. So, when firing of individual nerve fibers within the stimulated group of nerves begins, those fibers that are closer to the source of the field gradient will fire (transmit) signals at a faster rate, while those further away will fire more slowly.

So, a nerve group will transmit a series of signals, i.e. a stream of nerve activity, over a period of time. More precisely, this "flowing" is indicated by the fact that some nerves in the group will trigger signals in sequence, one after the other, whereby the later triggering will be with a lower speed than the previous one.

The current appears during the last part of the impulse stimulus of an individual nerve. The length of the current of the group of nerve fibers is limited by the delay of the start of the current at the beginning of the pulse (not earlier than 0.1 millisecond according to the strength-duration curve from Fig. 1A) and the time remaining until the end of the stimulus pulse. This behavior is shown in the upper part of the graph from Fig. 1A line labeled "Frequency of Nerve Firing". The latent stage is the delay between the start of the stimulus impulse and the reaction (activation) of the nerve fiber.

Figure 1A shows that the latency for each nerve is different as defined by the strength-duration curve and the field gradient. In practice, it is desirable that the start of streaming is delayed by more than 0.2 ms. Since the latency is shortened by increasing the stimulus amplitude, the compensatory component required to keep the differential latencies constant during streaming requires increasing the stimulus to excessive amplitude for a system with a long streaming time (as in a 4-channel system). The minimum latency time also determines the time overlap for neighboring channels, i.e. the time in which the neighboring or other nerve fibers must be stimulated in order to continue the transmission of the total signal after the channel of the underlying (original) nerve fiber has entered the regeneration phase (returning to the initial state ).

In addition, individual fibers of the 8th nerve cannot transmit the stimulus indefinitely. After receiving and transmitting stimuli, the nerve must undergo a period of regeneration. If it fails, the nerve will fatigue (lose its properties) and stop transmitting. The characteristics of nerve regeneration limit the rate of repetition of stimulation of an individual channel. Finally, nerves are damaged by prolonged stimulation with the average component of direct current (DC). All such stimuli must be performed with alternating current (AC).

When an electric field acts on auditory sensory branches of the 8th nerve including the brainstem or spiral ganglion, the angle of arrival produces a field gradient across the entire nerve group, which can cause sequential nerve firing. This is shown in Fig. 1C, which shows how the strength of the electric field weakens across the cross-section of a group of nerve fibers, between the cathode and the anode. Due to this weakening of the electric field, the nerve fibers are activated (triggered) one after the other. Additionally, due to the combination of the electric field and the strength-duration characteristics of nerve fibers, as the distance from the cathode increases, the time between successive firings of the nerve signal also increases. In contrast, in the arrangement shown in Figure 1D, all nerve fibers are subjected to essentially the same electrical potential and will therefore fire largely simultaneously.

If the amplitude of the stimulus is small, so that it does not produce a carrier frequency high enough to be above the audible range, the streaming frequency can be heard. Its frequency is a function of the location of the stimulus on the intensity-duration curve relative to the stimulated nerve fiber and the slope of the curve. This changes with time and stimulus amplitude. As mentioned before, the signal sent by the nerves is non-specific, but the number of nerves sending it and the frequency of firing transmits the information to the brain, and the brain translates it into sound. For example, as the amplitude increases, the rate of successive firing of the nerve fiber increases. If the angle of the stimulus is almost vertical, there will be a high speed of successive transmission because individual nerve fibers of the channel receive almost the same stimulus. If the angle is small, successive triggering is at a lower frequency because the difference in stimulation across the section of individual stimulated nerve fibers will have a larger range.

Well-known devices designed to help completely deaf people by means of electrical stimulation of the 8th nerve and based on principles guided by the Location Theory function primarily because of this angle and stimulus dependence, but with results that are not predictable, repeatable or optimized. In the U.S. In patent 3,449,768 issued to James Doyle, the system was not designed (implemented) based on the principles of the Locational Theory of Hearing, but was designed to produce a carrier of neural activity based on sequential stimulation of multiple channels at a rate high enough to result in a carrier frequency of neural activity suitable for modulation by information sound. That patent discloses a device for applying electrical stimuli to the 8th cranial nerve and includes a system of electrodes to be placed near the auditory nerve, elements (means) for delivering pulses to a number of transmission channels and a modulator that modulates the time-amplitude integral of each pulse.

This system is limited by, for example, the number of channels required and because the time allowed for regeneration of each channel is too short to allow prolonged stimulation without causing nerve fatigue. No attention was paid to the characteristics of the latency of nerve fibers (the delay between the start of the stimulus impulse and the reaction /triggering/ of the nerve fiber). Additionally, the earlier Doyle system failed to provide compensation in pulse strength to maintain a constant trigger intensity and avoid the current attenuation present, as previously described. Finally, the previous Doyle patent did not acknowledge or indicate that the carrier frequency (frequency or density of background neural activity) is independent of the intensity (rate) at which individual channels fire and the number of individual channels. These limitations or deficiencies have resulted in a system with poor sound clarity, a signal to sound ratio that is lower than otherwise achievable, and constant current background (hum) or tone received by the patient.

Summary of the invention

The object of the present invention is to improve systems for stimulating the auditory nerve of the human body.

Another object of the present invention is to improve the system disclosed in the U.S. Patent no. 3,449,768, issued by James Doyle.

A further object of the present invention is to produce sustained neural activity that mimics the spontaneous neural activity present in normal hearing and to modulate that activity with an audio signal to achieve audibility.

A further object of this invention is to provide a new system for stimulating groups of bundles of nerve fibers of the 8th cranial nerve in such a way as to cause flow in the channel at a constant intensity (speed).

Another goal of the present invention is to achieve this constant flow intensity in a way that applies to the strength-duration characteristics of the nerves.

A further object of the present invention is to modulate this flow in the channel with audio information to produce a sense of hearing in the totally deaf or in those with other hearing impairments.

The subject invention enables a new system for stimulating groups or bundles of nerve fibers of the 8th cranial nerve in such a way as to achieve a flow in the channel of constant intensity. With the recommended embodiment of the present invention described in detail herein, unlike any prior device, this constant intensity flow is achieved in a manner that matches the strength-duration characteristics of the nerves. Furthermore, the system modulates this channel flow with audio information to produce the sensation of hearing in the totally deaf or those with other hearing impairments. The subject invention further provides a method for stimulating these groups or bundles of nerve fibers.

In accordance with a preferred embodiment of the present invention, an electrical stimulus is applied to activate the nerve to fire a signal at a constant intensity. The increase in stimulus strength enabled by the device according to the present invention is controlled by the strength-duration characteristics or the behavior of the nerve fibers or ganglia. By adjusting the field gradient produced by the electrodes placed near the 8th nerve for the nerve characteristics shown by the strength-duration curve, individual nerve fibers located within the field gradient need not fire simultaneously when stimulated, as has been interpreted so far, but will fire in sequence neural activity as time progresses.

Preferably, the stimulus does not appear at the endpoints of the intensity-duration curve. This is because the high voltage required to maintain a very short latent stage can produce unwanted electromechanical reactions. Likewise, a long latent stage results in excessive channel overlap and reduces the available time per channel for streaming. It is recommended, for example, to reflect the latent stage between 0.1 and 4.0 milliseconds; more preferably, the latent stage can be maintained between 2 and 3 milliseconds. It should be emphasized that the recommended latent stage may vary depending on the specific subject or patient, and in some individuals it may be appropriate or advisable to work outside the previously described ranges.

