Cochlear Implant Fitting Based on Neuronal Status

Abstract

Methods and arrangements are described for developing a virtual channel matrix for mapping analysis channels to stimulation channels for a cochlear implant patient by selecting a stimulation channel and measuring the amplitude growth function for the selected stimulation channel in response to commands to the cochlear implant to apply electrical stimulation pulses for the stimulation channel, where each stimulation pulse comprises a negative and a positive phase separated in time by a first inter-phase-gap; and measuring the amplitude growth function for the selected stimulation channel in response to commands to the cochlear implant to apply electrical stimulation pulses for the stimulation channel, where each stimulation pulse comprises a negative and positive phase separated in time by a second inter-phase-gap and whereby the first and second inter-phase-gaps are different. Thereafter Determining the slopes of the measured amplitude growth functions for the stimulation channel measured with the first and second inter-phase-gaps, and calculating an indicator based at least in part on the difference of the slopes of the amplitude growth functions indicative of the local neural survival for that stimulation channel. Thereafter Repeating this process for each stimulation channel where an indicator shall be derived and selecting for the virtual channel matrix the stimulation channels with best local neural survival by optimizing a function based at least in part on the calculated indicators of the stimulation channels.

Claims

1. A method for developing a virtual channel matrix for mapping analysis channels to stimulation channels for a cochlear implant patient comprising: Select a stimulation channel and measure the amplitude growth function for the selected stimulation channel in response to commands to the cochlear implant to apply electrical stimulation pulses for the stimulation channel, where each stimulation pulse comprises a negative and positive phase separated in time by a first inter-phase-gap; measure the amplitude growth function for the selected stimulation channel in response to commands to the cochlear implant to apply electrical stimulation pulses for the stimulation channel, where each stimulation pulse comprises a negative and positive phase separated in time by a second inter-phase-gap and whereby the first and second inter-phase-gaps are different; determine the slopes of the measured amplitude growth functions for the stimulation channel measured with the first and second inter-phase-gaps, and calculate an indicator based at least in part on the difference of the slopes of the amplitude growth functions indicative of the local neural survival for that stimulation channel; and repeat this process for each stimulation channel where an indicator shall be derived; select for the virtual channel matrix the stimulation channels with best local neural survival by optimizing a function based at least in part on the calculated indicators of the stimulation channels.

2. A method according to claim 1, wherein the first inter-phase-gap has a duration that is as short as possible and just enough to allow measurement of an amplitude growth function.

3. A method according to claim 1, wherein the second inter-phase-gap has an as long as possible duration that is at or close to the maximum acceptable loudness to the patient.

4. A method according to claim 1, wherein the first and second inter-phase-gaps are at least 2 microseconds.

5. A method according to claim 1, wherein the difference between the first and second inter-phase-gaps is at least 8 microseconds.

6. A method according to claim 1, wherein the difference of the first and second inter-phase-gaps is at least 30 or 7.9 microseconds and the inter-phase-gaps are at least 2.1 microseconds.

7. A method according to claim 1, wherein the ratio between first and second inter-phase-gaps is at least 4.

8. A method according to claim 1, wherein the function further includes at least in part a component supporting uniform distribution of selected stimulation channels over the frequency range covered by the stimulation channels where an indicator shall be derived.

9. A method according to claim 1, wherein optimizing the function includes forming stimulation channel groups by grouping a pre-defined number of adjacent stimulation channels and selecting the stimulation channel with the largest calculated indicator in this group, reflecting the highest estimated local neuronal survival.

10. A method according to claim 1, wherein optimizing the function includes selecting the stimulation channels where the calculated indicator exceeds a pre-defined threshold.

11. A method according to claim 1, wherein optimizing the function includes selecting the stimulation channels that minimize the variance of the calculated indicators.

12. A method according to claim 1, wherein optimizing the function includes forming at least two stimulation channel groups by grouping a pre-defined number of adjacent stimulation channels and selection of one stimulation channel per group such that the variance of the calculated indicators from the selected stimulation channels for all groups is minimized.

