Cochlear implant system with improved electrode selection scheme

Abstract

A cochlear implant system with improved electrode selection scheme is disclosed, comprising: a microphone unit configured to receive an acoustic signal and transmit an audio signal based on the acoustic signal, a processor unit configured to receive the audio signal and process the audio signal into a plurality of electrode pulses, an electrode array comprising a plurality of electrodes configured to stimulate an auditory nerve of a user of the cochlear implant system based on the plurality of electrode pulses; the processor unit configured to: assign an importance value to one or more electrodes of the plurality of electrodes, wherein each importance value is determined based on a status of an electrode pulse assigned to a respective electrode, select a primary set of electrodes from the plurality of electrodes during a time window, wherein an importance value of each selected electrode in the primary set of electrodes is greater than or equal to an importance threshold, enable electrode stimulation of the auditory nerve by electrodes of the primary set of electrodes based on an electrode pulse of the plurality of electrode pulses, and maintain electrodes of the primary set of electrodes in a reserved mode during a reserved time period.

Description

TECHNICAL FIELD

[0001] The present invention relates to cochlear implant systems with improved electrode selection schemes. BACKGROUND

[0002] In the normal (acoustic) hearing system, one of the main functions of the cochlea is to spectrally resolve the incoming acoustic signal and encode the resulting spectral components into neural excitation of the auditory nerve. The spectral resolution and neural encoding occurs along the length of the cochlea in response to frequency and can be modeled by a filter bank. Multi-channel / electrode cochlear implant (CI) systems also perform spectral resolution and convert the resulting spectral components into electrical current pulses, or electrode pulses, which are delivered directly to the auditory nerve from electrodes inserted into the cochlea that correspond to the frequency response map. Ideally, for the same acoustic input signal, the current electrode pulse delivered by a given electrode should selectively stimulate the same population of nerve fibers as would be restored in the normal acoustic hearing system. However, in practice, only a relatively small number of electrodes are available in current implants (between 12 and 22), and these electrodes typically produce broad current fields that elicit neural responses in substantially broader clusters of nerve fibers. This results in poor spectral resolution and a large amount of overlap of excitation elicited by different electrodes. Such overlap of excitation produces significant cross-electrode interference whereby stimulation on one electrode disrupts neural excitation elicited by another electrode.

[0003] To mitigate the effects of cross-electrode interference, various CI coding strategies employ a "M-of-N" channel selection approach to limit the number of stimulating electrodes in a given time window (i.e., time epoch) to a subset (N) of the total number of available electrodes (N avail ). In such schemes, N electrodes are selected that contain the "most important" information from the potential acoustic input signal in a given time window (i.e., pulses or any spectral-temporal signal features from which the electrode pulses are ultimately derived), and the remaining electrodes are deactivated to ensure that the most important pulses are presented at the implant output with minimal interference from stimulation on electrodes that do not encode important information. The most important events are encoded with minimal interference from stimulation on electrodes that do not encode important events. Such an approach can also employ an importance threshold to discard events that do not satisfy a benchmark importance criterion prior to applying the M-of-N selection. Criteria for assessing electrode event importance can be based on, but not limited to, energy, psychophysical masking, periodicity in the underlying acoustic signal, signal coherence across channels and ears, etc.

[0004] "M out of N" type electrode or channel selection is commonly employed in sound coding strategies that generate electrode pulses or events at a constant rate over time and across electrodes. In such systems, the time window within which the M out of N selection is performed is set so that an equal number of events (i.e., electrode pulses) can occur on each electrode or channel. In this case, no electrode or channel enjoys a selection advantage at any given time due to the rate at which events occur. However, when the sound coding strategy is based on events that occur at a rate that varies over time and across channels, channels with higher event rates can enjoy a selection advantage over lower rate channels due to the fact that more events are contained in the time window in which the M out of N selection is performed. The problem described so far is addressed in the following sections for fixed and variable rate strategies, respectively. Figure 2A and 2B As shown in the figure, although modern CI systems usually use N values ​​of 8-12 for M values ​​of 16-22, Figure 2A and 2B A simplified case is described where N = 2 and M = 3. Both stimulation strategies are applied to the same hypothetical acoustic signal. During each time window of the potential acoustic input signal, the importance of the channels is the same between the coding strategies, but the events generated in response to the input signal are different. The horizontal axis on these panels refers to the time window number, i.e., the time window in which one or more channels are being evaluated and selected, the vertical axis refers to the channel number, and the value in each grid refers to the importance value of the hypothetical event. For the fixed rate strategy, i.e., Figure 2A , events occur on each channel in each time window, and the selection method can test the importance of each channel or electrode compared to other channels or electrodes in any given time window. Thus, the channel or electrode with the most important events over all time epochs is selected.

[0005] Crucially, since the selection scheme is influenced only by the importance value of the electrode or channel, channels containing unimportant / noisy information are not selected in any given time window. Figure 2B In the variable rate strategy, events occur in fewer time windows on the lower rate channel than on the higher rate channel. This leaves the high rate channel unchallenged during the selection procedure in some time windows. Thus, a standard "choose N out of M" type defaults to selecting the high rate channel whenever no events are present on the lower rate channel. Consequently, the overall coding strategy will encode events on the high rate electrode more frequently, even if those events reflect unimportant / noisy information that may be perceived by the listener and may interfere with information on the more important electrodes. This problem is particularly problematic for values ​​of N and M that are more suitable in practice than in the present example (e.g., N=8, M=20), and can be applied to sound coding strategies that derive impulses from acoustic features that are unrelated to the importance of the underlying acoustic content. Summary of the Invention

[0006] In the present specification, "channel" and "electrode" are used interchangeably, and "event" and "pulse" are used interchangeably.

[0007] An aspect of the present application is to provide a cochlear implant system targeting to overcome the drawbacks of the known solutions described above.

[0008] An aspect of the present application is to provide a cochlear implant system including an electrode selection scheme that prevents potentially unimportant / noisy information from being inappropriately utilized by higher event rates.

[0009] An aspect of the present application is to provide a cochlear implant system including an electrode selection scheme that minimizes resolution of spectral components of pulses transmitted through electrodes on the implant from an audio signal across electrode interference amounts.

[0010] The aspect of the present application is achieved by a cochlear implant system including a microphone unit configured to receive an acoustic signal and transmit an audio signal based on the acoustic signal, a processor unit configured to receive the audio signal and process the audio signal into a plurality of electrode pulses, an electrode array including a plurality of electrodes configured to stimulate an auditory nerve of a user of the cochlear implant system based on the plurality of electrode pulses.

[0011] The aspect of the present application is achieved by a cochlear implant system including a microphone unit configured to receive an acoustic signal and transmit an audio signal based on the acoustic signal, a processor unit configured to receive the audio signal and process the audio signal into a plurality of channels for generating a plurality of electrode pulses, the cochlear implant system further including an electrode array including a plurality of electrodes configured to stimulate an auditory nerve of a user of the cochlear implant system based on the plurality of electrode pulses.

[0012] The processor unit can be configured to assign an importance value to one or more of the plurality of channels or one or more of the plurality of electrodes, wherein each importance value can be determined based on a status of electrode pulses assigned to the respective channel / electrode.

[0013] The status of the electrode pulses can be a measure of a parameter / feature of the electrode pulses, which can include an estimated pulse energy level estimated within a frequency subset of a frequency range of the electrode pulses.

[0014] The parameter / feature of the electrode pulses can include an estimated signal-to-noise ratio of the sampled audio signal within a frequency subset of a frequency range of the electrode pulses.

[0015] The parameter / feature of the electrode pulses can include periodicity in the audio signal within a frequency subset of a frequency range of the electrode pulses, audio coherence across electrodes of the plurality of electrodes, and / or audio coherence across a second electrode array of another cochlear implant system.

[0016] The processor unit can also be configured to select a main group of electrodes from the plurality of electrodes during a time window, i.e. an epoch, analysis frame, wherein each selected electrode in the main group of electrodes has an importance value greater than or equal to an importance threshold value. The importance threshold value can for example be the maximum of the Nth highest electrode importance value or the absolute minimum importance value. The importance threshold value can for example be:

[0017] - an energy / charge / level value of the minimum allowable masking weighting (resulting in fewer electrodes being enabled to stimulate the auditory nerve);

[0018] - a minimum signal-to-noise ratio in the acoustic signal from which the electrode pulse is derived;

[0019] - a minimum estimated pulse energy level of the electrode pulse;

[0020] - a minimum value of the autocorrelation amplitude of the underlying acoustic signal; or

[0021] - a minimum interaural coherence value of the underlying acoustic signal received at both ears.

[0022] The main group of electrodes is selected in a first time window of a plurality of time windows, wherein the main group of electrodes or a subgroup of the plurality of electrodes can be selected during one or more other time windows of the plurality of time windows. The first time window is a precursor to the other time windows of the plurality of time windows.

[0023] The processor unit can also be configured to enable electrodes of the main group of electrodes to stimulate the auditory nerve based on an electrode pulse of the plurality of electrode pulses, and to keep these electrodes of the main group of electrodes in a "reserved" mode during a "reserved" time period.

[0024] The processor unit is configured to enable an electrode by delivering an electrode pulse of the plurality of electrode pulses to the respective electrode.