While this flow of nerve activity is in progress, the device according to this invention modulates the stimulus pulse in order to change the intensity of the nerve activity and cause the signal to be transmitted to the brain via the 8th nerve or ganglion, which is perceived as hearing. It also modulates the combined stimulus pulse and its audio modulation to ensure that the stream remains constant when there is no sound. It follows that when increasing the amplitude of the stimulus pulse, the percentage of modulation due to sound remains constant. In other words, as the amplitude of the stimulus increases, so does the audio modulation component.

In embodiments of the present invention in which more than one channel is used, adjacent channels overlap due to the latency and regenerative characteristics of the nerve. This is done by transmitting a constant stream of summarized channels. The device according to the present invention therefore provides a constant flow of nervous activity independent of the audio modulation acting as a carrier wave, but this will not be perceived by the brain as sound but only as "active silence" in the sense of the name accepted in this field of science. In addition to stimulating the nerve fibers in such a way as to create a wave of carrier frequency, the subject invention further uses this carrier wave to stimulate the sensation of hearing by modulating it at appropriate moments during the duration of the stimulus signal and the response of the nerve fibers.

Devices made in accordance with the instructions of the present invention include means for generating background neural activity that the brain perceives as silence. Such devices use this background neural activity as a carrier of audio information and modulate that neural activity with audio information. This modulation causes variations in the density of background neural activity that the brain perceives as sound. It is important, in any case, that such modulation be performed in such a way that the frequency or density of background neural activity is independent of the number of channels used by the device (as long as that number is greater than one) and the intensity with which any given channel is stimulated. Thus, the intensity with which a stimulus is applied to a channel, and the duration of any pulse, is secondary in function to the main goal of this invention. Indeed, the summed carrier frequency of the nerve can then have more than 1000 cycles (oscillations) per second, independent of the modulation frequency.

A procedure was discovered that stimulates the current of nerve fiber activity that results in the background state of nerve activity on the audio transmission branch of the 8th nerve to flow into the brain and the modulation of this current (pseudo carrier) with audio information. A device for stimulating the auditory transmission branch of the 8th nerve is described. It uses electrodes designed to limit the electric field to the zone of their corresponding nerve fiber group and to create such a field gradient for each channel that the latent characteristics of nerve fibers in a given channel cause sequential firing (current) of nerve fibers during part of the channel's stimulus pulse. In the analog system, nerve fiber channels are stimulated sequentially, but their stimulus overlaps with their previous channel by an amount equal to the shortest latent period of the channel.

During the period in which the stimulus causes the nerve fiber to flow, the amplitude of the impulse is modulated by the audio information. In addition, during the current of the nerve fiber, either an electrical element or in the form of a tentacle or both compensate, through the strength of the stimulus, the Strength-Duration characteristics of individual nerve fibers so that, when sound is not present, the intensity of the current is mostly constant, resulting in "active silence". i.e. with a minimum of sound perception.

Thus, the stimulus pulse is divided into two periods. In the first period, the nerves, to which the impulse is delivered, do not send (do not trigger) impulses. All nerve firings occur in the second period of the stimulus impulse. It may be noted that the two periods are usually not equal in length; in fact, with the specific examples detailed here, the second period is generally longer than the first period. This desired flow - that is, constant nerve activity with uniform intensity - can be initiated by changing the stimulus impulse during the second period, or during both the first and second periods. In any case, the audio modulation of the stream is performed only in the second period of the stimulus pulse.

In a digital system, there is no current because there is only single firing of the nerve fibers in the channel. Channel overlap exists over multiple channels and the audio modulation is in the form of frequency modulation of the streaming frequency and is independent of the frequency of the channel order.

One aspect of the present invention involves producing or amplifying a carrier frequency of background neural activity that is perceived as silence and modulating the background neural activity (pseudo carrier) to create the sensation of sound and restore some degree of hearing when the auditory organ is completely damaged.

Another aspect of the preferred embodiment of the present invention includes a system that converts sound into a corresponding electrical signal and has an encoding device to convert the analog signal into a sequence of nerve stimulus pulses that is applied simultaneously to at least 2 groups of nerve channels. In addition, the recommended system includes a transmitter link element, a receiver link element, a multi-channel gradient probe for delivering the electrical stimulus to the nerve bundle, and elements for independently adjusting the stimulus amplitude of each channel.

A further aspect of the recommended embodiment of the subject invention includes a system of electrodes inside the ear vestibule (vestibulum) and/or ear cochlea (cochlear apparatus) for stimulating individual nerve fibers of the auditory nerve in them, by generating the activity of the nerve fiber that transmits a simple impulse pattern to the brain, whereby the background neural activity (pseudo carrier frequency) is modulated by audio information, and the modulation of the background neural activity density is perceived as sound. See fig. 20.

An additional aspect of the present invention involves generating a carrier frequency of background activity that does not depend on the number of stimulus channels and allowing activated nerve fibers sufficient time to regenerate, so that fatigue of the stimulated nerve fibers does not occur.

A further aspect of the present invention involves stimulating any number (N) of different fiber groups or portions of the spiral ganglion of the sensory branch of the 8th nerve, phase-shifted at N spaced time intervals with an overlap of a portion of each adjacent group. The time interval between repetitions of stimuli of any group is generally longer than the time of natural regeneration of one nerve fiber or part of a spiral ganglion after electrical stimulation. In the recommended embodiments, five milliseconds are chosen to avoid fatigue of the nerve group after the application of the stimulus. This represents about 5 time constants of the regeneration period of the nerve, with a remainder of about 1% from the previous stimulus.

The advantage of the subject invention is that it allows a long enough time for the regeneration of the nerve fibers from the previous stimulus, so that the fatigue of the nerve fibers does not occur. A further advantage is that the present invention provides a continuous stream of nerve fiber activity that is not directly related to channel repetition rate, avoiding the limitations of the system described in U.S. Pat. Patent no. 3,449,768 in which a channel rate was required that was imposed by the criteria of sampled data theory.

It is advantageous that the strength required for nerve stimulation is reduced due to the proximity of the stimulation electrodes to the nerves of the audio transmission branch of the 8th nerve. Potentials of less than one volt are sufficient to trigger a nerve fiber without causing metal/tissue electrochemical reactions.

Application of single-cycle stimulus pulses with an average direct current (DC) value of 0, reduces the possibility of metal/tissue electrochemical reactions. Hereinafter, the term "biphasic" is used to refer to a stimulus pulse having an average DC voltage of 0. U.S. Patent no. 5,674,264 states that manufacturers of cochlear implant systems must be careful to control electrode voltages to maintain them in the zone where all electrochemical reactions occur at rates too slow to cause damage. It is advantageous that the embodiments according to the present invention effectively eliminate these reactions.

There are two limitations to the use of the strength-duration curve. If the latent period is too short, the stimulus amplitude will be large, but, more importantly, compensation to maintain a constant current over an extended period of time will require a very high stimulus. This can lead to the danger of the electrodes causing electrochemical reactions. As the number of channels is increased, this effect is less pronounced.

The present invention provides constant background activity during the course of a single nerve fiber. The subject invention solves this by canceling the non-linear characteristics of the nerve fiber current, as shown by the time constant of the strength-duration curve. Such cancellation may be accomplished in one or both of the following ways: (a) by modulating a stimulus pulse with a similar canceling time constant; or (b) shaping the field gradient originating from the electrodes in such a way as to cancel out the strength-duration curve.

The embodiments of the subject invention enable the receiver to experience normal sound. For those who have heard in the past, no special exercises are needed to interpret the sound.

The subject invention will also produce the sensation of sound for those persons whose hair cells and nerve cells have been destroyed. Often, the cause of deafness or hearing loss is the destruction of hair cells in the ear. Under such circumstances, nerve stimulation in the manner described here will create a sensation of sound. However, in some circumstances, the nerves that go to the hair cells in the ear of totally deaf or hearing-impaired patients are also destroyed. In such cases, the subject invention will produce sound sensations by electrically stimulating the spiral ganglion in the manner established herein, or by stimulating it at a higher level such as the brainstem.