13. A system for fitting a virtual channel matrix for mapping analysis channels to stimulation channels of an implantable hearing system with a cochlear implant for electrical stimulation of the cochlear, the system comprising: at least one hardware implemented, and programmable processor adapted to perform the steps of: selecting a stimulation channel and measuring the amplitude growth function for the selected stimulation channel in response to commanding the cochlear implant to apply electrical stimulation pulses for the stimulation channel, where each stimulation pulse comprises a negative and positive phase separated in time by a first inter-phase-gap; measuring the amplitude growth function for the stimulation channel in response to commanding the cochlear implant to apply electrical stimulation pulses for the stimulation channel, where each stimulation pulse comprises a negative and positive phase separated in time by a second inter-phase-gap and whereby the first and second inter-phase-gaps are different; determining the slopes of the measured amplitude growth functions for the stimulation channel measured with the first and second inter-phase-gaps, and calculating an indicator based at least in part on the difference of the slopes of the amplitude growth functions indicative of the local neural survival for that stimulation channel; and repeating this process for each stimulation channel where an indicator shall be derived; selecting for the virtual channel matrix the stimulation channels with best local neural survival by optimizing a function based at least in part on the calculated indicators of the stimulation channels.

14. The system according to claim 13, wherein the hardware implemented, and programmable processor is further adapted to allow setting the first and second inter-phase-gaps.

15. The system according to claim 14, wherein the hardware implemented, and programmable processor is further adapted to restrict setting the first and second inter-phase-gaps shorter than 2 microseconds and/or the ratio between first and second inter-phase-gaps not less than 4.

16. The system according to claim 13, wherein the hardware implemented, and programmable processor is further adapted to store default inter-phase-gaps.

17. The system according to claim 16, wherein the default inter-phase-gaps for the first and second inter phase gaps are at least 2.1 microseconds and the difference between the first and second inter-phase-gaps are at least 30 or 7.9 microseconds.

18. The system according to claim 13, wherein the hardware implemented, and programmable processor is further adapted to apply an optimizing function that is further based at least in part on the uniform distribution of selected stimulation channels over the frequency range covered by the stimulation channels where an indicator shall be derived.

19. The system according to claim 13, wherein the hardware implemented, and programmable processor is further adapted to apply an optimizing function that includes forming stimulation channel groups by grouping a pre-defined number of adjacent stimulation channels and selecting the stimulation channel with the largest calculated indicator in this group, reflecting the highest estimated local neuronal survival.

20. The system according to claim 13, wherein the hardware implemented, and programmable processor is further adapted to apply an optimizing function that includes selecting the stimulation channels where the calculated indicator exceeds a pre-defined threshold.

21. The system according to claim 13, wherein the hardware implemented, and programmable processor is further adapted to apply an optimizing function that includes selecting the stimulation channels that minimize the variance of the calculated indicators.

22. The system according to claim 13, wherein the hardware implemented, and programmable processor is further adapted to apply an optimizing function that includes forming at least two stimulation channel groups by grouping a pre-defined number of adjacent stimulation channels and selection of one stimulation channel per group such that the variance of the calculated indicators from the selected stimulation channels for all groups is minimized.

23. A computer program product for fitting a hearing device of a patient, the computer program product comprising a computer usable medium having computer readable program code thereon, the computer readable program code comprising instructions for carrying out the method steps according to claim 1.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0019] FIG. 1 shows anatomical structures of a human ear and some components of a typical cochlear implant system.

[0020] FIG. 2 shows various functional blocks in a signal processing arrangement for a cochlear implant according to a prior art arrangement.

[0021] FIG. 3 shows details of stimulation pulses for stimulation channels utilizing virtual channels.

[0022] FIG. 4 shows a stimulation pulse comprising a negative and positive phase separated in time by an inter-phase-gap and the thereafter measurable ECAP amplitude.

[0023] FIG. 5 shows the estimated slopes of exemplary measured amplitude growth functions with various inter-phase-gap stimuli and the difference of the slopes, indicative of the local neuronal survival.

[0024] FIG. 6 shows an example selection of stimulation channels based on optimized local neuronal survival.

DETAILED DESCRIPTION

[0025] Various embodiments of the present invention are directed to a new coding and fitting algorithm suited for electrical stimulation specifically directed to relatively more intact neural structures nearby the electrode array. In addition to this site optimization, the present invention is directed to a new fitting and coding algorithm suited for the preservation or improvement of tonotopic optimized perception of the electrical stimulation. Various embodiments of the present invention are directed to a cochlear implant, adapted for the new coding algorithm and suitable for use with the new fitting algorithm.