[0025] The processor unit can comprise an electrode selection scheme comprising the steps of assigning importance values to one or more electrodes, selecting the one or more electrodes to a main group of electrodes, and keeping those electrodes of the main group of electrodes in a reserved mode during a reserved time period.

[0026] In the proposed electrode selection scheme, each electrode can enter a reserved mode for a duration, i.e. a reserved time period, after an electrode pulse on that electrode has been selected. While in the reserved mode, the electrode will influence the electrode selection during subsequent (future) epochs, irrespective of whether an electrode pulse occurs on that electrode in those time windows.

[0027] The duration of the reservation mode can be constant or vary across time and electrodes, and can be statically or adaptively assigned. For example, the reservation time period on a given electrode can be statically assigned as the inverse of the channel center frequency of the electrode, or adaptively assigned as the inverse of a short-term estimate of the frequency of a key spectral feature in the acoustic signal. During the arrival of a new electrode pulse on an electrode that is already in the reservation mode, the reservation mode of that electrode can be deactivated, and that electrode will compete to be selected again by the M-of-N scheme. As an alternative, if an electrode pulse, i.e., event, arrives at an electrode that is already in the reservation mode, if it has to compete to be selected by the M-of-N scheme, the electrode can be checked whether it would be selected. If it would be selected, the electrode pulse is selected, and the electrode will enter its reservation mode again. If not, the electrode pulse can be ignored, and the electrode can keep its current reservation mode until the mode expires, i.e., the reservation time period expires. As an alternative, the electrode can keep the reservation mode until the next occurrence of an electrode pulse on that electrode. In each time window, N k electrodes with the highest importance values are selected, where the number of selected electrodes, N k may vary across time windows depending on the number of non-stimulating electrodes in the reservation mode and N k ≤ N, where N is the total number of available electrodes.

[0028] The reservation time period can be different or the same for each electrode of the main set of electrodes during the time window and / or during the upcoming time window.

[0029] The reservation time period can be equal to or longer than the time window.

[0030] The reservation time period determined for each electrode can be based on the frequency content of each electrode or electrode pulse. For example, electrodes with higher frequency content will produce pulses more often than low frequency electrodes. This would give an undue advantage to electrodes with higher frequency content compared to electrodes with lower frequency content. This is avoided by determining the reservation time period for each electrode based on the frequency content.

[0031] If a selected electrode is reserved and has an assigned electrode pulse, the reservation time period of that electrode is extended.

[0032] In the case where more than two active electrode pulses have the same importance metric value but only a subset of those electrodes can be selected, a "tie-breaker" importance value can be used to select the subset. The tie-breaker importance value must be different from the previously defined importance values, and can include: channel center frequency, pulse energy, signal-to-noise ratio, interaural coherence, periodicity, etc.

[0033] Within a time window, an electrode can be assigned an importance value and a tie-breaker importance value.

[0034] Within a time window, an electrode may be assigned a tie-breaking importance value, where the tie-breaking importance value must be different from the importance value assigned to the electrode in a previous time window.

[0035] The processor unit can be configured to assign a tie-breaking importance value to an electrode in the plurality of electrodes, wherein the tie-breaking importance value is different from the assigned importance value. The processor unit can be configured to select one or more electrodes as part of a tie-breaker, wherein the tie-breaking importance value of each selected electrode is greater than or equal to an importance threshold of the corresponding electrode. The tie-breaking importance value can include: channel center frequency, pulse energy, signal-to-noise ratio, interaural coherence, periodicity, and / or cross-electrode interference.

[0036] The proposed electrode selection scheme can be applied to unilateral or bilateral sound coding strategies, where the electrodes in the latter can be evaluated and selected by comparing the stimulation signals provided at the user's two ears.

[0037] The cochlear implant system may include a memory unit configured to store importance values ​​of one or more electrodes for a current time window and possibly for one or more (or no) previous time windows. The processor unit is configured to update the importance values ​​based on changes in the state of electrode pulses assigned to the corresponding electrodes.

[0038] The memory unit is connected to the processor unit and is configured to receive and transmit importance values ​​of one or more electrodes.

[0039] The processor unit may be configured to select an electrode subset of the main electrode group during a retention time period, and no other electrodes of the plurality of electrodes are allowed to be selected, wherein each electrode of the electrode subset has an importance value greater than or equal to an importance threshold, and wherein the processor unit is configured to enable electrodes of the electrode subset to stimulate the auditory nerve based on electrode pulses in the plurality of electrode pulses.

[0040] The processor unit may be configured to select a subset of electrodes from the plurality of electrodes, wherein each selected electrode of the subset of electrodes is assigned an electrode pulse and an importance value greater than or equal to an importance threshold, and wherein the subset of electrodes comprises a maximum number N determined by N (the maximum allowable active electrodes within a time window, i.e., an epoch) minus the number of electrodes currently having no active pulse but in a reserve mode k In other words, the electrode array or the plurality of electrodes includes a total of M electrodes, N of which k are allowed to be active during time window k, where N k Can be different across time windows.If a "soft" retention mode is employed, an electrode in retention mode will only block an electrode from delivering an electrode pulse if the calculated importance value of the event does not exceed the importance value assigned to the retention mode.

[0041] This can preferably be used to prevent high importance events from being blocked by previously entered into the reservation mode with lower importance values. The importance value of an electrode in the reservation mode can be assigned and kept constant as the importance value at the moment the channel / electrode was reserved, or can vary over time (e.g. decay over time). As an alternative, the importance value of a reserved electrode at any given time can be calculated independently of the importance value calculated at the time the electrode became a reserved electrode.

[0042] Rather than keeping the importance value of an electrode constant at the time the electrode enters a "soft" reserved state, the importance value of the electrode can be continuously updated in each time window while the electrode is reserved. During the time window when an electrode pulse occurs on the electrode and the electrode pulse has an importance value greater than or equal to the importance threshold, the electrode will be selected to enter the reserved mode. By updating the importance value of a reserved electrode in subsequent time windows, the ability of the electrode to prevent other electrodes carrying active electrode pulses from being selected will depend on the importance value of the (acoustic) content of the channel of the electrode in the current time window, rather than the importance value during the time window when the electrode entered the reserved mode.

[0043] The processor unit can be configured to select a subset of electrodes from the plurality of electrodes (i.e. including electrodes that are not part of the main set of electrodes) and / or the main set of electrodes during a reservation time period, wherein each electrode of the subset of electrodes has an importance value greater than or equal to an importance threshold, and wherein the processor unit is configured to enable stimulation of the auditory nerve by the electrodes of the subset of electrodes based on electrode pulses in the plurality of electrode pulses.

[0044] The importance threshold is determined as follows: determine a minimum importance threshold, determine a minimum importance value of the subset of electrodes and / or the main set of electrodes, if the minimum importance value is greater than or equal to the minimum importance threshold, determine the importance threshold to be equal to the minimum importance value, if the minimum importance value is less than the minimum importance threshold, determine the importance threshold to be equal to the minimum importance threshold.

[0045] For example, the importance value can be based on a noise level, and the cochlear implant system can define an acceptable noise level of 40 dB SPL. Each electrode with a noise level above this acceptable level plus a noise error margin (+3 dB) is not selected. Similarly, the system can also have a noise floor of 20 dB SPL. Each electrode with a signal level below this noise floor will not be selected.

[0046] The processor unit can be configured to replace an electrode of the main set of electrodes with a new electrode, wherein the importance value of the replaced electrode is lower than the importance threshold, and wherein the importance value of the new electrode is equal to or greater than the importance threshold. Thereby, a more important electrode pulse can be selected compared to a less important electrode pulse that was in the hold mode. The cochlear implant system becomes more flexible to sudden changes in the audio signal, e.g. sudden changes in the importance value of each electrode can occur during the hold period in noisy situations.

[0047] The processor unit can be configured to update the main set of electrodes by adding a new electrode of the plurality of electrodes to the main set of electrodes, wherein the importance value of the new electrode is greater than or equal to the first importance threshold.

[0048] The number of electrodes of the main set of electrodes or the sub-set of electrodes can not exceed the total number of available electrodes (N avail ).

[0049] The processor unit can be configured to update the main set of electrodes by lengthening the hold period of an electrode of the main set of electrodes when an event of generating a new pulse occurs on an electrode whose importance value is greater than or equal to the importance threshold.

[0050] The processor unit can be configured to update the main set of electrodes by removing an electrode from the main set of electrodes when the hold period of the electrode has expired and before the hold period is lengthened.

[0051] The processor unit can be configured to sample the audio signal in a frequency range, and wherein the status of an electrode pulse of the plurality of electrode pulses comprises a pulse energy level estimated in a frequency sub-set of the frequency range, an estimated signal-to-noise ratio of the audio signal sampled in a frequency sub-set of the frequency range, periodicity in the audio signal in a frequency sub-set of the frequency range, audio coherence across electrodes of the plurality of electrodes, and / or audio coherence across a second electrode array of another cochlear implant system.