In summary, the present invention generates continuous neural activity that mimics spontaneous neural activity in normal hearing and modulates that neural activity with an audio signal to achieve the sensation of hearing.

Brief description of the drawing

Figure 1A is a plot of the strength-duration curve and illustrates the gradient of the electric field acting on the linear matrix of nerve fiber terminals and the effect of the strength-duration curve characteristics on the nerve in the time each nerve fiber fires in relation to the strength of the received stimulus. Figure 1A also shows the logarithmic nerve firing rate or current;

Figure 1B is a graph illustrating in two dimensions one channel of a gradient probe and the field gradient acting on nerve endings near the probe.

Figures 1C and 1D show electric fields at two different angles relative to a group of nerve fibers.

Figure 2 is a schematic representation of the clock generator, modulator and four-phase signal generator for a 16-channel system.

Figure 3 is a schematic view of the sawtooth signal generator for a 16-channel system.

Figure 4 is a schematic view of a 16-channel multiplexer.

Figure 5 is a diagram showing the temporal relationship between the 16 channels, electrical waveforms including amplitude-duration compensation, and continuous nerve activity of the auditory branch of the 8th nerve.

Figure 6A is a diagram showing increasing stimulus strength to compensate for the strength-duration characteristics of the nerve to produce a constant intensity of nerve activity.

Figure 6B shows the audio modulation applied to the last part of the stimulus pulse.

Figure 7 is one configuration of a multichannel gradient probe in a 16-channel system.

Figure 8 is a block diagram of a 4-channel system.

Figure 9 is a schematic representation of a 4-channel clock generator, sequencer and modulator.

Figure 10 is a schematic representation of one of the four compensation circuits of the strength-duration curve.

Figure 11 is a block diagram of the interface between Figure 8 and Figure 9.

Figure 12 is a schematic representation of the logic circuit of the latent period.

Figure 13 is a schematic representation of the output attenuator of the 4-channel system and one of the four digital-to-analog converters used to adjust the channel gain.

Figure 15A is a drawing of a 4-channel probe using electrodes perpendicular to the probe and an enlarged ground plate.

Figure 15B is a drawing of a 4-channel probe showing its conformation to the shape of the scala tympany - the part of the cochlea below the spiral plate - and its position relative to the spiral ganglion.

Figure 16 is a block diagram of the digital system.

Figure 17 is a schematic representation of the external power source, microphone, frequency modulator and RF link element of the digital system.

Figure 18 is a schematic view of the internal unit of the digital system.

Figure 19 is a diagram of digital system channel waveforms.

Figure 20 is a diagram showing changes in nerve VIII activity as a result of acoustic displacement of the basilar eardrum. (Honrubia V, Strelioff D, Stiko S; Ann Otol Rhinol Larvngol 85:697-701, 1976)

Figure 21A graphically shows the input-output relationship, in decibels, for the malleus and basilar eardrum. (Rhode WS: Ann Otol Rhinol Larvngol 86: 610-6126. 1974)

Figure 21B shows the vibration patterns of the cochlear septum for the loess sample at different frequencies. (Beckesv: Experiments in Hearing, NewYork, McGraw-Hill, 1960.)

Figure 22 shows the block diagram of the system. It includes an implant that uses a microprocessor, an external patient unit, a test computer and a computer screen display.

Figure 22A shows the ability to independently adjust the amplitude of each stimulus channel, both amplitude and gain, and select the channel frequency.

Figure 22B shows the ability to adjust the strength-duration compensation for each channel, to choose a constant initial time for all channels, and the ability to fine-tune the strength-duration compensation for each channel separately.

Figure 22c shows the ability to connect the optimal channels from the probe to the stimulator, the ability to adjust the stimulus level for each channel, control the intensity of the soft start, and enter the identification number of the stimulator for permanent records in the patient's record.

Detailed description of the recommended performance

The nerve fibers in the 8th nerve that transmits audio information to the brain are divided into N separate sections. Each section consists of a certain number of nerve fibers or parts of a spiral ganglion. Each section is independently stimulated with an electrical impulse that is divided into two time periods. During the first period, the stimulus amplitude of each section is kept constant, so that the first activated nerve fibers have a latent period of approximately the same time length as the other sections. During the second period, the amplitude of the impulse will vary in such a way as to induce some of the remaining nerve fibers in that section to fire at a constant rate until the end of the stimulus impulse. This compensation can also occur during the first period, since it remains fixed and does not contain an audio component. During the second period, audio modulation is added while changing the speed of nerve firing.

If the compensation of the amplitude of the pulse includes the first period, it is still important to understand that the audio modulation is imposed only in the second period, i.e. when the flow appears in the group of nerve fibers. Sections are stimulated cyclically. The stimulation of section N+1 begins at a time determined so that its 2nd period begins at the end of the stimulus pulse of section N. During the second period of the stimulus pulse of each channel, the audio information modulates the stimulus pulse. Since the 2nd (second) period of each subsequent channel begins without a time gap between them, the flow of audio modulation information is continuous. The electrical probe used to stimulate a section of nerve fibers is shaped so that the amplitude of stimulation is different for different nerve fibers in that section. This causes the latency periods to be different for different nerve fibers, thereby causing the nerve fibers to fire in the correct order during period 2.

Figure 1A provides a graphical representation of an electric field gradient 12 imposed on a linear array of nerve fiber terminals. The graph illustrates the strength duration curve 14 in (relative) volts per millisecond. The intensity-duration curve is a graphic representation of the intensity threshold just sufficient for the excitation of one axon and its relationship to the duration of the stimulus current. The strength-duration curve expresses the relationship between the minimum power of the imposed current and the minimum time in which the current must flow to reach the excitation threshold. There is a minimum current density below which no excitation occurs. The strength-duration curve does not show the influence of subthreshold stimuli on excitability. The current below the specified threshold can be advantageously used in the recommended embodiment of the present invention as a means for overlapping adjacent channels or for amplifying the background state of nerve activity.

Figure 1A further shows (to scale) 16 the firing order of a group of nerve fibers when a stimulus pulse is applied with a gradient potential such that it produces a stimulus amplitude on each nerve fiber different from that on its adjacent fibers.

Figure 1B provides a two-dimensional plot showing one channel of the gradient probe and the field gradient. Figure 1B shows the location of active and ground electrodes 20 and 22 for one stimulus channel. The active electrode and any associated ground electrode create a field gradient between the electrodes and at a short distance above the electrodes. The gradient probe is placed so that the nerves to be stimulated are either between the electrodes or directly next to the electrodes. Figure 1B also shows the gradient of the field 24 in terms of voltage. The voltage range shown is given as an example and does not represent the actual voltage used. Those skilled in the art will appreciate that nerve fibers can be stimulated with very low voltages such as 100 millivolts.

16-CHANNEL STIMULATOR ASSEMBLY

A 16-channel system using a repetition rate of 200 Hz produces a channel cut-off frequency of 3,200 Hz. It requires an average of 7.5 nerve fiber currents per channel to achieve a neural carrier frequency of 24 kHz. Amplitude modulation of the stimulus pulse (to achieve sound perception) is converted into frequency modulation of the neural carrier frequency. Precision is not required in following the strength-duration curve, considering that only a small part of the strength-duration curve is used. With that number of channels, the streaming time per channel is only 0.25 ms and the slope change in that time is small.

Figure 2 is a schematic view of the clock transmitter, modulator and signal generator. Its function is to generate a sequence of two-phase pulses and a corresponding modulation of part of the two-phase pulses.

