Self pitch matching as basis for acoustic coordinated reset neuromodulation

Abstract

An illustrative method includes determining a preliminary estimate of the patient's tinnitus tone by allowing the patient to listen to and vary frequency and intensity of a first test tone within a predetermined first frequency range and a predetermined first intensity range to approximate the first test tone to the patient's tinnitus tone; defining a second frequency range and a second intensity range around the preliminary estimate of the patient's tinnitus tone; and determining a final estimate of the patient's tinnitus tone by allowing the patient to listen to and vary frequency and intensity of a second test tone within the second frequency range and the second intensity range to approximate the second test tone to the patient's tinnitus tone.

Claims

1. A method of determining a pitch of a patient's tinnitus tone, the method comprising the steps: presenting a first test tone to the patient; varying within a predetermined first frequency range or around a predetermined first intensity range a frequency or an intensity of the test tone by the patient to approximate the first test tone to the patient's tinnitus tone, thus providing a tinnitus tone; defining a second frequency range and a second intensity range around a preliminary estimate of the patient's tinnitus tone; determining a final estimate of the patient's tinnitus tone by allowing the patient to listen to and vary frequency and intensity of a second test tone within the second frequency range and the second intensity range to approximate the second test tone to the patient's tinnitus tone, wherein the method is implemented on a digital processing unit of a device comprising an output unit for delivering the test tones to the patient and a user interface for variation of the frequency and the intensity of the test tones by the patient.

2. The method of claim 1, wherein the user interface comprises a touch screen defining a two-dimensional plane, the touch screen configured so that sliding in a first dimension leads to an increase or decrease of the frequency and sliding in a second dimension leads to an increase or decrease of the intensity.

3. The method of claim 1, wherein the first frequency range is different to the second frequency range and/or wherein the first intensity range is different to the second intensity range.

4. The method of claim 1, wherein the first frequency range and the first intensity range are determined based on a masking level of the patient, wherein the masking level is the lowest intensity of a given frequency band required to mask the patient's tinnitus tone.

5. The method of claim 4, wherein the masking level is determined by presenting at least one frequency band individually to the patient and increasing intensity of each of the at least one frequency band by the user until the patient detects that one of the at least one frequency band masks the patient's tinnitus tone.

6. The method of claim 1, wherein intensities within the first intensity range and/or intensities within the second intensity range are scaled relative to same perceived intensities, wherein the same perceived intensities are intensities, which the patient perceives of substantially the same intensity at frequencies within the first frequency range and/or at frequencies within the second frequency range.

7. The method of claim 6, wherein the same perceived intensities are determined based on the patient's threshold of perception of a plurality of test tones covering a plurality of frequencies adjacent and/or within the first frequency range and/or covering a plurality of frequencies adjacent and/or within the second frequency range.

8. The method of claim 1, further comprising: determining a deviation of the preliminary and the final estimate of the patient's tinnitus tone and determining reliability of the final estimate of the patient's tinnitus tone based on the deviation, preferably a sum of a factorized deviation in a frequency and intensity space, wherein the reliability of the final estimate of the patient's tinnitus tone is defined to be higher the smaller the deviation, or vice versa.

9. The method of claim 1, further comprising: performing at least one pair-wise testing, wherein in each of the at least one pair-wise testing the patient is allowed to listen to one pair of tones, the one pair of tones consisting of the final estimate of the patient's tinnitus tone and a test tone different thereof, and to choose the tone which closer matches the patient's tinnitus tone.

10. The method of claim 9, wherein the an intensity of one or more of the test tones is adapted to have the same perceived intensity with the final estimate of the patient's tinnitus tone.

11. The method of claim 9, wherein the method is registered as passed when the patient has chosen the final estimate of the patient's tinnitus tone as the tone which closer matches the patient's tinnitus tone in the majority, preferably in all, of the at least one pair-wise testing.

12. A non-transitory storage medium storing instructions that are executable by a digital processing device to perform the method of claim 1.

13. (canceled)

14. A digital processing device comprising: an output unit for delivering test tones to a patient; an user interface for varying frequency and intensity of the test tones by the user; and a control unit in functional communication with the output unit and the user interface, the control unit comprising a digital processing unit for performing the method of claim.

15. The digital processing device of claim 14, wherein the user interface comprises a touch screen defining a two-dimensional plane, which is configured so that sliding in a first dimension leads to an increase or decrease of the frequency and sliding in a second dimension leads to an increase or decrease of the intensity.

Description

[0043] Further aspects, embodiments and advantages of the present invention become apparent from the following detailed description, the figures and claims, which follow after the brief description of the drawings.

[0044] In the drawings:

[0045] FIG. 1 illustrates a masking level-based approach for determining first intensity and frequency ranges according to one embodiment of the present invention.

[0046] FIG. 2 illustrates an approach for determining first intensity and frequency ranges based on threshold of perception determined for different frequencies, according to one embodiment of the present invention.