[0052] The status of an electrode pulse can be a measured parameter of the plurality of electrodes, which can be made during fitting and / or during operation of the cochlear implant system. The pulse energy level of each electrode is determined by the processor unit alone or based on a measurement made by a sensor or electrode array. The measurement sensor is connected to the processor unit and configured to measure the stimulation provided by the electrode. The measured stimulation can comprise the pulse energy level and / or a noise estimate. The measurement signal is given to the processor unit. The measurement signal comprises, for example, the pulse energy level and / or the noise floor level. The processor unit can be configured to determine the signal-to-noise ratio of the electrode based on the measurement signal. The noise floor level of each electrode is measured by the measurement sensor when no electrode pulse is applied to the respective electrode. The periodicity in the audio signal and the audio coherence are determined by the processor unit.

[0053] The importance value of the electrodes of the sub-set of electrodes is greater than the importance value of the electrodes that are not selected.

[0054] The importance threshold can be determined such that the electrode master set and / or the electrode sub-set comprises between 2 and 5 electrodes, between 2 and 10 electrodes, between 2 and 15 electrodes, between 2 and 25 electrodes or more than 25 active electrodes.

[0055] The ideal maximum number of stimulating electrodes N within a given time frame can vary across patients and across stimulation.

[0056] For example, for patients with low amounts of cross-electrode interference, a larger number of electrodes can be stimulated without introducing significant electrode interaction, i.e. excitation spread and spatial masking, compared to patients with high amounts of cross-electrode interference. Conversely, patients suffering from high amounts of excitation spread can be stimulating too many electrodes and thus experience significant electrode interaction, reducing the power consumption of the cochlear implant system for those patients. Furthermore, if there is considerable variability in the cross-electrode interference (within a patient), the optimal number of active electrodes can actually vary depending on the specific selection of electrodes. It is therefore advantageous to determine the optimal N by:

[0057] - fitting and adjusting the number of active electrodes for the individual patient / user; and / or

[0058] - adjusting the number of active electrodes on a short-term basis depending on the selected electrodes.

[0059] For a given patient of a cochlear implant system, the excitation spread can be estimated on each electrode. This can be achieved, for example, by using an electrically evoked compound action potential (eCAP) measurement. Using standard eCAP measurement techniques, an excitation spread function can be estimated, for example, for a given electrode by stimulating on that electrode, i.e. the stimulating electrode, and measuring the ECAP response on each electrode, i.e. the probe measurement electrodes. Thus, the excitation spread function can be determined to include the measured ECAP responses as a function of the number of electrodes for a given stimulating electrode. The measured ECAP responses can be expressed as a normalized ECAP magnitude of the measured ECAP responses, which reflects the amount of spatial masking contribution each measurement electrode experiences in response to the stimulation provided by the stimulating electrode, i.e. the higher the ECAP magnitude the higher the amount of spatial masking contribution. A measure of the spread width describing the spatial masking contribution of the stimulating electrode can be defined, for example, to include the 3 dB bandwidth of the function or a multiple of the standard deviation of the function. The excitation spread function and the spread width measure can then be calculated for each electrode, and the optimal number of active electrodes can be calculated based on the excitation spread function and the spread width measure of each electrode to result in an allowed spatial masking contribution, i.e. electrode interference amount. One possible method of calculating the number of active electrodes is given by:

[0060]

[0061] where N array is the number of electrodes on the electrode array, the expression in the denominator is the average spread width of all electrodes, and round() describes the process of rounding the result of the equation to an integer value (rounding up, rounding down, or rounding to the nearest).

[0062] The method of determining the optimal number of active electrodes can be performed automatically by the implant fitting software, whereby, for each combination of stimulating electrodes and measuring electrodes, one or more eCAPs are measured, resulting in an excitation spread function for each electrode.

[0063] The excitation spread function and / or the spread width metric for each electrode can be stored in a memory unit of the cochlear implant system.

[0064] The processor unit can be configured to calculate the optimal number of active electrodes for a given audio signal based on the excitation spread function and the spread width metric for each electrode, resulting in the allowed spatial masking contribution, i.e. the amount of electrode interaction.

[0065] The optimal number of active electrodes can vary depending on the particular electrodes of the electrode array that are selected. The optimal number of active electrodes can be lower when electrodes with a wider spread width are selected than when electrodes with a narrower spread width are selected. The electrodes can be repeatedly selected within a given time window, i.e. epoch, until the sum of the spread widths of those selected electrodes exceeds a predetermined maximum allowed sum of spread widths. Other parameters such as the pulse energy of the electrodes can also be considered in calculating the sum of spread widths, weighting the spread width of each electrode by those parameters. Furthermore, a predetermined maximum sum of spread widths can be established for a particular group of adjacent electrodes of the electrode array, i.e. a particular region of the electrode array, and the electrodes of that group of adjacent electrodes can be repeatedly selected until the group allowed sum of spread widths is exceeded. In this case, the optimal value of active electrodes varies across the electrode array and locally in different groups of adjacent electrodes of the electrode array. This can have the benefit of further reducing the amount of electrode interaction occurring between stimulating electrodes of the electrode array.

[0066] The fitting software can employ a current spread model that predicts the excitation spread function on each electrode and can be fitted to the patient of the cochlear implant system by collecting a few eCAP measurements of different stimulating electrodes of the electrode array. For example, if the excitation spread of each electrode is assumed to be modeled by a particular function, such as a Gaussian function, the fitting software is configured to measure eCAPs of a subset of trial electrode combinations and then fit the assumed function using a standard curve fitting method. More complex models can also be employed. The use of these models can reduce the total time required to collect eCAP data and thus the time required to determine the optimal number of active electrodes during a clinical fitting.

[0067] The excitation spread function can include a spatial masking contribution and / or a temporal masking contribution.

[0068] The number of electrodes in the main set of electrodes and / or the sub-set of electrodes can be determined by a method of determining an optimal number of active electrodes.

[0069] The fitting software can be part of a fitting system.

[0070] The importance value is determined based on a status of the electrode pulse assigned to the given electrode. The status of the electrode pulse in the plurality of electrode pulses can be determined by estimating a pulse energy level of the electrode pulse if the frequency of the electrode pulse is within a frequency sub-set of the frequency range. If the pulse energy level is increasing, the importance value will also increase.

[0071] The status of the electrode pulse in the plurality of electrode pulses can be determined by estimating a pulse energy level of the electrode pulse and applying an importance value dependent on a frequency range in the acoustic signal from which the electrode pulse is derived.

[0072] The electrode pulse having a high pulse energy level can indicate that the electrode pulse comprises important information that can result in an improved perceptual ability of the user.

[0073] Thus, by selecting those electrodes whose importance value is greater than or equal to an importance threshold value, and wherein each importance value is determined based on a pulse energy level, the perceptual ability of the user is improved. The improved perceptual ability is provided for by the cochlear implant system being able to select those electrode pulses that deliver important information and ignore those electrode pulses that are assumed to not comprise relevant information, such as noise.

[0074] The status of the electrode pulse in the plurality of electrode pulses can be determined based on a signal-to-noise ratio of the sampled audio signal within a frequency sub-set of the frequency range, wherein if the signal-to-noise ratio of the sampled audio signal is increasing, the importance value is also increasing. The perceptual ability of the user is further improved when selecting those electrode pulses that have the highest signal-to-noise ratio.

[0075] Thus, by combining the pulse energy level and the signal-to-noise ratio in determining the importance value, the perceptual ability of the user will be further improved. By combining the pulse energy level and the signal-to-noise ratio, the ability of the cochlear implant system to select those electrode pulses that have important information is further improved.

[0076] The status of an electrode pulse of the plurality of electrode pulses can be determined based on periodicity in the audio signal within a subset of frequencies of the frequency range. The cochlear implant system is then configured to select those electrode pulses that include audio information having a certain periodicity that will result in an improved pitch perception by the user. The relationship between pitch perception and periodicity is set forth in "Periodicity and pitch perception" (Pierce, John R., Center for Computer Research in Music and Acoustics, Department of Music, Stanford University, Stanford, California 94305, PACS number 43.66.MK, 43.66.Hg [WAY]), which is hereby incorporated by reference.

[0077] The status of an electrode pulse can be determined by the coherence (maximum magnitude of cross-correlation) between the audio signals received at two or more microphones within the frequency range associated with that electrode. These microphones can be located on the same sound processor at the same ear, or distributed across sound processors at both ears (i.e., interaural coherence). Higher coherence values represent that the information conveyed by the pulse is more likely to come from a single sound source direction (i.e., a single sound source) in space, rather than multiple sound source directions. Higher interaural coherence values also represent that the binaural cues conveyed by the pulse more reflect one sound source direction than multiple directions. Thus, assigning importance based on coherence enhances the user's ability to focus on a single sound in a complex environment, and in the case of interaural coherence, to localize sounds in space.

[0078] The status of an electrode pulse of the plurality of electrode pulses can be determined based on the amount of masking imposed on that electrode pulse by other electrode pulses of the plurality of electrode pulses. For example, if an electrode pulse receives a large amount of masking from other electrode pulses, the importance value of the electrode that includes that electrode pulse will be low, or if an electrode pulse receives a small amount of masking from other electrode pulses, the importance value of the electrode that includes that electrode pulse will be high. The status of an electrode pulse of the plurality of electrode pulses can then be determined based on the amount of masking received by that electrode pulse.