On the left side of Figure 2 are the terminals (connectors) 30a and 30b for the +6 volt and -6 volt power source. Both + and - terminals have capacitors C1 and C2 to ensure stable voltages without drifting due to changes in current demand. Schmitt trigger IC1 (1/6 of 74C14) together with R1 and C3 form an oscillator, which provides the clock (rhythm) of the system. The values of R1 and C3 determine the clock frequency. The output from the Schmitt trigger drives (integrated circuit) IC2, a 4-bit binary counter (74C93). That counter divides the clock signal by 16.

The two least significant digits of the output from counter IC2 send the least significant inputs (input data) of 4028 BCD (binary coded decimal digits) to the decimal decoder IC3. The 2 most significant inputs are grounded making the 4028 effectively a 2 digit binary decoder. Four output signals representing logic values 0 to 3 (4 separate states) drive the 4001 Quad 2-input NOR circuit IC4. The 4 outputs of the NILI circuit are two-phase signals that rotate between the +6V and -6V power supplies and are phase-shifted by 90° from each other (see Fig. 2). These outputs drive 220K resistors R2, 3, 4 and 5. These outputs will be summed with the modulation signal.

In the bottom right of Figure 2, the modulation input goes through a 10K resistor R14 to input one half of the 4052 dual 4-channel analog multiplexer IC5. The channel select input signals of integrated circuit IC5 are paralleled with the inputs of 4028 IC3. The outputs from IC3 are modulation signals delayed by 90° from the output from IC4. The outputs of IC5 drive 220K resistors R 6, 7, 8 and 9 which are summed with the outputs of R 2, 3, 4 and 5. Timing (clock) selection causes modulation to be imposed on only the second half of the two-phase pulses.

R14 is a resistor in series with the modulating signal. It provides a 10K input resistance to IC5 from the modulation source.

The outputs from the 4052 also go through resistors R 10, 11, 12 and 13 to electrical ground. Their function, including R14, is to keep the impedance roughly constant at the input to resistors R 6, 7, 8, and 9, so that the value of the summed outputs remains largely independent of which channel IC5 has selected. There are three ways out of that circle. At the bottom left is the 4-digit ADDRESS BUS (ADDRESS BUS) to which the outputs from IC2 arrive. On the right side is the SIGNAL BUS, which contains 4 phase-shifted two-phase analog stimulus signals including their modulation, and in the middle below are signals for the management (control) of the linear increase. In the lower part to the right of the middle are the LINEAR INCREASE CONTROL (RAMP CONTROL) outputs.

Figure 3 is a schematic view of the sawtooth signal generator. On the left side of the picture are the three control signals coming from the linear gain control circuit of Figure 2. Line A excites the control line of the bidirectional switch 4066, circuit IC7 and through the converter IC6 goes to the second control line of circuit IC7. Phase A and Phase B signals coming from Figure 2 charge capacitors C4 and C5 through resistors R32 and R15. When the IC7 switches are open, there is a linear rise in voltage. IC8 and IC9 are operational amplifiers each having a positive gain of 11 due to negative feedback through resistor networks R16 and R17 and R18 and R19. Resistors R20 and R21 introduce positive feedback which causes a linear gain to C4 or C5 to produce a linear gain at the output of the amplifier with a rising curve similar to the power-duration curve. R22 and R23 sum these sawtooth output signals from IC8 and IC9 on the SIGNAL BUS with signals A and B from Figure 3. IC10 and IC11 are operational amplifiers connected with a gain of -1. This is determined by resistors R24 and R25 and R26 and R27. Resistors R28 and R29 reduce the DC deflection by keeping the two inputs of each amplifier at the same impedance. R30 and R31 sum the output signals from IC10 and IC11 to lines C and D of the SIGNAL BUS. Resistors R22 and R23 sum the output signals from IC8 and IC9 to lines A and B of the SIGNAL BUS.

Figure 4 is a 16-channel multiplexer. The ADDRESS BUS lines from Figure 2 are connected to the address bus input in the lower left part of the picture. The ADDRESS BUS drives the address input lines of circuit IC12, which is a 74C154 4-line to 16-line decoder. The output from the decoder IC12 has 16 lines (0 to 15) which are cyclically switched on one after the other. The first four lines O to 3 go to the input of the 4-input NI logic IC13A (1/2 of the 4012 dual 4-input NI-circuit IC13). The output from the NI circuit IC13A is high in the first 4 positions of the 16-line decoder. Similarly, lines 1 to 4 of circuit IC12 go to the inputs of NI circuit IC13B. Its output will be delayed by one counter beat from circuit IC13A and so through the eight integrated circuits 4012, IC13 to IC20. It should be noted that this is done cyclically, so that the output from logic circuit IC20B starts one clock pulse before the output from IC13A. These outputs which last 4 clock pulses and are overlapped by three clock pulses from the triple 2-channel analog multiplexer integrated circuits 21A, B, C to control the adjacent channel via IC26A.

The multiplexer, when turned on, selects the two-phase signal A to connect to its output when turned on and the ground signal when turned off. Similarly circuit IC21B applies signal B, IC21C applies signal C, IC22A applies signal D, IC24B applies signal A and so on to IC26A and then back to IC21A. When the switches are not connected to one of the signal lines A, B, C, D, they are connected to ground (ground) to prevent interference (crosstalk) and leakage currents. The outputs of the multiplexer switches have a low impedance and provide a voltage output. Resistors R33 to R48 change the voltage output into a current to drive (excite) the nerve probe. The value of their resistance is generally higher than the resistance of the electrodes of the nerve probe, which ensures that nerve stimulation is relatively independent of the resistance of the probe on the way to the nerve.

Fine adjustment of the output current can be done by changing the supply voltage, which has little effect on the clock frequency, or by placing parallel connected variable resistors on each signal line A, B, C and D and grounding the signal, which will reduce the voltage deviation from these points and thus reduce the current inflow in the nerve. Each variable resistor only affects 1/4 of the 16 outputs, so there are 4 variable resistors in the group.

Figure 5 shows the waveforms of the 16 channels. As shown in the figure, each channel is delayed by 90 degrees from a full pulse cycle. The amplitude of the stimulus pulse is adjusted so that the current (or trigger) starts at the beginning of the slope compensation. It should be noted that slope compensation can occur at the beginning of the stimulus pulse (not shown in Fig. 5 ) or in the second part of the stimulus pulse (as shown in Fig. 5 ). In any case, the modulation does not occur until the current is in the right place.

Figure 6A shows the intensity-duration curve 40 with the stimulus pulse 42 intersecting the intensity-duration curve and compensation added to the stimulus pulse to obtain a constant firing rate of the nerve fibers. Figure 6B shows the audio modulation 44 added to the last period of the stimulus pulse. Both images show the stimulus in the positive direction. This is for simplicity only and does not necessarily represent the corresponding polarity of the stimulus pulse.

Figure 7 is a drawing of a 16-channel probe. It connects to the channel outputs of Figure 4. Figure 15A shows a 4-channel probe. Its performance is common for 4 channels of a 16-channel probe.

4-CHANNEL STIMULATOR ASSEMBLY

A 4-channel analog system is the configuration with the minimum number of channels recommended. Therefore, he requires that the recommended maximum possible number of active nerve fibers and the recommended maximum compensation of the strength-duration characteristics of the nerves be stimulated per channel. To achieve a neuron carrier frequency or streaming frequency of 24 kHz, assume a repetition rate of 200 pulses per second times 4 channels results in a total of 800 channel stimulus pulses per second. If 30 nerve fibers are activated in a uniform sequence per channel, a carrier neural frequency of 24 kHz will be obtained. Amplitude modulation of the stimulus (to achieve the sensation of hearing) is then effectively converted into frequency modulation of the neural carrier frequency (with firing, more or less than 30 nerve fibers per impulse).