[0026] FIG. 1 shows anatomical structures of a human ear and some components of a typical cochlear implant system as known in the art. FIG. 2 shows various functional blocks in a signal processing arrangement for a cochlear implant according to a prior art arrangement where the present invention finds advantageously application. The details of such an arrangement and its functioning have been set forth in the above discussion and more details are described below to the extent necessary to fully understand the application of the present invention in known or modified cochlear implant systems. In particular, the elements of Pulse Generator 204 are more fully described. It is readily understood by those skilled in the art, that the invention can be used in other implementations of virtual channels in cochlear implants systems, and the following description is understood as a mere example without limiting the invention and its application in any way.

[0027] Pulse Generator 204 weights each analysis channel signal Y.sub.2 to Y.sub.k with a virtual channel matrix W of stimulation amplitude weights that reflect local neural survival to produce a set of electrode stimulation channel signals Z.sub.2 to Z.sub.m for providing an electric stimulation signal to the optimal location in the cochlea to nearby intact neural tissue. Matrix weighting of the stimulation pulses is described further in U.S. Patents U.S. Pat. No. 9,180,295 and U.S. Pat. No. 8,532,782, which are incorporated herein by reference. Equation 1 shows a typical virtual channel matrix W of size m×k:

[00001] $\begin{matrix} W = (\begin{matrix} W_{1 1} & W_{1 2} & .Math. & W_{1 k} \\ W_{2 1} & W_{2 2} & .Math. & W_{2 k} \\ .Math. & .Math. & .Math. & .Math. \\ W_{m 1} & W_{m 2} & .Math. & W_{m k} \end{matrix}) = (\begin{matrix} 1 & 0 & 0 & .Math. & 0 \\ 0 & 0.25 & 0 & .Math. & 0 \\ 0 & 0.75 & 1 & .Math. & 0 \\ .Math. & .Math. & .Math. & .Math. & .Math. \\ 0 & 0 & 0 & 1 & α \\ 0 & 0 & 0 & 0 & 1 - α \end{matrix}) & Equation 1 \end{matrix}$

where m is the number of stimulation channels. The number of stimulation channels may be equal to, larger or smaller than the number of physical stimulation electrodes on the electrode array. Here k is the number of analysis channels and α is an exemplary stimulation channel weighting factor in the range from 0 to 1 associated with analysis channel k. A negative weighting factor W.sub.ij indicates an electrical stimulation pulse with inverted polarity. Electrode stimulation signals Z.sub.1 to Z.sub.m are obtained from analysis channel signals Y.sub.1 to Y.sub.k by matrix multiplication {right arrow over (Z)}= custom-character ×{right arrow over (Y)}, or component wise expressed by Z.sub.l=Σ.sub.n=1.sup.kW.sub.ln.Math.Y.sub.n. For example, for the k.sup.th analysis channel, analysis channel signal Y.sub.k is mapped by virtual channel matrix W to the associated stimulation channel represented by stimulation channel signals Z.sub.1 to Z.sub.m:

[00002] $Z = (\begin{matrix} 1 & 0 & 0 & .Math. & 0 \\ 0 & 0.25 & 0 & .Math. & 0 \\ 0 & 0.75 & 1 & .Math. & 0 \\ .Math. & .Math. & .Math. & .Math. & .Math. \\ 0 & 0 & 0 & 1 & α \\ 0 & 0 & 0 & 0 & 1 - α \end{matrix}) \times (\begin{matrix} 0 \\ 0 \\ 0 \\ .Math. \\ 0 \\ Y_{k} \end{matrix}) = (\begin{matrix} 0 \\ 0 \\ 0 \\ .Math. \\ α .Math. Y_{k} \\ (1 - α) .Math. Y_{k} \end{matrix}) .$

[0028] In a further embodiment, the cochlear implant comprises a tonotopicity correction in Stimulation Timing 203 through stimulation rate-adjustment. In one embodiment Stimulation Timing 203 generates fixed sequences of stimulation timing signals in fixed time-frames that are subsequently adjusted. For example, for a cochlear implant system comprising 12 stimulation channels, each time-frame comprises a time sequence at fixed time-intervals t of stimulation timing signals X.sub.1, . . . , X.sub.s as follows:

X(1)=(1 0 0 0 0 0 0 0)

X(2)=(0 1 0 0 0 0 0 0)

X(12)=(0 0 0 0 0 0 0 1)

where X(t) represents the stimulation timing signals at time-instant t within the time-frame and a one indicates that for the respective stimulation channel the Pulse Generator 204 generates a stimulation pulse. A zero value indicates no stimulation pulse generation for the respective stimulation channel. All stimulation channels in this example are stimulated with the same fixed rate, i.e. inversely proportional to the time-frame duration. In one exemplary embodiment, Stimulation Timing 203 may include a timer and a rate-adjustment vector per stimulation channel. The stimulation channel specific rate-adjustment vector may in part depend on the analysis channel signal. For example, the rate-adjustment may depend on the magnitude of the respective analysis channel signal and may be determined with the inventive fitting method as further described below. For each stimulation channel the timer is first initialized with a base-value defining the frame-rate and the rate-adjustment factor read from the rate-adjustment vector and subsequently started. Upon timer expiry, a respective stimulation timing X(t) signal is generated and provided to Pulse Generator 204 that generates a stimulation pulse for that stimulation channel. The timer is re-initialized as before and re-started and the before described process starts again.

[0029] In one embodiment, reading the rate-adjustment factor from the rate-adjustment vector includes calculating from the analysis channel signal an index and returning the vector component with that specific index. The calculation may involve normalizing the analysis channel signal to within the range from 0 to 1. In one specific example for stimulation channel e, adjustment-vector {right arrow over (a)} may be given through {right arrow over (a)}=(a.sub.0 a.sub.1 a.sub.2)=(0 2 −5). The analysis channel signal Y.sub.e is normalized Y.sub.n,e=Y.sub.e/maxvalue(Y.sub.e) where maxvalue (Y.sub.e) is the maximal value Y.sub.e can become. In this example Y.sub.n,e is assumed to be 0.3. The index i may be calculated as follows:

i=int[0.3.Math.(length({right arrow over (a)})−1)]=int[0.3*2]=1

and the returned rate-adjustment factor is the i's component of {right arrow over (a)}: a.sub.i=a.sub.1=2. The stimulation channel timer may be initialized with the calculated rate-adjustment factor added to the base-value of e.g. 1000, which yields 1002. The channel specific stimulation rate is therefore slightly reduced, because it takes longer for the timer to expire. A negative adjustment factor, e.g. as for a.sub.2=−5, would increase the channel specific stimulation rate. Similarly, the base-value can be changed by the rate-adjustment factor in various other ways and without limitation by multiplication, dividing or any other functional relationship. It may further be possible that the rate-adjustment factor includes the base-value, in which case the timer is initialized by the rate-adjustment factor only. In a further embodiment, rate-adjustment vector may include only a single scalar rate-adjustment factor that may be determined with the fitting method described further below. This simplification reduces computation complexity and has the advantage of lower power consumption as compared to the former method.

[0030] In another embodiment, Stimulation Timing 203 may continuously estimate the instantaneous dominating frequency of the audio signal for at least some analysis channels. For example, with the method as described in U.S. Patent Publication 20180311500 (incorporated herein by reference). The base-value is determined from the such estimated dominant instantaneous analysis channel frequency by calculating the multiplicative inverse. This has the additional advantage, that not only the stimulation channel with optimal neuronal survival is stimulated, but further that the optimal tonotopicity mapping may be simultaneously achieved and adapted to the input audio signal. In a further embodiment, the base-value b may be low-pass filtered, for example with a first-order, unity-gain IIR-filter:

b.sub.t+1=(1−β).Math.b.sub.t+β.Math.b.sub.new.