[0079] The status of an electrode pulse of the plurality of electrode pulses can be determined based on the amount of masking imposed on that electrode pulse by other electrode pulses of the plurality of electrode pulses. For example, if an electrode pulse imposes a large amount of masking on other electrode pulses, the importance value of the electrode that includes that electrode pulse will be low, or if an electrode pulse imposes a small amount of masking on other electrode pulses, the importance value of the electrode that includes that electrode pulse will be high. The status of an electrode pulse of the plurality of electrode pulses can then be determined based on the amount of masking imposed by that electrode pulse.

[0080] The status of an electrode pulse of the plurality of electrode pulses can be determined based on a masking model of the cross-electrode interference imposed on the electrode pulse by other electrode pulses of the plurality of electrode pulses.

[0081] Ideally, for the same acoustic input signal, the electrode pulses delivered by a given electrode should selectively stimulate the same population of nerve fibers as in the case of a normal acoustic hearing system. However, in practice, the broad current field generated by cochlear implant stimulation evokes neural responses in a substantially broader population of nerve fiber clusters, resulting in poor spectral resolution and substantial overlap of excitation evoked by different electrodes. This overlap of excitation creates significant cross-electrode interference, i.e., masking, whereby stimulation at one electrode consumes some of the neural resources at the site of an adjacent electrode, thereby disturbing the neural excitation evoked by the stimulation. Thus, by applying a masking model approach to the determination of the importance values, the cochlear implant system will be configured to select those electrode pulses that result in a reduction of cross-electrode interference. The advantage of this is that the spectral resolution of the electrode pulses converted from the audio signal can be improved.

[0082] The status of an electrode pulse of an electrode of the plurality of electrodes can include an amount of cross-electrode interference caused by one or more electrode pulses of other electrodes of the plurality of electrodes on the electrode pulse of the electrode determined based on a masking model approach, wherein the masking model approach includes determining a spatial masking contribution of each of the one or more electrode pulses of the other electrodes on the electrode pulse of the electrode based on a spatial separation between the electrode and each of the other electrodes. The effect of the spatial separation on the masking is affected by the stimulation level, e.g., higher stimulation levels result in more spread than lower stimulation levels, and can vary between patients and between specific electrodes within a patient. The plurality of electrodes can be arranged in an order starting from electrode 1 to electrode N, the distance between electrodes being determined based on the respective stimulation levels of the electrodes. For example, an electrode pulse with a high stimulation level will have a shorter distance to an adjacent electrode than a low stimulation level.

[0083] The status of an electrode pulse of an electrode of the plurality of electrodes includes an amount of cross-electrode interference caused by one or more electrode pulses of other electrodes of the plurality of electrodes on the electrode pulse of the electrode determined based on a masking model approach, wherein the masking model approach includes determining a temporal masking contribution of each of the one or more electrode pulses of the other electrodes on the electrode pulse of the electrode based on a pulse time difference between a first time of the electrode pulse of the electrode and a second time of each of the one or more electrode pulses of the other electrodes, wherein the second time precedes the first time. The first time and the second time are defined to be within a time window, or the first time can be defined to be within a first time window and the second time can be defined to be within the first time window or within a previous time window. The amount of masking decays with increasing pulse time difference.

[0084] The masking model approach can include a determination of spatial masking contributions and a determination of temporal masking contributions. An advantage of applying the masking model approach to the determination of the importance values is that the information transmitted to the auditory nerve is maximized by including only electrode pulses that impose less cross-electrode interference and contribute to most perceptual effects and reduce power consumption.

[0085] An advantage of applying both spatial masking contributions and temporal masking contributions simultaneously is that the space itself will not take into account the order of the pulse arrival and will not take into account the time difference between them, the temporal action itself will take into account the pulse order and timing but not the electrode position difference, while having both in the masking model approach will enable a more appropriate prediction of which pulses will be masked.

[0086] For a CI user, it is important to include the temporal action because the pulses occur at discrete times. A pulse that arrives earlier will mask those that arrive later, but not the other way around.

[0087] The spatial masking contribution determined from each electrode pulse on other electrodes can be multiplied by a temporal masking decay function that includes a pulse time difference between the pulse of that electrode and each of the pulses on the other electrodes. For example, the electrode pulse can have a first time, and a first other electrode pulse can have a second time, the pulse time difference can be between the first time and the second time, and a second other electrode pulse can have a second time, and wherein the pulse time difference can be between the first time and the second time of the first other electrode pulse and / or between the first time and the second time of the second other electrode pulse. The temporal masking function can be fixed across all electrodes, can vary across electrodes, or can vary across pairs of electrodes.

[0088] The pulse time difference can be changed by adjusting the first time of the electrode pulse and / or the second time of the other electrode pulses. The adjustment of the first time and / or the second time can be based on the determined cross-electrode interference imposed on the electrode pulse and / or based on the determined cross-electrode interference caused by the other electrode pulses.

[0089] The pulse time difference can be changed by adjusting the timing between the first time window and the second time window or by changing the time length of the first time window and / or the second time window.

[0090] The changing of the pulse time difference can be performed by a processor unit.

[0091] Adjusting the timing of the electrode pulses provides a method of reducing cross-electrode interference without changing the spatial separation between the electrode pulses, which can result in acoustically unnatural signals. Thus, a user will experience an enhanced perception compared to an application that only changes the spatial separation for the purpose of reducing cross-electrode interference.

[0092] The temporal masking decay function may include a time constant that may be fixed or vary across electrodes in a plurality of electrodes (this will produce an electrode-specific decay function). The decay function may be determined by psychophysical and / or objective neurophysiological measures such as electrically evoked compound action potentials (ECAPs). The amount of masking imposed on a given electrode pulse may be calculated using all preceding pulses that fall within the same time window, i.e., the analysis epoch, or alternatively, pulses from previous time windows may be considered. The latter is useful if the previous time window is sufficiently shorter than the time scale at which masking decay occurs. In this case, masking from previous pulses within the previous time window may be included if the temporal masking factor has not yet decayed below a certain decay threshold, or the amount of masking (i.e., spatial masking multiplied by the temporal decay factor) still exceeds the masking threshold.

[0093] For a given patient, the spatial masking contribution can be determined based on a spatial masking function that can be derived directly from obtaining objective measurements (such as eCAP) or behavioral psychophysical metrics. When both methods are used simultaneously, the process can be accelerated by collecting a smaller number of objective or behavioral measurements and by using a spatial masking model that is appropriate for the patient. masker That is, the stimulation on one electrode from other electrodes is imposed on one electrode masked The masking can be achieved by adding electrodes e masker Stimulation and use of electrodes masked The eCAP response is measured and determined by repeatedly changing the electrode e masked , while keeping the stimulating electrode fixed on the e masked On, by e masker The spatial masking function MS(e masked ,e masker ) can be derived. This can be repeated for multiple stimulation electrodes to derive the masking function associated with the stimulation on each electrode. Alternatively, when using the "masker-probe" eCAP method, by presenting the probe stimulus to electrode e masked and presenting the masking stimulus to e masker and through the electrode e masked The eCAP response on the upper or adjacent electrode can determine the MS (e masked ,e masker By repeating this process for different combinations of "probe" and "masker" electrodes, a complete set of spatial masking functions can be determined. A variety of behavioral psychophysical tests can also be used to determine the spatial function. An example of such a test may include a fixed level at electrode e. masker On stimulation, and by placing electrodes at different levels masked The stimulation (using standard psychophysical methods) was determined in the presence of e masker Detection of stimulation on maskedthe minimum level of stimulation required on the electrode. Again, by repeatedly varying the e masked fixed, the spatial masking function associated with stimulation on the e masker fixed, the spatial masking function associated with stimulation on the e masker fixed, the spatial masking function associated with stimulation on the e

[0094] The temporal masking contribution of each of the one or more other electrodes can be expressed by a temporal masking function. The temporal masking function can also be derived directly by taking objective or behavioral measures of a range of time differences between the preceding and lagging pulses. This process can be expedited by fitting a model of the temporal masking decay function using a smaller number of measurements. The decay of masking with increasing pulse time difference can be determined with eCAPs by varying the difference between the offset time of the stimulating electrode pulse and the time at which the eCAP response is measured. When using the "probe-masker" approach, it can also be measured by measuring eCAPs with different inter- stimulus intervals between the masker and the probe. Similarly, a behavioral temporal masking test can be performed in which, for a given preceding pulse level, the detection threshold level of the lagging pulse is measured for different time differences. In both cases, the effects of the temporal and spatial separation between the preceding and lagging pulses can also be measured. For eCAPs, the electrode and the time at which the measurements are made can be varied. Similarly, for behavioral tests, the electrode and the timing of the lagging pulse can be varied.

[0095] The processor unit can be configured to control the cross-electrode interference between the electrode pulses of the plurality of electrodes by varying the first time of the electrode pulse of the electrode and / or the preceding time of each of the one or more electrode pulses of the other electrodes. For example, if the cross-electrode interference imposes a large cognitive load on the user, the cochlear implant system can comprise a sensor configured to measure the cognitive load of the user, such that, based on a measurement signal comprising the measured cognitive load, the processor unit is configured to control the cross-electrode interference. The sensor can be a part of the electrode array, the implant part or an external part connected to the cochlear implant system. The sensor can comprise one or more electrode pads made of Ir02.