Figure 8 is a block diagram of a 4-channel system. On the left side is the microphone to which the audio amplifier and limiter/automatic gain regulator are connected. The output signal of the audio amplifier is fed to an external transmitter. The internal receiver drives the clock modulator, the stimulus generator, and the amplitude-duration sawtooth generator.

The output from this module is sent through an attenuator to a 4-channel multiplexer. The output from the multiplexer then drives the probe.

Figure 9 is a circuit of the audio preamplifier, clock generator, and sequencer. In the lower left of the picture is the clock oscillator 60. IC27 is one section of the 4093 used as a Schmitt trigger. R48 and R49 provide feedback to the input. C6 together with the sum of R48 and R49 determine the clock frequency. R49 enables adjustment of the clock frequency. The output from the clock generator is fed to the input of the 4-bit 74C393 counter IC28 and also to the 62 sawtooth signal strength-duration (SD) generator. The output signal from circuit IC28 is fed from 4 lines to 16-line decoder IC29. On the lower right of the picture is the wiring of a 4093 quad-wire 2-input Schmitt trigger Nl circuit wired to form set/reset circuits. In the upper left of the picture are 6 4093 circuits, IC30 to IC35, wired as shown to form the set/reset circuits. The first set/reset circuit IC30A is set to the O position of the output circuit IC29. The second set/reset circuit IC30B is set to position 4, the third circuit IC31A to position 8 and the fourth circuit IC31B to position 12. Similarly, the first set/reset circuit is again returned to position 5, the second circuit is reset to position 9 , the third circuit is reset to position 13 and the fourth circuit IC31B is reset to position 1. The output from these four set/reset circuits is routed through resistors R50, R52, R54 and R56 for summing with audio modulation.

In the lower center of figure 9 is the microphone input 64 which drives the closed loop amplifier IC36 with a gain of approximately 80. The ratio of R58, the microphone impedance and R59 determine the gain, the AC capacitor C7 couples to the next stage via the volume control R60. The value of C7 can be chosen to achieve pre-amplification. IC37 adds an additional gain of 3 which is determined by the values of R61 and R62. The output from amplifier IC37 is an audio signal. It is fed into 4 single pole (SPDT) CMOS (complementary MOS-metal-oxide semiconductor) switches IC39B, IC39C, IC40A and IC40B. Set/reset circuits IC32A, IC32B, IC33A and IC33B control these switches. The switching arm is connected through resistors R51, R52, R55 and R57 so that it is summed with the channel stimulus output signals. The set/reset circuits IC32A, IC32B, IC33A and IC33B perform the time regulation of the switches. Circuit IC32A sets on circuit IC29 position 1 and resets (resets) position 5, IC32B sets position 5 and resets position 9. IC33A sets position 9 and resets position 13. IC33B sets position 13 and resets position 1 .The summed outputs for channels 1 to 4 are fed to CMOS switches IC38A, IC138B, IC138C and IC139A. The output of these switches selects either the summed outputs of channels 1 to 4 or the ground reference level.

In the lower left part of Figure 9 are 4 output signals 66 that go to the strength-duration compensation circuit. These come from IC41 and IC42. IC43 -74C08, which is an l-circuit, controls their setup and reset times as follows. RESET sets CH1 to positions 0 and 6 (a low output sets the hold circuit) and resets to positions 5 and 9. Similarly, RESET CH2 sets to positions 5 and 9 and resets to positions 8 and 14. RESET CHS sets to positions 8 and 14 and resets to positions 1 and 13. RESET CH4 sets to positions 1 and 13 and resets to positions 0 and 6.

Figure 10 is a schematic representation of one of the four channels of the strength-duration compensation circuit. In the lower left part of the picture come the input signals from the clock generator reset signals from Figure 9. The clock signal enters the clock input of 74C393, IC45 and the clock input of 74C174, IC46. Circuit IC45 is held at zero by a reset/clear signal. When the reset signal stops, the counter starts counting down. IC46 prevents timing errors in the countdown by capturing the counter value after any ripple (ripple) has settled. The output from register IC 46 drives two digital-to-analog converters that include 4053 CMOS switches IC47, IC48, IC49 and IC50. These switches excite two binary chain-connected networks LR1 and LR2. In the upper left part of Figure 10, the input signal from channel 1 from Figure 9 is fed to the decoupling amplifier IC50, which drives the first circuit switches IC47 and IC48. The output from LR1 is fed to decoupling amplifier IC51, which drives the reference of the other DAC switches IC49 and IC50. In this way, a rectangular curve is created as a mirror image of the strength-duration curve. That curve generated by the channel 1 signal also has the audio modulation generated in Figure 9. A second decoupling amplifier l C 54 sends the channel 1 signal through a 200K resistor R63 to the output of the second analog-to-digital amplifier. The outputs are summed and sent to decoupling amplifier IC52. This power-duration curve compensation circuit is repeated for each of the four channels as shown in Figure 11.

In the lower central part of Figure 11 there is a block called LATENCY LOGIC 70 (LATENCV LOGIC). When it is activated, the impulse to each channel is shortened to the resting time (latency), the time before the first nerve fiber in the channel is activated. It enables the adjustment of the stimulus amplitude of each channel so that it has an appropriate waiting time and is used to establish the strength-duration characteristics for each nerve group during the test. Figure 12 is a schematic representation of the LATENCY PERIOD LOGIC. The signals from Figure 9 lower right drive circuit IC53 which is a 74C08 l-circuit. 100K resistors R64, R65, R66 and R67 ground one input of each l-logic circuit. When SW1 is in the OFF position, a strong signal is applied to IC54, which turns on the output of IC54, 74C32. The output from IC54 drives the analog switches IC55 and IC56 to select ground and the corresponding output of the signal channel or to enable pass through all channels.

Figure 13 is a circuit that adjusts the amplitude of each channel. IC57 is a quadruple Schmitt two-input logic circuit that, in combination with R69 and C7, generates a clock pulse. The clock pulse is enabled either when the UP switch SW2 or the DOWN switch SW3 is activated. When the down switch is activated, a weak signal drives the up/down counters IC58 and IC59 to count down. In this way, the counter will count up or down according to the command. Out of the possible 8 bits from the up/down counter, only 6 bits are used. Diodes D1 to D8 are used to prevent the meter from running away. When the count has reached the end of the scale, it will not be switched to O, and likewise when counting down to O, the counter will not return to full scale. The output from counters IC58 and IC59 drives four digital-to-analog converters (DAGs) configured as a current-duration compensation circuit. This circuit actually replaces the quad potentiometer. The inputs to the four digital-analog converters are the outputs of the four signal channels of the logic circuit of the latency period from Figure 12.

Figure 14 shows the waveforms for each of the 4 channels along with their modulation component. That figure shows an overlap of 0.2 or 0.3 milliseconds between channels and the addition of modulation and compensation of the strength-duration characteristics of the nerve fibers throughout the period except for the first 0.2 or 0.3 milliseconds of each channel. The stimulus pulse of each channel is biphasic to avoid introducing a DC component into the system. Modulation is only in part of the nerve firing impulse. Since the modulation over time has an average value of O, it is not necessary to modulate the part of the stimulus pulse that follows the modulation. The portion of the stimulus pulse that follows the modulation is inverted and results in an average DC value of 0. What is not shown in this figure, but is typically shown in Figure 5, is the background state of neural activity when no sound is present. This background state, when modulated, results in the perception of sound.

Figure 15A shows a four-channel probe 80. The conduction zone of the electrodes is vertical and creates a field gradient between the electrodes and provides maximum surface area for each electrode. In this way, the contact surface is not determined by the proximity of the electrodes and the correct field gradient is achieved.