Here b.sub.t+1 is the updated base-value at time-instant t+1, b.sub.t is the former base-value at time-instant t, b.sub.new is the new calculated base-value from the estimated dominant instantaneous analysis channel frequency and β the filter coefficient in the range from 0 to 1. The filter coefficient β may be a function of b.sub.new and a pre-defined per analysis channel target base-value b.sub.target. In one embodiment the function may be of sigmoidal form:

[00003] $β = \frac{1}{1 + e^{+ a .Math. .Math. b_{new} - b_{target} .Math.}}$

where a is a properly chosen scaling parameter. The target base-value b.sub.target may be determined from the respective analysis channel center frequency and may include the rate-adjustment factor. This has the advantage, that filtering is more robust against frequency estimation errors and the target base-value b.sub.target is prioritized.

[0031] Finally, patient specific stimulation is achieved by generating mono-, bi- or triphasic monopolar pulse shapes for each electrode stimulation signal Z.sub.l to Z.sub.m to form a set of output electrode pulses E.sub.l to E.sub.m to the physical electrode contacts in the implanted electrode array through which adjacent nerve tissue is stimulated. FIG. 3 shows details of simultaneous applied bi-phasic stimulation pulses at adjacent physical electrode contacts of the electrode array. Whenever one of the stimulation timing signals X.sub.l to X.sub.S indicates a stimulation event, the respective number of physical electrode contacts is activated with a set of output electrode stimulation pulses E.sub.l to E.sub.m. Instead of applying a single output stimulation pulse for each selected stimulation event, frequency specific stimulation pulse sequences for one or more stimulation channels may be generated, for example channel specific sampling sequences (CSSS) as described in U.S. Patent U.S. Pat. No. 6,594,525 to Zierhofer, which is incorporated herein by reference. Such stimulation pulse sequences can vary in inter-pulse intervals and amplitude shape. Amplitude shapes can be based on templates, or the amplitudes can fall with a decay, e.g. with an exponential characteristic. Optionally, in addition or alternatively, the electrode stimulation signals Z.sub.l to Z.sub.m may be prior to generating pulse shapes adjusted based on a non-linear mapping that reflects patient-specific scaling from the fitting process, e.g., detection thresholds (THR) and maximum comfortable loudness (MCL) levels.

[0032] A fitting method for use in such a cochlear implant is described in the following in relation to FIGS. 4 to 6. In a first step for developing a virtual channel matrix for mapping analysis channels to stimulation channels for a cochlear implant patient according to the present invention a stimulation channel is selected. The amplitude growth function for the selected stimulation channel is measured in response to a stimulation pulse comprising a negative and positive phase separated in time by a first inter-phase-gap. The stimulation pulse is generated by the implantable part of the cochlear implant after receiving the command and applied to the physical electrode contacts representing the stimulation channel as described before. The stimulation pulse may have the negative phase followed by the positive phase, or the positive phase followed by the negative phase or, in another example may have pulses with ramped amplitudes. Other suitable stimulation pulse shape forms may equally be applied without departing from the invention. Next the amplitude growth function for the stimulation channel in response to a stimulation pulse comprising a negative and positive phase separated in time by a second inter-phase-gap is measured. The first inter-phase-gap and second inter-phase-gap are different. Both slopes for the before measured amplitude growth functions are determined, and an indicator based at least in part on the difference of the slopes of the amplitude growth functions, which is indicative of the local neural survival for the selected stimulation channel, is calculated. The former steps are repeated for each stimulation channel where an indicator shall be derived. Finally, for the virtual channel matrix the stimulation channels with best local neural survival are selected by optimizing a function based at least in part on the calculated indicators of the stimulation channels.