[0096] The processor unit can be configured to control the cross-electrode interference between the electrode pulses of the plurality of electrodes by applying a time delay between a first time window and a second time window, wherein, in both time windows, the processor unit is configured to select electrodes from a subset of electrodes of the plurality of electrodes and / or to select electrodes from a main group of electrodes of the plurality of electrodes.

[0097] The first time, the preceding time and the time delay can be continuously determined by the processor unit.

[0098] The first time, the preceding time and the time delay can be continuously determined by the processor unit based on masking measures provided by the electrodes of the plurality of electrodes during fitting of the cochlear implant system and / or during operation of the cochlear implant system.

[0099] The processor unit can be configured to control the cross-electrode interference based on a subjective measure, e.g. a questionnaire, presented to the user via a graphical user interface. The user can receive one or more questions relating to the perceptibility of the generated stimulation provided to the user's auditory nerve. The graphical user interface comprises an input interface for receiving the user's answers to the questions. The processor unit can receive a command signal which determines the control of the cross-electrode interference. The command signal can be determined by an external server or computer connected to the graphical user interface based on the questionnaire and the answers from the user. The command signal can for example comprise a change in the amount of the pulse time difference and / or a change in the time window and / or a delay between the time windows, e.g. the first time window and the second time window. The graphical user interface can be part of a smartphone, tablet or any computing device. The input interface can be separate from the graphical user interface.

[0100] The processor unit can be configured to determine the state of an electrode pulse by subtracting the amount of cross-electrode interference caused by one or more electrode pulses from other electrodes from the estimated pulse energy / charge / level of the electrode pulse to determine a masking adjusted energy / charge / level.

[0101] The processor unit can be configured to determine the state of an electrode pulse by determining a masking weighted energy / charge / level which comprises the estimated pulse energy / charge / level of the electrode pulse multiplied by a cross-electrode interference scaling factor (i.e. between 0 and 1), the scaling factor comprising the effective energy / charge / level of the electrode pulse taking into account the amount of cross-electrode interference introduced to the electrode pulse from one or more electrode pulses from other electrodes, whereby the effective energy / charge / level provides an estimate of the energy / charge / level that would produce the same amount of activity in the auditory nerve as the pulse of interest would produce in the absence of cross-electrode interference.

[0102] These masking adjusted energy / charge / levels will substantially reflect the amount of information transfer that each electrode pulse will be able to pass onto the auditory nerve, and thus reflect the prominence / intensity of the perceptual effect it will elicit, given the level and relative timing of the pulses on each electrode. Since the parameter / feature of interest is not directly related to the masking weighted energy / charge, a channel / electrode with a high parameter / feature value can not necessarily have their information reliably passed onto the auditory nerve due to masking.

[0103] The parameter / feature can for example be periodicity / time coherence, envelope modulation depth and shape, interaural coherence and cross-microphone coherence of the audio signal.

[0104] The status of the electrode pulses can be determined based on the values of the parameters / features. For example, in a first stage, a set of electrodes can first be identified by calculating the parameter / feature values on each electrode to which an electrode pulse has been assigned and selecting the electrode with the highest value, i.e. the electrode with the highest parameter / feature value. These selected electrodes will then constitute the electrodes of the cochlear implant system that best convey the perceptual feature of interest to the user of the cochlear implant system, if there is no masking effect. In a second stage, the masking adjusted energy / charge / level can then be calculated for all possible electrodes or only for the sub-set of highest value electrodes. These masking adjusted energy / charge / levels will essentially reflect the amount of information that each channel / electrode pulse will be able to transmit onto the auditory nerve and thus the prominence / intensity of the perceptual effect it will evoke, given the level and relative timing of the electrode pulses on each electrode. Since the parameter / feature of interest is not directly related to the masking adjusted energy / charge, electrodes with a high importance value based on the parameter / feature of interest can not necessarily have their information reliably transmitted onto the auditory nerve due to masking. Therefore, a further processing can be introduced whereby the energy / charge of the electrodes with the highest importance value is increased to increase their masking weighted energy / charge values and thus the ability of those electrodes to transmit information onto the auditory nerve. The range of possible energy / charge adjustments should be limited to prevent excessive (and unsafe) variations of the pulse energy / charge. The amount of masking adjusted energy / charge / level applied to a given electrode can also be adjusted by the parameter / feature value of that electrode, such that electrodes with the highest parameter values are adjusted to have higher masking adjusted energy / charge / levels than electrodes with lower parameter / feature values. After these energy / charge adjustments, the determination of the masking adjusted energy / charge / level can be applied using all electrodes or only the sub-set of highest value electrodes. The result is that the electrodes with the highest masking adjusted energy / charge / level will be selected for stimulation, whereby some of the energy / charge / levels on the electrodes can have been boosted to compensate for the masking effect.

[0105] The processor unit can then be configured to amplify the pulse energy / charge / level of the electrode pulses, thereby boosting the energy level and compensating for the masking effect, i.e. the cross-electrode interference.

[0106] The processor unit can be configured to set the cochlear implant system into an energy saving mode by increasing the importance threshold. By increasing the threshold, more electrode pulses will not be selected by the processor unit. In other words, the processor unit is configured to remove electrode pulses that are assumed to not provide a perceptual benefit, thereby saving energy and transmitting fewer pulses.

[0107] Definitions

[0108] In the present specification, a cochlea stimulation system or a hearing aid comprising a cochlea stimulation system refers to a device adapted to improve and / or enhance the hearing ability of a user by receiving acoustic signals from the user's environment, generating corresponding electric audio signals, possibly modifying the electric audio signals, and providing the possibly modified electric audio signals as audible signals to at least one ear of the user via an array of electrodes.

[0109] More generally, a hearing aid comprises an input transducer for receiving acoustic signals from the user's environment and providing corresponding input audio signals and / or a receiver for receiving input audio signals electronically (i.e. wired or wirelessly), a (typically configurable) signal processing circuit (such as a signal processor, e.g. comprising a configurable (programmable) processor, e.g. a digital signal processor) for processing the input audio signals, and an output unit for providing audible signals to the user in dependence of the processed audio signals. The signal processor can be adapted to process the input signals in the time domain or in a plurality of frequency bands. In some hearing aids, an amplifier and / or a compressor can constitute the signal processing circuit. The signal processing circuit typically comprises one or more (integrated or separate) storage elements for executing programs and / or for holding parameters used (or possibly used) in the processing and / or for holding information suitable for the function of the hearing aid and / or for holding information used in connection with an interface to the user and / or to a programming device (such as processed information provided by the signal processing circuit, for example). In some hearing aids, the output unit can comprise a transducer, e.g. a vibrator for providing a structure-borne or liquid-borne acoustic signal. In some hearing aids, the output unit can comprise one or more output electrodes for providing an electric signal (such as a multi-electrode array for electrically stimulating the cochlea nerve).

[0110] In some hearing aids, the vibrator can be adapted to transmit a structure-borne acoustic signal transdermally or by the skin to the skull. In some hearing aids, the vibrator can be implanted in the middle ear and / or in the inner ear. In some hearing aids, the vibrator can be adapted to provide a structure-borne acoustic signal to the middle ear bones and / or to the cochlea. In some hearing aids, the vibrator can be adapted to provide a liquid-borne acoustic signal to the cochlea fluid, e.g. through the oval window. In some hearing aids, the output electrodes can be implanted in the cochlea or on the inner side of the skull and can be adapted to provide an electric signal to the hair cells of the cochlea, to one or more auditory nerves, to the auditory brainstem, to the auditory midbrain, to the auditory cortex and / or to other parts of the cerebral cortex.

[0111] A "hearing aid system" refers to a system comprising one or two hearing aids, e.g. one BTE unit and one cochlear implant. A "binaural hearing aid system" refers to a system comprising two hearing aids and adapted to cooperatively provide audible signals to both ears of a user. The hearing aid system or binaural hearing aid system can further comprise one or more "auxiliary devices" which communicate with the hearing aid and influence and / or benefit from the functionality of the hearing aid. The auxiliary device can e.g. be a remote control, an audio gateway device, a mobile phone (e.g. a smartphone) or a music player. The hearing aid, hearing aid system or binaural hearing aid system can e.g. be used to compensate for a loss of hearing ability of a hearing impaired person and / or to enhance the hearing ability of a normally hearing person and / or to transmit electronic audio signals to a person. The hearing aid or hearing aid system can e.g. form part of or interact with a public address system, an active ear protection system, a hands-free telephone system, a car audio system, an entertainment (e.g. karaoke) system, a teleconference system, a classroom amplification system, etc.