The distance between the grounding electrodes can be as small as 20 microns and over 200 microns. Hair cells are usually 10 microns apart. A gap of 20 microns allows only 1 row of hair cells between the active electrode and its corresponding ground electrode. In any case, if the probe is tilted so that the hair cells are separated between the active and ground electrodes, current will occur. Even though the hair cells are non-functional in a completely deaf person, the location of the hair cells is an indication of the location of the nerve fiber endings that transmit the sensation of sound to the brain. Behind the grounding and the active electrode is a layer of insulation. This is to avoid creating an electric field from the back. Behind the insulator is an additional grounding plate to increase the maximum grounding area. The maximum ground area is desirable because it reduces the contact impedance of the ground and the conducting fluid and provides a field gradient that minimizes interference (crosstalk) of the channel. Also, not shown in the figure, insulating material can be placed on the sides of each channel to prevent current from flowing outside the sides.

Figure 15B shows a four-channel probe placed near the spiral ganglion. As shown in Figure 15A, the electrodes are installed perpendicular to the field gradient. The diameter of the probe at its largest cross-section is about 2 mm, and the example of each electrode at its thickest point is about 1 mm. The electrodes of the probe are embedded in the elastic structure of the insulation, allowing the probe to be shaped according to the shape of the Sca/a Tympany or Sca/e Vestibule. The length of the surface of the probe that closes the electrodes is between 2 mm and 8 mm. In a system with more than 4 channels, the distance between the ground electrodes would be smaller.

DIGITAL SYSTEM

The limitation of this system leads to a system where only one nerve fiber per channel is stimulated and the system becomes purely digital. In order to achieve a carrier frequency of 24 kHz in a digital system, where the firing rate of a single nerve fiber is 200 impulses per second, 120 channels are required (120 x 200 = 24,000). In this digital system, instead of tracing the amplitude-duration curve for each channel, as is done in analog systems, a large-amplitude stimulus pulse is used to locate the firing of a nerve fiber in the sloping (steep) part of its amplitude-duration curve. This minimizes the differences in activation time between channels. The electric field is limited to only 1 or a small constant number of nerve fibers firing simultaneously, creating a neural carrier frequency of 24 kHz. In this digital system, modulation is performed by frequency modulation of the carrier frequency. No modulation of the stimulus pulse amplitude is required. It is evident from this that in a system that is in the transition zone between analog and digital, both modulations can be used advantageously, i.e. amplitude modulation and frequency modulation.

Figure 16 is a block diagram of the digital system. On the left side of the picture is the external unit 102. It consists of a microphone, an audio amplifier, a limiter/automatic gain controller, an oscillator, a frequency modulator, a loop antenna (in the form of a loop) and a power supply. In the central part of the picture is the indoor unit 104, receiving loop antenna, diodes for securing both positive and negative voltage, voltage regulators and counter/decoder 128 positions and individual circuits for setting/resetting each channel. On the right side of the image is probe 106, which is placed near the nerve fibers that conduct sound to the brain.

Figure 17 is one version of the outdoor unit. On the left side of the picture, the condenser microphone M1 is connected to the input of the Schmitt circuit IC62 with feedback through resistors R72 and R73 which form a frequency modulated oscillator. R73 is variable to adjust the center frequency. The output from the oscillator is divided by 2 in IC63, which is a 74C93 counter. The output from IC63 is temporarily stored in IC64 to prevent capacitive loading on the counter output. The output from the buffer IC64 drives circuit IC65 which is a MOSPOVVER driver circuit in the half-bridge design of Si 9950. The output from IC65 is a low-impedance switch that switches from one power line to another. The output is routed through C9 to the frame antenna L1. Capacitor C9 is in resonance with the inductance of the loop antenna forming a tuned circuit. Capacitor C8 is connected in parallel to the power terminals of IC65. The layout is such that there is a minimum area within the current path, considering that the resonance currents through the loop can be high, and any area outside the frame antenna will reduce efficiency.

Figure 18 is the internal unit of the digital system. In the upper left part of the picture is the frame antenna L2 and its tuning capacitor C10. The output signal from the loop antenna goes through diodes D8 and D9 to produce + and - voltages. Capacitors C11 and C12 filter DC voltages. Voltage regulators VR1 and VR2 regulate voltages. C13 and C14 ensure the stability of the output of the voltage regulator. R74 is connected to the output of the loop antenna to provide the clock for the implanted system. Diodes D10 and D11 limit the voltage swing at the input of IC66, the Schmitt trigger. IC64 sends a clock signal to the decoder of the 7 bit counter IC67. The 128 counter positions of IC67 (0 to 127) drive the set/reset circuits IC68 to IC95 which form 74COO CMOS logic circuits.

As shown in the lower left part of Figure 18, these circuits are set to a given number N, with the first circuit resetting at N+4 and the second circuit resetting at N+8. The following pairs of set/reset circuits are moved down one number as shown in the drawing. The outputs of the first pair of set/reset circuits are fed to the 4053 IC196A. When the switch IC196A is in the OFF position, the output from the switch is at ground potential or zero potential. When the switch is in the ON position, the output of the first circuit is connected to the output. This will occur in one full cycle of the first circuit. That way, the average potential of a given channel will be 0. The output from those switches has their voltage changed to current through resistors like R73 on channel 1 and also through capacitors like C15 to further ensure that there is no long-term DC voltage component.

Figure 19 shows the waveforms of individual channels of the digital system. The circles on the waveform indicate the time when the nerve fires. The delay between channels is 40 microseconds. The amplitude of each stimulus pulse is such that the excitation of the nerve fiber occurs in the oblique part of the strength duration curve after 120 microseconds and that the stimulus pulse lasts after the excitation of its nerve fiber, in this drawing, for 160 microseconds. The repetition rate of each channel is 5.12 milliseconds or about 5 times the constant, which allows regeneration (recovery) of the nerve by about 1% of its initial state before the stimulus. In all systems, both analog and digital, a carrier frequency of approximately 25 kHz is used as an example only. Higher frequencies would require a higher stimulus amplitude which would cause a higher current velocity and lead to a higher modulation response frequency. It is also accepted that, as in any carrier frequency system, audio modulation must be limited to less than 1/2 the carrier frequency.

PROCEDURE

The procedure for implanting the device in question in the human ear is as follows:

The placement of the gradient probe is critical and requires precise placement for optimal performance in the application. In an adult patient who was able to hear in the past, the probe is placed while the patient is awake. Using local anesthesia, the activated probe is moved into place and the patient indicates when he hears a clear sound. Then, for fine-tuning, each channel is stimulated separately to determine that each channel requires approximately the same stimulus intensity to achieve sound perception and to confirm the intensity-duration characteristics. The patient is again asked to confirm that their hearing is normal and the probe is then permanently fixed in place. During this procedure, a fastening element placed on the patient's head is used to hold the probe so that the movement of the head does not affect the position of the probe in relation to the nerve fibers.

In patients who have had no hearing in the past, the probe is located as previously stated, except that the last step involves confirming that a minimum of sound is heard when there is no external sound and that the sounds are pleasant when produced externally.

For children who cannot directly assist with placement, other means of establishing higher-level neural activity or brain activity can be used. One way is to measure nerve activity in the cochlear apparatus.