[0033] FIG. 4 shows a stimulation pulse comprising a negative phase 401 and positive phase 402 separated in time by an inter-phase-gap 403 as generated and delivered through the stimulation channel of the electrode array to nearby neural tissue in response to a command to the cochlear implant to apply the stimulation pulse. Thereafter the neural response is recorded through physical electrode contacts, typically adjacent to the physical electrode contacts where the stimulation pulse is applied. From the recorded signal the ECAP magnitude, which is the difference between neural response amplitudes N1 and P2, is estimated. Various known methods may be used, for example those described in the background art above. The stimulation pulse may be a single, charge balanced, rectangular biphasic pulse, i.e. positive and negative phase, where the pulse has an amplitude in a monopolar mode of stimulation. In another embodiment, the pulse shape can also be triphasic, for example compromising a negative phase 401a, followed by a positive phase 402 and a third negative phase 401b, separated in time by two inter-phase-gap intervals 403. In one example, phase durations are uniform across the electrode array and equal to 30 μs per phase unless longer phase durations are required to reach MCL level on one or more electrode contacts. In this case all phase durations are increased to the minimum phase durations required to reach MCL on all electrode contacts. For determining the phase duration prior to recording ECAPs for measuring amplitude growth functions, MCLs on all electrode contacts may be measured for all physical stimulation channels using, for example, 500 ms bursts of stimulation pulses with cathodic leading phase and an inter-phase-gap of 2.1 μs at a rate of 45 stimulation pulses per second. Any other known suitable stimulation pulses for measurement of MCLs may equally be used without departing from the spirit of the invention, including monitoring for maximum loudness during the ECAP measurement procedure. MCLs for virtual stimulation channels may be estimated via linear interpolation from the MCLs of the immediate neighbored physical stimulation channels according the weighting factor a for channel k as follows:

MCL.sub.k=(1−a).Math.MCL.sub.k−1+a.Math.MCL.sub.k+1

This interpolation can be applied to any not measured physical stimulation channels in the same way. Measuring the MCL can be obtained in various ways, including behavioral and objective measurement techniques, for example without limitation by measurement of the stapedius reflex.

[0034] In general, measurement of the amplitude growth function includes measurement of only a part of the amplitude growth function sufficient to determine the slope. In one advantageous embodiment, the measured part of the amplitude growth function is from the largest stimulation pulse amplitude at MCL-level for that stimulation channel toward lower stimulation pulse amplitudes where a reliable an ECAP amplitude can still be measured. In addition, the pulse amplitude may be preferably in the range to capture the linear portion of the amplitude growth function, for example by adjusting the used MCL for amplitude growth measurement. The adjustment from the measured stimulation channel MCL may be dependent on the stimulation pulse inter-phase-gap. The inventors have found, that the adjusted MCL for a stimulation pulse having an inter-phase-gap of 30 μs may be automatically set to a value 20 percent lower than the MCL used for a stimulation pulse having an inter-phase-gap of 2.1 μs, generally without producing uncomfortably loud percepts. In one embodiment, the amplitude growth function for a stimulation channel and a specific inter-phase-gap may be measured at a minimum of five uniformly spaced stimulation pulse amplitudes between 50% and 100% of the stimulation channel MCL. The slopes of the amplitude growth function may be calculated with a linear or nonlinear regression, maximum, mean or minimal slope taken from one or more measured amplitude growth functions per stimulation channel. For nonlinear regression a sigmoidal model function may be used. Various other suitable and known estimation and regression methods may be equally applied to calculate the slope, for example with MED-EL AutoART or as described in EP3104931B1 of applicant with title “Determination of neuronal action potential amplitude based on multidimensional differential geometry”.

[0035] The inventors also found that the calculated indicator indicative of the local neural survival is not affected by whether the stimulation channel is a physical stimulation channel or an virtual stimulation channel. However, for calculating the indicator based at least in part on the difference of the slopes of the amplitude growth functions indicative of the local neural survival for the stimulation channel, the inter-phase-gap for the stimulation pulse shall be at least 2 μs. In another embodiment, the difference between the first and second inter-phase-gaps may be at least 7.9 μs. In a further embodiment, the ratio between the first and second inter-phase-gap may be at least 4. In still a further embodiment the first and second inter-phase-gap may be 2.1 μs and 30 μs, respectively. Maximizing the difference between the first and second inter-phase-gaps may be advantageous in certain cases to improve the accuracy of the estimation of the local neuronal survival and it is thus a further embodiment of the invention to set the inter-phase-gaps accordingly. This may be achieved by finding and setting the inter-phase-gaps such that, the first inter-phase-gap has a duration that is as short as possible, and in one example includes no inter-phase-gap at all, and just enough to allow measurement of an amplitude growth function and/or the second inter-phase-gap has a maximum possible duration and that is at or as close as possible to the maximum acceptable loudness to the patient.