[0112] A "unit" is a device with technical and functional features. A "unit" is considered a device within the present invention. BRIEF DESCRIPTION OF DRAWINGS

[0113] The various aspects of the present invention will be best understood with the aid of the following detailed description and accompanying drawings. These drawings are all schematic and simplified for clarity and they only show details which are necessary in order to explain the present invention, other details being omitted. Throughout the specification like drawing references have been used to indicate like or corresponding parts. Each feature of each aspect can be combined with any or all features of any other aspect. These and other aspects, features and / or technical effects will be apparent from and elucidated with reference to the drawings, in which:

[0114] Figures 1A-1C An example of a cochlear implant system is shown;

[0115] Figure 2A And 2B A known "M-of-N" type electrode selection often employed in sound coding strategies is shown;

[0116] Figures 3A-3C An example of a processor unit selecting and retaining N electrodes out of a plurality (M) of electrodes is shown;

[0117] Figure 4 An example of determining an importance threshold is shown;

[0118] Figure 5 An example of an electrode array arranged in the cochlea of a user of a cochlear implant system is shown;

[0119] Figure 6A And 6B An example of a masking model scheme including a spatial masking contribution is shown;

[0120] Figures 7A-7E An example is shown of a masking model scheme including a determination of a temporal masking contribution;

[0121] Figure 8 An example is shown of a masking model scheme including a determination of a spatial masking contribution and a determination of a temporal masking contribution;

[0122] Figures 9A-9D Different examples of cochlear implant systems are shown. DETAILED DESCRIPTION

[0123] The detailed description set forth below, in connection with the appended drawings, is intended as a description of various configurations and is not intended to limit the scope of the concepts. The detailed description includes specific details for the purpose of providing a thorough understanding of the various concepts. However, it will be apparent to those skilled in the art that these concepts can be practiced without these specific details. Numerous specific details are set forth in order to provide a thorough understanding of the subject matter. Well-known structures have been presented without detail in order to not obscure the subject matter. It will be apparent to one of ordinary skill in the art, however, that the concepts can be practiced with only some of the specific details presented and without all of the others. Aspects of the apparatus and methods are set forth by various blocks, functional units, modules, elements, circuits, steps, processes, algorithms, etc. (collectively referred to as "elements"). Such elements can be implemented using electronic hardware, computer software, or any combination thereof.

[0124] The structural features of the apparatus described above, in the detailed description of the "DETAILED DESCRIPTION", and defined in the claims can be combined with the steps of the method for determining temporal fine structure parameters when appropriately replaced by corresponding processes.

[0125] As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, i.e. having the meaning of "at least one". It will be further understood that the terms "has", "have", "having", "include" and / or "comprise" and / or "comprising", as used herein, are intended to indicate the presence of the stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements can be present. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items. Unless explicitly stated otherwise, the steps of any method disclosed herein need not be performed in the order in which they are disclosed.

[0126] It is to be appreciated that a reference to "one embodiment" or "an embodiment" or "aspect" or "may" in the specification or claims of this application means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one implementation of the application. The appearances of the phrase "in one embodiment" or "an embodiment" or "aspect" in various places in the specification are not necessarily all referring to the same embodiment, nor are they necessarily referring to one specific embodiment. Furthermore, the particular features, structures, or characteristics can be combined in any suitable manner on one or more implementations. The foregoing description has been presented for purposes of illustration.

[0127] The claims are not limited to the various aspects shown in this disclosure, but include the full scope potentially consistent with the language of the claims, where the use of the terms "one," "a," or "the" are not intended to foreclose the use of "plurality of," "a second" or "another" with respect to a particular element. The terms "some" and "another" are defined with reference to at least one of the claims.

[0128] Thus, the scope of the application should be determined by the claims.

[0129] Figures 1A-1C An example of a cochlear implant system 1 is shown, which comprises a microphone unit 2 configured to receive an acoustic signal and to transmit an audio signal based on the acoustic signal. The cochlear implant system 1 comprises a processor unit 3 configured to receive the audio signal and to process the audio signal into a plurality of electrode pulses, and an electrode array 4 comprising a plurality of electrodes 5 configured to stimulate a hearing nerve of a user of the cochlear implant system based on the plurality of electrode pulses.

[0130] In Figures 1A-1C the electrode array 4 is arranged within a cochlea 10 of a user of the cochlear implant system 1.

[0131] In Figures 1A-1C the processor unit 3 is configured to assign an importance value to one or more electrodes of the plurality of electrodes 5, wherein each importance value is determined based on a status of an electrode pulse assigned to the respective electrode. The processor unit 3 is further configured to select a set of primary electrodes MS of the plurality of electrodes M during a time window TW, wherein the importance value of each selected electrode of the set of primary electrodes MS is greater than or equal to an importance threshold value. The processor unit 3 is configured to enable electrodes of the set of primary electrodes to stimulate the hearing nerve based on electrode pulses of the plurality of electrode pulses, and to maintain the electrodes of the set of primary electrodes MS in a hold mode during a hold period.

[0132] In Figure 1A and 1BIn FIG, a cochlear implant system 1 includes an external portion 20 disposed on the head of a user of the cochlear implant system 1 and an implant portion 30 disposed beneath the user's skin 50. The external portion 20 includes a first sensing interface 21, and the implant portion 30 includes a second sensing interface 31, wherein the external portion 20 is configured to communicate with the second sensing interface 31 of the implant portion 30 via the first sensing interface 21. The implant portion 30 is connected to the electrode array 10.

[0133] exist Figure 1A In the embodiment of the present invention, the processor unit 3 is disposed in the external part 20, and the external part 20 includes a memory unit 22. In another example, the memory unit 22 may be disposed in the implant part 30. The memory unit is configured to store importance values ​​of one or more electrodes, and the processor unit may be configured to continuously update the importance values ​​based on changes in the states of electrode pulses assigned to the corresponding electrodes.

[0134] exist Figure 1B In the embodiment, the processor unit 3 is arranged in the implant part 30.

[0135] exist Figure 1C In the embodiment of the present invention, a cochlear implant system 1 includes an implant part 30, wherein the implant part 30 includes a microphone 2, a processor unit 3, and a memory unit 22. Optionally, the implant part may include a communication interface configured to communicate with an external device such as a remote processor unit, a smartphone, or any computing device inductively or via an electromagnetic link such as an RF link.

[0136] Figure 2A and 2B The figure shows the known "M-of-N" type of electrode selection often employed in phonocoding strategies that generate electrode pulses at a fixed or variable stimulation rate over time and across electrodes. In cochlear implant systems with a fixed stimulation rate, the time window (TW1-TW10) within which the M-of-N selection is performed is set so that an equal number of events (i.e., pulses) are likely to occur on each electrode, and no electrode has a selection advantage at any given time due to the stimulation rate at which the events occur. However, when the phonocoding strategy is based on events occurring at a stimulation rate that varies over time and across electrodes, electrodes with higher event rates can have a selection advantage over electrodes with lower rates due to the inclusion of more events in the time window in which the M-of-N selection is performed.

[0137] exist Figure 2AIn this example, the total number of electrodes is set to 3 (M) and the subgroup of electrodes is set to 2 (N). In this example, the selection of electrodes is applied to a fixed rate vocoding strategy. The stimulation rate is such that each time window (TW1-TW10) contains one event, i.e. an electrode pulse, on each electrode 5. The abscissa refers to the time windows and the ordinate corresponds to the electrode numbers. The values shown in each electrode-time window cell indicate the significance value of the event contained. The electrodes selected in each time window are marked with a circle. For example, in the first time window TW1, two electrodes are selected, having significance values 4 and 5, respectively. These two electrodes are selected because they have the highest significance values among the three electrodes. The selection of the two electrodes is maintained in time windows TW1 to TW4, and in time window TW5, a third electrode is selected in preference to one of the two previously selected electrodes. Again, the significance values of the two selected electrodes are the highest among the three electrodes. The selected electrodes are maintained in time window TW6.

[0138] In Figure 2B , the total number of electrodes is set to 3 and the subgroup of electrodes is set to 2. In this example, the selection of electrodes is applied to a variable rate vocoding strategy, in which the stimulation rate increases with the electrode number. The abscissa refers to the time epochs, the ordinate corresponds to the electrode numbers, and "-" indicates the absence of an event, i.e. an electrode pulse. The values shown in each electrode-epoch cell indicate the significance value of the event contained. The situation illustrated in this figure can for example represent the case of speech with high energy low frequency content in the presence of higher frequency noise. The electrodes selected in each epoch, i.e. time window, are marked with a circle. In this example, there are time windows in which a low significance value on a high rate channel is selected, because there is no event on the other channels. These low significance values can interfere with more significant events on other electrodes, indicated by the dashed boxes.

[0139] Figures 3A-3C An example of a processor unit selecting and maintaining N electrodes out of a plurality (M) of electrodes with a variable rate vocoding strategy, in which the stimulation rate increases with the electrode number, is shown. Furthermore, Figures 3A-3C An example of how the present application solves the problems of the known "select N out of M" type electrode selection schemes shown in Figure 2A and 2B is shown.

[0140] In Figure 3A , the abscissa refers to the time epochs, the ordinate corresponds to the electrode numbers, and "-" indicates the absence of an event. The values shown in each electrode-epoch cell indicate the significance value of the electrode pulse. The electrodes selected in each time window are marked with a circle, those in the maintenance mode are indicated with a boxed "-", and those blocked from selection by another electrode in the maintenance mode are indicated with a cross.