The procedure according to this invention directly stimulates the nerve fibers of the audio transmission part of the 8th nerve by means of electrical signals typical of the received audible sounds in a sequence, thereby enabling the hearing sensation for the deaf patient. The procedure involves the implantation of a receiver with a loop antenna and connections to an electrode probe, which consists of a grid of electrodes shaped to produce multiple field gradients in the patient and produces an electrical signal typical of the audible sounds being received. The electrical signal is divided into time-multiplexed channels, where each multiplexed channel is connected to a corresponding gradient probe channel and contains an audible representation of the entire audio spectrum and means for limiting the audio spectrum. Each channel is processed to produce a biphasic stimulus signal that overlaps (in time) with that of its adjacent biphasic stimulus signal, but does not overlap with its component representing the audible sounds.

The procedure further includes matching the external transmitter/receiver frame antenna with the receiver/transmitter frame antenna of the implant, so that the outer antenna is separated from the inner antenna by a distance equal to at least the thickness of the patient's skin, which enables both power transmission and the transmission of sound displays to the implant receiver, and locating gradient probes near the nerve fibers that transmit the sensation of sound to the brain. In addition, the procedure includes compensating for the patient's head motion during placement of the gradient probe; and permanent fixation of the gradient probe.

Given that some embodiments of the subject invention are presented here, it is understandable that experts in this field can make changes and modifications, therefore it is intended that the related patent claims cover all such changes and modifications within the scope of the subject invention. The previously described circuits show standard, well-known components. This was for the purpose of giving examples of how certain functions can be achieved. This is not to suggest that programmable logic arrays, microprocessors, or other components are excluded from application. In fact, devices such as Intel's 8XL51FX commercial/high-speed low-voltage CHMOS single-chip 8-bit microcontroller are excellent choices for implementing the functions of this invention. When operating at a clock frequency of 3.5 MHz, it draws less than 6 mA.

Figure 22 shows a block diagram including the application of all these devices. In the upper left part of the picture is a sketch of the external module 120 that will be worn by the patient. It is adapted to be worn behind the ear in a manner similar to conventional hearing aids. It houses a high-energy battery as a power source. When not in use, this external module can be placed in the charger shown behind the module. The external module has a volume control, a zero background tone adjustment, a power switch and a microphone located in the external unit behind a small opening. The antenna coil is connected to the device via a small cable and mounted directly next to the receiving coil of the implant. To the right of the external module assembly is a block diagram of 122 circuits. The microphone drives the auto gain control circuit. The volume control provides the input signal to the analog-to-digital converter. Above the automatic gain controller (AGC) block are the up and down controls that drive the adder and subtracter (up/down) counters. It is a fine-tuning (adjustment) of the stimulus level. The sum of this value and the output from the digital-to-analog converter (DAG) is formatted and sent to the modulator, which adds the data to the output signal of the oscillator and sends it to the antenna.

Below the external module and charger, the figure shows a sketch of the doctor's office computer module 124 which consists of a laptop computer with a microphone, a transmitter/receiver and includes software support to perform the following functions:

1) Establishing a two-way communication link between the external computer and the implant computer,

2) Setting the channel selection speed (or channel frequency), stimulus amplitude for each channel individually and in groups,

3) Adjusting the strength-duration curve compensation. This adjustment is in the form of a computer table. It adjusts the curve of the time constant, which can be fine-tuned at all points of the curve in order to cancel the variations of the carrier frequency of the nervous activity. This curve is related to the adjustment of the stimulus amplitude since its values are expressed as a percentage of the stimulus amplitude. The time axis of the table is independent of the channel selection speed.

4) Adjusting the audio level (coarse) and soft start speed... The soft on/off function encourages the background carrier frequency to gradually turn on or off with minimal disturbance to the user.

5) Setting the number of channels to be used. Four is the minimum and eight is the maximum. This function also selects which probe channels will be used. The doctor first searches (scans) each channel and establishes its functionality with regard to sensitivity and strength-duration characteristics. In the event that 4 channels are selected for use, then the doctor selects which probe channels will be used. All unused channels are grounded to common return electrodes. In the case of a 6-channel system, the 6 best probe channels are used. During this selection procedure, the electrode matrix switch remains on the electrodes of the selected channel, and all other channels are grounded. Figure 22C shows the selection of 4 channels of an 8-channel probe.

Figures 22A, 22B and 22C are representations of various functions on a computer screen.

In the middle part of Figure 22 is a block diagram 130 of the implantation module. To the far left in the block diagram is the receive/transmit loop antenna. It sends a signal to the receiver that develops a voltage that is then regulated and provides protection against element failure, amplification and reduction of the power of the built-in microprocessor, high voltage for writing in non-erasable (permanent) memory and soft start and soft shutdown of the probe stimulus. The second output from the receiver is a serial output for data transmission. From it, the signal is taken to the data decoder and then to the microprocessor. The programmed function values of the microprocessor are set in non-volatile memory. Eight digital-to-analog converters (these can be either of the analog type or can be performed by pulse width modulation, given that the stimulus below the threshold acts in an additive way and has the effect of increasing the ability to excite the membranes of nerve cells) to drive the eighth power source, of which four to eight. The outputs from these power supplies produce both positive and negative currents to achieve an average zero DC potential. These current source (power) channels are then selected by an electrode matrix switch to transmit through resistors, reducing the possibility of a residual DC component, into the probe electrodes.

As those skilled in the art will appreciate, the subject invention may be carried out in many different specific ways, and the specific details disclosed herein are merely exemplary of the practical application of the invention. For example, many alternating block and circuit diagrams (circuit diagrams) are equivalent to the diagrams shown in the example drawings and may be used to practice the present invention. Likewise, the necessary electrical circuits can be manufactured as integrated circuits or they can be embedded in a microprocessor, or a combination of integrated circuits and microprocessors can be used. In fact, a microprocessor, because of its programmable capabilities and small size, may be a particularly advantageous device for use in the present invention.

it is easily clear that the invention disclosed here was conceived and developed in such a way as to fulfill the previously set tasks, it is valuable that numerous modifications and implementations can be based on the patent claims that cover it, all of which modifications and implementations fall within the scope and spirit of this invention.

Claims (32)