[0036] FIG. 5 shows amplitude growth function slopes estimated from the measured amplitude growth functions with a stimulation pulse having an inter-phase-gap of 2.1 μs and 30 μs respectively over the stimulation channels in the upper diagram. The electrode array has twelve physical electrode contacts numbered from 1 to 12, each representing a physical stimulation channel. The lower diagram shows the difference of the amplitude growth function slopes, indicative of the local neuronal survival over the stimulation channels. Two virtual stimulation channels between neighbored physical stimulation channels with a=0.67 and a=0.33 were measured. The value of the difference of the slopes indicates local neural survival, where “local” means local to the stimulation channel. The higher the value of the difference, the higher the local neural survival for the stimulation channel. For example, the lower diagram shows for physical stimulation channel of physical electrode contact numbers 1 to 11 on the electrode array and virtual stimulation channels with a=0.33 and a=0.67. For example, the virtual channel between physical electrode contacts number 1 and 2 is represented through stimulation channel signals

[00004] $Z = (\begin{matrix} α .Math. Y_{k} \\ (1 - α) .Math. Y_{k} \\ 0 \\ .Math. \\ 0 \\ 0 \end{matrix}) .$

The slope-difference is approximately 0.4 for physical stimulation channels number 1 and 4, suggesting relatively high local neural survival for these two stimulation channels.

[0037] FIG. 6 shows the lower diagram of FIG. 5 where in addition the stimulation channels with highest estimated local neural survival are marked with a circle symbol. Physical stimulation channels are marked with solid black filled triangle symbols and named “Physical electrodes”, while virtual stimulation channels are marked with a triangle symbol. For the virtual channel matrix, the stimulation channels with best local neural survival by optimizing a function based at least in part on the calculated indicators are selected. In one embodiment, optimizing the function includes selecting the stimulation channels where the calculated indicator exceeds a pre-defined threshold. Optimizing this way has the advantage that the highest total overall estimate of neural survival is selected. In another embodiment, optimizing the function includes selecting the stimulation channels that minimize the variance of the calculated indicators, this is

f(S)=min(var(S))

with the variance var(S) and the mean value <(x(S)> defined through

[00005] $var (S) = \frac{1}{.Math. S .Math.} \underset{i \in S}{.Math.} {(x_{i} - .Math. x (S) .Math.)}^{2} .Math. x (S) .Math. = \frac{1}{.Math. S .Math.} \underset{i \in S}{.Math.} x_{i}$

[0038] where S is the set of indicators for selected stimulation channels comprising elements x.sub.i, the calculated indicator for stimulation channel i. First a pre-defined number to be selected stimulation channels is determined, i.e. the cardinality of the set S is determined. Typically, this number corresponds to the physical electrode contacts on the electrode array. Next, those elements for set S may be selected that represent uniformly spaced stimulation channels over the electrode array. The variance var(S) and mean value <x(S)> is evaluated for set S and stored. Set S is modified by replacing elements with the respective neighboring stimulation channel where the indicator is either closer to the mean value or higher. The variance var(S) and mean value <x(S)> is evaluated for the modified set S and compared to the stored variance. If the new variance is smaller than the stored, the new calculated variance replaces the stored one. The steps are repeated until the variance is smallest.

[0039] In a further embodiment, the function further includes at least in part a component on uniform distribution of selected stimulation channels over the frequency range covered by the stimulation channels where an indicator shall be derived. In one example this part of the function may have the following form:

[00006] $f_{d} (S) = .Math. \frac{1}{.Math. S .Math.} \underset{x_{i} \in S}{.Math.} .Math. p (x_{i}) - p (x_{i + 1}) .Math. - (.Math. S .Math. - 1) .Math. Δ .Math.$

where p(x.sub.i) is the position of stimulation channel i on the electrode array and Δ is the physical electrode spacing. The position of virtual stimulation channels is estimated to be between two physical stimulation channels proportional to weighting factor a for that virtual stimulation channel. In this example, optimizing the overall function may include optimizing:

f(S)=min(var(S)+f.sub.d(S))

[0040] In a further embodiment, optimizing the function includes forming stimulation channel groups by grouping a pre-defined number of adjacent stimulation channels and selecting the stimulation channel with the largest calculated indicator in this group, reflecting highest estimated local neuronal survival. FIG. 6 shows a selection based on this embodiment. Each group is formed by a physical stimulation channel and its immediate virtual stimulation channel neighbors. Thus, the group size is three stimulation channels and the number of groups corresponds to the number of physical stimulation channels. For each group from these three stimulation channels, the one with the highest calculated indicator indicative of local neural survival is chosen. This has the advantage, that the entire electrode array is used for stimulation and even regions with poor local neuronal survival remain selected and thus stimulated.