[0141] In the first time window TW1 the processor unit 3 has selected two electrodes with an importance value that is equal to or greater than the importance threshold. The selected electrodes are part of a main set of electrodes (MS, 41) of the plurality of electrodes (4, 5). In this example the importance threshold is 3. In the time windows TW2 and TW3 the reserved electrodes are inactive, which means that no electrode pulse is assigned to these electrodes. During these time windows, i.e. TW2 and TW3, the processor unit 3 is not allowed to select an active electrode because the importance value of that electrode / electrode pulse is lower than the importance threshold. However, if the importance value of an electrode that is not reserved is equal to or higher than the importance threshold, see for example the time windows TW5 and TW6, the processor unit 3 will not be allowed to select that electrode. This electrode reservation is denoted as a "hard" reservation.

[0142] In the time window TW4 the processor unit 3 is configured to select a sub-set of electrodes 42 in the main set of electrodes 41 because the importance value of the selected electrodes is greater than or equal to the importance threshold.

[0143] During the time window TW1 the processor unit 3 enables two electrodes 5 in the electrode array 4 to stimulate the auditory nerve of the cochlea of the user. In the time window TW4 only one electrode is selected to stimulate the auditory nerve, and so on for the other time windows.

[0144] In Figure 3B In the example of Fig. 6 the processor unit 3 is configured to select a plurality of electrodes (40, 5) and / or a sub-set of electrodes 42 of the main set of electrodes 41 during the reserved time period, see for example the time windows TW4 to TW7 and TW10, wherein each electrode of the sub-set of electrodes has an importance value that is greater than or equal to the importance threshold. For example in the time window TW6 the processor unit 3 has selected two electrodes, one selected from the main set 41 and one selected from the plurality of electrodes 40 that comprises electrodes that are not part of the main set of electrodes. In TW5 the processor unit 3 has selected one electrode that is part of the plurality of electrodes 40.

[0145] The processor unit 3 is configured to enable the electrodes 5 in the electrode sub-group 42 to stimulate the auditory nerve based on the electrode pulse in the plurality of electrode pulses.

[0146] In Figure 3C , the importance values of the reserved electrodes are updated at each time window (TW1-TW10) during the duration of the reservation period. The abscissa refers to the time window, and the ordinate corresponds to the electrode number. The value displayed in each electrode-time window cell indicates the importance value of the contained electrode pulse, i.e. event, the larger font number designating the importance value of the electrode pulse, and the smaller font number designating the importance of the electrode pulse in the reserved mode. To illustrate the difference between this embodiment and the embodiment shown in Figure 3B , the electrodes that have been additionally selected or rejected are marked with a circle or a cross, respectively.

[0147] Figure 3C In time window TW9 of , an example of a tie-breaker is seen between three electrodes carrying active electrode pulses that have the same importance scale value. Only the subgroup of these electrodes can be selected. In this case, a tie-breaker importance value can be assigned by the processor unit 3 to the electrodes that are part of the tie-breaker. The tie-breaker importance value must be different from the aforementioned importance values or previously defined importance values, e.g. the importance values in time window TW8, and can include: channel center frequency, pulse energy, signal-to-noise ratio, interaural coherence, periodicity, etc.

[0148] The tie-breaker importance value of the electrodes is not shown in Figure 3C .

[0149] The processor unit 3 can be configured to update the electrode main group 41 by adding a new electrode 5 in the plurality of electrodes 4 to the electrode main group 41, wherein the importance value of the new electrode is greater than or equal to the first importance threshold (Thimp_1).

[0150] The processor unit 3 can be configured to prolong the reservation period of the electrodes 5 of the electrode main group 41 when a new electrode pulse generating an event occurs on an electrode 5 whose importance value is greater than or equal to the importance threshold (Thimp).

[0151] The processor unit 3 can be configured to remove an electrode 5 from the electrode main group 41 when the reservation period of the electrode 5 has expired and before the reservation period is prolonged.

[0152] Figure 4An example of determining the importance threshold is shown. The minimum importance threshold (Thimp_min) is determined by summing the noise floor NF of the electrode array 4 (of the cochlear implant system 1) and a safety margin ΔS of 1-3 dB. If the noise estimate NF is on average 20 dB SPL and the selected safety margin ΔS is 3 dB, then the resulting minimum importance threshold (Thimp_min) is set to 23 dB SPL.

[0153] The minimum importance threshold (Thimp_min) can be any measurable parameter of the electrode pulse, such as the center frequency, the signal-to-noise ratio, the noise floor and the electrode pulse energy.

[0154] Thereafter, the minimum importance value (Imp_min) of the electrode sub-group 42 and / or the electrode main group 41 can be determined based on the assigned importance values of the electrode sub-group 42 and / or the electrode main group 41, respectively. In this example, the electrodes numbered E6, E5 and E2, which are part of the main group 41, have an importance value of 5, 5 and 4, respectively. The electrodes numbered E3, E1 and E0, which are part of the sub-group 42, have an importance value of 4, 5 and 4, respectively.

[0155] If the minimum importance value (imp_min) is greater than or equal to the minimum importance threshold (Thimp_min), then the importance threshold (Thimp) can be equal to the minimum importance value (imp_min).

[0156] If the minimum importance value (imp_min) is less than the minimum importance threshold (Thimp_min), then the importance threshold (Thimp) can be equal to the minimum importance threshold (Thimp_min).

[0157] Figure 5 An example of an electrode array 4 arranged within the cochlea 10 of a user of a cochlear implant system 1 is shown. The cochlea comprises a plurality of auditory nerves 11 which are to be stimulated by the electrodes (5A, 5B) of the electrode array 4. In this example, the electrodes (5A, 5B) produce an excitation overlap 61 which produces a significant cross-electrode interference, i.e. masking, whereby the stimulation on one of the two electrodes (5A, 5B) consumes some of the neural resources of the auditory nerves at the neighboring electrode (5A, 5B) sites, thus disturbing the neural excitation evoked by the stimulation.

[0158] Figure 6A and 6B An example of a masking model scheme including a spatial masking contribution is shown. In Figure 6A the processor unit 3 has selected and activated two electrodes (5A, 5B) within a first time window (TW), and in Figure 6B, the processor unit 3 has selected and enabled two electrodes (5A, 5B) within a second time window (TW). The spatial separation between the two electrodes enabled within the first time window is smaller than the spatial separation between the two electrodes enabled within the second time window. The spatial separation is determined based on the respective stimulation levels of the two electrodes, and in the second time window, the processor unit 3 has reduced the stimulation level of at least one of the two electrodes to reduce the spatial masking contribution. Alternatively, the spatial separation may be increased by selecting electrodes (5A, 5C) that are physically arranged far apart from each other. Figure 6A In , activation of two electrodes (5A, 5B) produces an overlap of excitation 61 that results in cross-electrode interference. Figure 6B In , the spatial separation has increased, which results in the elimination of excitation overlap 61. The change in spatial separation is accommodated by a masking model scheme that determines the spatial masking contribution of each of the two electrode pulses for the two electrodes (5A, 5B).

[0159] Figures 7A-7E An example of a masking model scheme including a temporal masking contribution is shown, more specifically, the cross-electrode interference between electrodes (5A, 5B) is determined by the temporal masking contribution of each of the two electrode pulses of the two electrodes (5A, 5B). Figure 7B and 7C In the example, the electrodes to be activated are the same as those in Figure 7A Seen in. Figure 7A , for the case where the pulse time difference is ΔT1 and the time delay ΔTW between the first time window TW1 and the second time window TW2 is zero, an excitation overlap 61 can be seen. In both time windows, the processor unit 3 is configured to select a main group of electrodes 41 from the plurality of electrodes, wherein the importance value of each electrode in the selected main group of electrodes 41 or electrode subgroup 40 is greater than or equal to an importance threshold.

[0160] exist Figure 7B , pulse time differences ΔT1 and ΔT2 are between a first time and a second time, respectively, of an electrode pulse assigned to two electrodes (5A, 5B). Processor unit 3 increases the pulse time difference from ΔT1 to ΔT2, which results in a reduction in cross-electrode interference. The increase in pulse time difference ΔT is performed based on a masking model scheme that determines a temporal masking contribution of each of the two electrode pulses of the two electrodes (5A, 5B).

[0161] Figure 7C An example is shown in which the time delay ΔTW1 between the first time window TW1 and the second time window TW2 is set to zero. In this example, cross-electrode interference is high. The processor unit 3 then increases the time delay between the two time windows (TW1, TW2) based on the masking model scheme, which results in a reduction in cross-electrode interference between the activated electrodes (5A, 5B).

[0162] Figure 7D An example of an electric stimulation spectrogram showing a sequence of electrode pulses within a time window, and Figure 7E An example of a temporal masking decay function MD(t3-t4) is shown, which is a function of the time difference between a preceding electrode pulse, such as electrode pulse t3, and a masked (lagging) pulse, such as electrode pulse t4. The temporal masking decay function MD(t3-t4) is associated with the masking of a pulse on channel / electrode 3 by a pulse on channel / electrode 4.

[0163] Figure 8 An example is shown in which the masking model scheme includes a determination of a spatial masking contribution and a determination of a temporal masking contribution. In another example, the processor unit 3 is configured to determine the importance value based on either or both of the spatial masking contribution and the temporal masking contribution. The processor unit 3 can switch between using both contributions or using only one of the contributions, and wherein the switching is based on the number of selected electrodes of the main group of electrodes. For example, if the processor unit 3 selects to include the temporal masking contribution, and no electrode has an importance value above the importance threshold, the processor unit 3 can switch to using the temporal and spatial masking contributions or only the spatial masking contribution.