1. Device for giving electrical stimuli to any branch of the 8th nerve, which includes: a stimulus generator for applying electrical stimuli to nerve fibers of the 8th nerve during at least two temporal channels to induce or enhance a stream of neural activity at a substantially constant rate, independent of audio modulation, capable of acting as a carrier wave and perceived by the brain as active silence . 2. The device according to Patent Claim 1, characterized in that the stimulus generator exerts electric fields on at least two groups of nerve fibers to cause currents of substantially uniform speed in the fibers in any group. 3. Device according to Patent Claim 2, characterized in that the stimulus generator acts on each group of fibers with a time-varying electric field thereby compensating the strength-duration characteristics of the fibers in the group to cause the fibers to flow in the group at a substantially uniform speed. 4. Device according to Patent Claim 2, characterized in that the stimulus generator acts on each group of fibers with a spatially variable electric field thereby compensating the strength-duration characteristics of the fibers in the group to cause the fibers in the group to flow at a substantially uniform speed. 5. Device according to Patent Claim 2, characterized in that the stimulus generator includes: multiple electric field generators, each of which is placed in the immediate vicinity of a corresponding group of nerve fibers for the effect of the electric field on the said adjacent group of nerve fibers, and a signal generator for providing a signal to the electric field generators to cause the electric field generators to exert an electric field on the nerve fibers to cause or enhance a flow of nerve activity at a substantially constant rate. 6. The device according to Patent Claim 1, characterized in that the stimulus generator includes means for modulating the carrier wave to create a sound sensation. 7. Device in accordance with Patent Claim 6, characterized in that: each acts during one of the time channels on the associated group of nerve fibers; each time channel includes a latent period (rest period) and a streaming period; when one of the stimuli is imposed on an associated group of nerve fibers during one of the time channels, the fibers in the associated fiber group conduct at a substantially uniform rate during the current period of the time channel to produce or amplify a carrier wave, and the means for modulating the carrier wave modulates said one electrical stimulus during the streaming period of the mentioned one of the time channels. 8. Device according to Patent Claim 2, characterized by that each stimulus is given to one of the fiber groups during one of the time channels; that each time channel includes a latent period; that when one of the stimuli is given to one of the groups of fibers during one of the time channels, said one of the stimuli includes compensation of the strength-duration characteristics of the fibers in said one group to cause the fibers in the group to flow at a substantially uniform rate, and until said compensation of the strength-duration characteristics comes after the latent period of one of the time channels specified. 9. The device according to Patent Claim 1, characterized in that the stimulus generator acts with electric fields on a large number of groups of nerve fibers in order to achieve that the fibers in each group fire mostly simultaneously. 10. The device according to Patent Claim 1, characterized in that the stimulus generator provides stimuli to a number of groups of nerve fibers in order to cause the flow of nerve activity. 11. The device according to Patent Claim 10, characterized in that the stimulus generator applies electric fields to groups of fibers to induce the flow of fibers in each group at a substantially uniform speed. 12. Device in accordance with Patent Claim 11, characterized in that the stimulus generator acts with electric fields on each group of nerve fibers during a corresponding one time channel and in a certain sequence, and characterized in that time-adjacent channels overlap in time to compensate for the latent period in nerve firing fibers. 13. The device according to Patent Claim 12, characterized in that the pairs of time-adjacent channels overlap in time for a substantially constant length of time. 14. Device according to Patent Claim 12, characterized in that during a certain time cycle, the stimulus generator acts with electric fields on a certain number of fiber groups; and yes in such a determined time cycle, the stimulus generator does not act with electric fields on each group of fibers in the remaining time period, in order to provide the nerve fibers in each group with time for recovery (regeneration). 15. Device according to Patent Claim 12, characterized in that during a certain time cycle, the stimulus generator acts with electric fields on a given number of fiber groups; and yes in such a determined time cycle, the stimulus generator acts with electric fields on all of the stated given number of fiber groups during mostly the same time period. 16. Device for providing electrical stimuli to any branch of the 8th nerve, which includes: stimulus generator for providing electrical stimuli to a number of sets of nerve fibers of the 8th nerve during a number of time periods channels, in order to achieve that fibers in each set fire substantially simultaneously and to produce a stream of neural activity at a substantially constant rate, independent of audio modulation, capable of acting as a carrier wave and perceived by the brain as active silence. 17. The device according to Patent Claim 16, characterized in that the amplitude of the electrical stimulus is such that each set of fibers fires within approximately 120 microseconds after one of the stimuli is given to the set of fibers. 18. Device according to Patent Claim 16, characterized in that the stimulus generator acts electric fields on each set of nerve fibers during the corresponding one time channel and in a certain sequence, and characterized in that temporally adjacent channels overlap in time to compensate for the latent period in triggering nerve fibers. 19. Device for providing electrical stimuli to the branch of the 8th cranial nerve, which includes in combination: means for stimulating a number (N) of different groups of nerve fibers of the 8th cranial nerve, wherein said groups of nerve fibers are phase-shifted in N spaced intervals; and means for repeatedly applying electric fields to groups of fibers for the purpose of inducing or amplifying a constant flow of nerve activity, independent of audio modulation, which is capable of acting as a carrier wave and which the brain perceives as active silence; and wherein the carrier wave, when modulated, results in the perception of sound; and whereby, for each group of fibers, an interval is ensured between the action of electric fields on the group of fibers, where the specified interval is not less than the natural recovery time of the nerve fibers in the group. 20. The device according to Patent Claim 19, characterized in that the electric field means act on each group of fibers with a variable electric field, thereby compensating the strength-duration characteristics of the fibers in the group to cause the fibers to flow in the group at a substantially uniform speed. 21. The device according to Patent Claim 20, characterized in that the means for acting on electric fields act on pairs of groups of fibers with electric fields in one overlapping period, in order to compensate for the latent period in the firing of nerve fibers. 22. A method for providing electrical stimuli to a branch of the 8th cranial nerve, comprising using a stimulus generator to provide electrical stimuli to nerve fibers of the 8th cranial nerve during at least two time channels to induce or enhance a stream of nerve activity at a substantially uniform rate, independent of audio modulation , which is capable of acting as a carrier wave and which the brain perceives as active silence. 23. The method of claim 22, wherein the step of applying the stimulus generator includes the step of applying electric fields to at least two groups of nerve fibers to cause the fibers in each group to conduct currents at a substantially uniform rate. 24. The method according to Patent Claim 23, characterized in that the step of applying the stimulus generator further includes the step of applying time-varying electric fields to each group of fibers, thereby compensating the strength-duration characteristics of the fibers in the group for the purpose of causing the fibers in the group to conduct currents with mainly at a uniform speed. 25. The method according to Patent Claim 23, characterized in that the step of applying the stimulus generator includes the steps: placing a suitable one electric field generator in the immediate vicinity of each of the groups of nerve fibers in order to apply electric fields to said adjacent group of nerve fibers; and using a signal generator to signal the electric field generators to the electric fields so that the electric field generators begin to exert electric fields on the nerve fibers to induce a generally constant flow of nerve activity. 26. The method of claim 22, further comprising the step of modulating the carrier wave to create a sound sensation. 27. The method according to Patent Claim 22, characterized in that the step of applying the stimulus generator includes the step of imposing a stimulus on a plurality of fiber channels in order to achieve said flow of neural activity. 28. The method according to Patent Claim 27, characterized in that the step of applying the stimulus generator includes the steps: effects of electric fields on each group of nerve fibers during the corresponding time channel and in a certain sequence; and temporal overlapping of adjacent channels to compensate for the latent period in the firing of nerve fibers. 29. The method in accordance with Patent Claim 28, characterized in that in time-adjacent pairs of channels there is an overlap for a mostly constant length of time. 30. The method according to Patent Claim 28, characterized in that the step of applying the stimulus generator further includes the steps: effects of electric fields on a certain number of groups of fibers during a certain time duration of the cycle; you in such a determined time cycle, ensuring each group of fibers a period of rest, during which the electric fields do not act on the group of fibers, in order to allow the nerve fibers in each group time to recover. 31. Procedure in accordance with Patent Claim 28, characterized in that: during a certain time cycle, the stimulus generator acts with electric fields on a given number of fiber groups; and yes in such a determined time cycle, the stimulus generator acts with electric fields on the total set number of fiber groups during mostly the same time period. 32. The procedure of direct stimulation of the nerve fibers of the 8th cranial nerve with electrical signals characteristic of audible sounds in a sequence in order to provide the deaf patient with the sensation of hearing, which procedure includes the following steps: implanting in the patient a receiver with an antenna and connections to an electrode probe, which consists of a series of electrodes shaped to produce multiple gradient fields in the region of the cochlear nerve (nerve of the cochlea of the ear); generating an electrical signal typical of audible sounds; dividing the electrical signal into time multiplexed channels, characterized in that each multiplexed channel is connected to a corresponding channel of the gradient probe and contains an audible representation of the entire audio spectrum and means for limiting the audio spectrum; processing each channel to produce a biphasic stimulus signal that overlaps with the biphasic stimulus signal of a temporally adjacent channel but does not overlap with its component representing audible sounds; aligning the external transmit/receive antenna with the implanted antenna such that the external antenna is separated from the internal antenna by a distance at least equal to the thickness of the patient's skin, indicated that the external antenna provides the means to transmit both the power and the sound image to the implanted receiver; placement of the field gradient probe near the nerve fibers that transmit sound sensation to the brain; and permanent fixing of the probe to its place in the patient.