[0041] In a further embodiment, instead of selecting the stimulation channel with the highest calculated indicator indicative of local neural survival, the variance of the calculated indicators from the selected stimulation channels for all groups is minimized. This may be accomplished with the functions as described before, where modifying the set S includes building all permutations of selected channels within the groups. It is readily understood by those skilled in the art, that the inventive fitting method can be exercised in various ways. For example, the fitting method can be started by a clinician or audiologist after initiating a programming session for fitting the cochlear implant system to the patient. In other embodiments, the fitting method can be initiated by the patient itself, for example through pressing a button on the external sound processor. In another example, the inventive fitting method may be also applied in a totally implantable cochlear implant (TICI). In other embodiments, the fitting procedure may be started automatically, for example periodically or based on some measurement initiated through the cochlear implant system itself. If for example the patient has steadily increased volume on its sound processor over time, this may indicate changes in sound perception. Thus, this increase over time may be monitored and may be used to initiate the inventive fitting method. The initiation may be fully automatic or manually through the patient, clinician or audiologist after notifying the patient, clinician or audiologist in various ways of the changes and recommending initiation of re-fitting.

[0042] Embodiments of the invention may be implemented in part in any conventional computer programming language such as VHDL, SystemC, Verilog, ASM, etc. Alternative embodiments of the invention may be implemented as pre-programmed hardware elements, other related components, or as a combination of hardware and software components.

[0043] Embodiments can be implemented in part as a computer program product for use with a computer system. Such implementation may include a series of computer instructions fixed either on a tangible medium, such as a computer readable medium (e.g., a diskette, CD-ROM, ROM, or fixed disk) or transmittable to a computer system, via a modem or other interface device, such as a communications adapter connected to a network over a medium. The medium may be either a tangible medium (e.g., optical or analog communications lines) or a medium implemented with wireless techniques (e.g., microwave, infrared or other transmission techniques). The series of computer instructions embodies all or part of the functionality previously described herein with respect to the system. Those skilled in the art should appreciate that such computer instructions can be written in a number of programming languages for use with many computer architectures or operating systems. Furthermore, such instructions may be stored in any memory device, such as semiconductor, magnetic, optical or other memory devices, and may be transmitted using any communications technology, such as optical, infrared, microwave, or other transmission technologies. It is expected that such a computer program product may be distributed as a removable medium with accompanying printed or electronic documentation (e.g., shrink wrapped software), preloaded with a computer system (e.g., on system ROM or fixed disk), or distributed from a server or electronic bulletin board over the network (e.g., the Internet or World Wide Web). Of course, some embodiments of the invention may be implemented as a combination of both software (e.g., a computer program product) and hardware. Still other embodiments of the invention are implemented as entirely hardware, or entirely software (e.g., a computer program product).

[0044] Although various exemplary embodiments of the invention have been disclosed, it should be apparent to those skilled in the art that various changes and modifications can be made which will achieve some of the advantages of the invention without departing from the true scope of the invention.

Cochlear Implant Fitting Based on Neuronal Status

Inventors

Cpc classification

Classification Explorer

A61N1/36039

HUMAN NECESSITIES

Classification Explorer

A61B5/4848

HUMAN NECESSITIES

Classification Explorer

A61B5/7221

HUMAN NECESSITIES

Classification Explorer

A61N1/0541

HUMAN NECESSITIES

Classification Explorer

A61B2505/09

HUMAN NECESSITIES

Classification Explorer

A61B5/24

HUMAN NECESSITIES

Classification Explorer

A61B5/686

HUMAN NECESSITIES

Classification Explorer

A61B5/6877

HUMAN NECESSITIES

Classification Explorer

A61B5/4047

HUMAN NECESSITIES

Classification Explorer

A61B5/4851

HUMAN NECESSITIES

International classification

Classification Explorer

A61N1/36

HUMAN NECESSITIES

Abstract

Claims

Description