[0164] Figures 9A-9D Different examples of the cochlear implant system 1 are shown, which include sensors (50A-50D) for measuring parameters to be used in determining the status of the electrode pulses. The measurements can be made during fitting and / or during operation of the cochlear implant system 1. In Figure 9A In the example shown in Fig. 1, the electrode array 4 is arranged within the cochlea 10 of a user. The electrode array 4 includes electrodes 5 configured to stimulate the auditory nerve 11 of the cochlea 10 and includes sensors (50A-50D). In Figure 9B In the example shown in Fig. 1, the electrode array 4 is arranged within the cochlea 10 of a user. The electrode array 4 includes electrodes 5 configured to stimulate the auditory nerve 11 of the cochlea 10 and includes sensors (50A-50D). In

[0165] Optionally, the second layer can be removed, thereby reducing the thickness of the electrode array 4.

[0166] The sensors (50A-50D) and / or the electrodes 5 can be used to make eCAP measurements.

[0167] In Figure 9C and 9D the sensor 50 is arranged on the implant part 30 or the external part 20. In this example, the sensor is configured to measure the cognitive load of the user, and the processor unit is configured to control the cross-electrode interference based on the measured cognitive load. The sensor can be part of the electrode array 4, the implant part 30 or the external part 20. The sensor can comprise one or more electrode pads made of Ir02.

Claims

1. A cochlear implant system comprising: - a microphone unit configured to receive an acoustic signal and to transmit an audio signal based on the acoustic signal; - a processor unit configured to receive an audio signal and process the audio signal into a plurality of electrode pulses; - an electrode array comprising a plurality of electrodes configured to stimulate an auditory nerve of a user of the cochlear implant system based on the plurality of electrode pulses; The processor unit is configured to: - assigning an importance value to one or more electrodes of the plurality of electrodes, wherein each importance value is determined based on a state of an electrode pulse assigned to the corresponding electrode; - selecting a main group of electrodes from a plurality of electrodes during a time window, wherein an importance value of each selected electrode in the main group of electrodes is greater than or equal to an importance threshold; - activating electrodes of the main group of electrodes to stimulate the auditory nerve based on an electrode pulse in the plurality of electrode pulses; and - holding electrodes of the main group of electrodes in a holding mode during a holding period, and wherein the held electrodes of the main group of electrodes influence the selection of electrodes of the main group of electrodes during a subsequent time window.

2. The cochlear implant system according to claim 1, comprising a memory unit configured to store importance values ​​of one or more electrodes, and the processor unit configured to update the importance values ​​based on changes in states of electrode pulses assigned to corresponding electrodes.

3. A cochlear implant system according to claim 1 or 2, wherein the processor unit is configured to select an electrode subset from the main electrode group during the retention time period, and no other electrodes of the plurality of electrodes are allowed to be selected, wherein each electrode in the electrode subset has an importance value greater than or equal to an importance threshold, and wherein the processor unit is configured to enable electrodes of the electrode subset to stimulate the auditory nerve based on electrode pulses in the plurality of electrode pulses.

4. The cochlear implant system of claim 1 , wherein the processor unit is configured to select an electrode subgroup from a plurality of electrodes and / or a main group of electrodes during a retention period, wherein each electrode of the electrode subgroup has an importance value greater than or equal to an importance threshold, and wherein the processor unit is configured to enable electrodes of the electrode subgroup to stimulate the auditory nerve based on electrode pulses from a plurality of electrode pulses.

5. The cochlear implant system according to claim 1 or 2, wherein the importance threshold is determined as follows: Determine the minimum significance threshold; determining a minimum importance value for the electrode subgroup and / or the electrode main group; If the minimum importance value is greater than or equal to the minimum importance threshold, the importance threshold is determined to be equal to the minimum importance value; if the minimum importance value is less than the minimum importance threshold, the importance threshold is determined to be equal to the minimum importance threshold.

6. The cochlear implant system according to claim 1 or 2, wherein the processor unit is configured to update the main set of electrodes by: - adding a new electrode to the plurality of electrodes, wherein the importance value of the new electrode is greater than or equal to a first importance threshold; - extending the retention period of an electrode of the main group of electrodes when an event generating a new electrode pulse occurs on an electrode whose importance value is greater than or equal to the importance threshold; and / or - Removing the electrode from the main set of electrodes when the retention period of the electrode has expired and before the retention period is extended.

7. The cochlear implant system of claim 1 or 2, wherein the processor unit is configured to sample the audio signal within a frequency range, and wherein the state of the electrode pulse in the plurality of electrode pulses comprises: - an estimated pulse energy level within a frequency subset of the frequency range, an estimated signal-to-noise ratio of an audio signal sampled within the frequency subset of the frequency range; - periodicity in the audio signal within a frequency subset of the frequency range, audio coherence across electrodes in the plurality of electrodes; and / or - Audio coherence between audio signals received by more than two microphones across a microphone unit.

8. The cochlear implant system of claim 1 or 2, wherein the state of an electrode pulse in the plurality of electrode pulses is determined based on a masking model scheme of cross-electrode interference imposed on the electrode pulse by other electrode pulses in the plurality of electrode pulses.

9. The cochlear implant system of claim 8 , wherein the state of the electrode pulses of an electrode of the plurality of electrodes comprises an amount of cross-electrode interference on the electrode pulses of the electrode caused by one or more electrode pulses of other electrodes of the plurality of electrodes, determined based on a masking model scheme, wherein the masking model scheme comprises: - determining a spatial masking contribution caused on the electrode pulse of the electrode by each of the one or more electrode pulses of the other electrodes based on the spatial separation between the electrode and each other electrode.

10. The cochlear implant system of claim 8 , wherein the state of the electrode pulses of an electrode of the plurality of electrodes comprises an amount of cross-electrode interference on the electrode pulses of the electrode caused by one or more electrode pulses of other electrodes of the plurality of electrodes, determined based on a masking model scheme, wherein the masking model scheme comprises: -Determining the time masking contribution caused by each of the one or more electrode pulses of the other electrodes on the electrode pulse of the electrode based on the pulse time difference between the first time of the electrode pulse of the electrode and the second time of each of the one or more electrode pulses of the other electrodes, wherein the second time is ahead of the first time.

11. A cochlear implant system according to claim 9 or 10, wherein the masking model scheme comprises determination of a spatial masking contribution and determination of a temporal masking contribution.

12. The cochlear implant system of claim 11, wherein the spatial masking contribution determined from each electrode pulse of the other electrodes is multiplied by a temporal masking decay function that includes a pulse time difference between the electrode pulse of that electrode and each electrode pulse of the other electrodes.

13. The cochlear implant system of claim 12, wherein the temporal masking decay function is an exponential factor comprising a time constant and / or a pulse time difference, and wherein the time constant is the same or different for each electrode of the plurality of electrodes.

14. A cochlear implant system according to any one of claims 10, 12 and 13, wherein the processor unit is configured to control cross-electrode interference by changing a first time of an electrode pulse of the electrode and / or a preceding time of each of one or more electrode pulses of other electrodes or by applying a time delay between a first time window and a second time window, wherein in both time windows the processor unit is configured to select an electrode subgroup from a plurality of electrodes and / or to select an electrode from a main electrode group of a plurality of electrodes.

15. The cochlear implant system according to any one of claims 9, 10, 12 and 13, wherein the processor unit is configured to determine the status of the electrode pulses by determining: - a masked adjusted energy / charge / level comprising the estimated pulse energy / charge / level of the electrode pulse minus the amount of cross-electrode interference induced at that electrode pulse from one or more electrode pulses at other electrodes; or The masked adjusted energy / charge / level comprises an estimated pulse energy / charge / level of the electrode pulse multiplied by an inter-electrode interference conversion factor, the conversion factor comprising the effective energy / charge / level of the electrode pulse after taking into account the amount of inter-electrode interference introduced into the electrode pulse from one or more electrode pulses from other electrodes, whereby the effective energy / charge / level provides an estimate of the energy / charge / level that will produce the same amount of activity in the auditory nerve as would be produced by the pulse of interest in the absence of inter-electrode interference.

16. A cochlear implant system according to claim 15, wherein the processor unit is configured to select electrodes from the main group of electrodes or the subgroup of electrodes that result in a maximized total masked-adjusted / weighted energy / charge / level, and wherein the total masked-adjusted / weighted energy / charge / level comprises the sum of the masked-adjusted / weighted energy / charge / level of each electrode pulse of the selected electrodes.

17. The cochlear implant system of claim 15, wherein the processor unit is configured to amplify the pulse energy / charge / level of the electrode pulses.

18. The cochlear implant system of claim 7, wherein the processor unit is configured to set the cochlear implant system to the energy saving mode by increasing the importance threshold, wherein the importance threshold is: - minimum permissible mask-weighted energy / charge / level value; - Minimum signal-to-noise ratio of electrode pulses; - the minimum estimated pulse energy level of the electrode pulses; - minimum value of the autocorrelation amplitude; or - Minimum interaural coherence value.

Images