[0047] FIG. 3 illustrates a technical implementation allowing the patient to vary frequency and intensity simultaneously in a 2D plane, according to a preferred embodiment of the invention.

[0048] FIG. 4 illustrates an approach to define the second frequency and intensity ranges based on the preliminary estimate of the patient's tinnitus tone, according to one embodiment.

[0049] FIG. 5 illustrates a general sequence of the method of the invention according to one embodiment.

[0050] FIG. 6 illustrates functionalities of the present invention according to various embodiments of the invention, implemented as a computer program (app) executed by a digital processing device (smartphone).

[0051] The present invention allows a health care professional or patient perform an accurate pitch matching within a short period of time, and provide input for neuromodulation-based therapies configured to treat subjective tinnitus. In the following, selected features and embodiments of the present invention as disclosed herein are described.

[0052] The method of the present invention (also referred to herein as pitch matching tool) is preferably implemented on a digital processing device (herein also referred to as electronic device or device), which has an appropriate user interface and high-quality sound output. As user interface, a touch screen with case sensitive user interfaces ideally support the pitch matching process. The present invention is not restricted to particular devices; most smartphones have both a high-resolution touch screen and a high-quality sound output, which renders them suitable for the present invention. The device may offer a comprehensive training and help function, to allow that an untrained patient can use the device without expert assistance.

[0053] As the patient may perceive different frequencies as of different intensity, information how intense a patient perceives different intensities may be collected in an initial step. To possible approaches are described in the following.

[0054] A first approach is based on band-pass filtered white noise. Although one may use 10, 12 or up to 27 noise bands, it is preferred that 5 noise bands are used. This allows for better handling to be done by the patient itself and at home. The 5 noise bands may be defined by the following low and high cut-offs: [0055] Band 1: 100-3.700 Hz [0056] Band 2: 3.700-5.300 Hz [0057] Band 3: 5.300-7.700 Hz [0058] Band 4: 7.700-12.000 Hz [0059] Band 5: 12.000-15.500 Hz

[0060] The stopband attenuation may be 60 dB. The term stopband defines the difference in power of the actual passband and the lower level outside the passband.

[0061] A (slight) overlap is possible. In particular, the used filters to realize the noise bands may not have an abrupt reduction of power to the stopband level at e.g. 3.700 Hz for the first band. Instead, the filter may have certain characteristics resulting in a certain slope of the spectrum resulting in an overlap.

[0062] The patient may then be exposed to each of these noise bands and asked to slowly increase the volume. If the tinnitus is in the frequency range of the noise band, there is a high probability that the patient's tinnitus is masked, i.e. no longer perceptible, by this noise band. The lower and upper frequency limits of the noise band where this masking was observed can then be used to determine the lower and upper frequency limits of the first frequency range. For instance, the lower and upper frequency limits F.sub.1, F.sub.2 of the noise band may by extended by 30% as shown in the example illustrated in FIG. 1. In this example, the minimal intensity I at which masking has been detected is extended by ?5 dB and +20 dB to define the first intensity range. On the right, the first intensity and frequency ranges are plotted in a 2D plane, which, in preferred embodiments, illustrate a screen, preferably a touch screen, on which the patient can subsequently vary frequency and intensity of the first test tone by moving a button (e.g. a smaller circle) on the touch screen.

[0063] According to a second approach (herein also referred to as audiogram-like test), the patient identifies the threshold of perception for several test tones. The test tones might range from 200 Hz to 10.000 Hz; for example tones with frequencies 200 Hz, 400 Hz, 1.000 Hz, 2.000 Hz, 4.000 Hz, 6.000 Hz, 8.000 Hz, 10.000 Hz can be used. Even higher frequencies (up to 14.000 Hz) can be used, but many patients might not be able to perceive such tones, even at highest sound output of the device. The threshold of perception for the different frequencies thus determined can then be used to determine the first frequency and intensity ranges.

[0064] For instance, as shown in the example illustrated in FIG. 2, the determined frequency-dependent threshold of perception (amplitude in %) is plotted over the respective frequency to define a curvature. This curvature shows that a frequency of ca. 4.000 Hz is perceived at a relatively low amplitude, whereas the patient hears lower and higher frequencies only at higher amplitudes. Moreover, the curvature is used to determine an upper and a lower corridor, which respective end points define the first intensity range. The lower corridor is defined by the curvature minus a certain dB value (5 dB in the example of FIG. 2), and the upper corridor is defined by the curvature plus a certain dB value (20 dB in the example of FIG. 2).

[0065] In some cases, it might be advantageous to allow the device to play the test tones, in particular, frequencies above 10 kHz, with maximum output volume, so that hearing-impaired patients may still hear the test tones.

[0066] The above measurement is preferably done for the left and the right ear separately. The output might be threshold information in dB SPL, dB HL or % sound output. For the functioning of the present invention, even the simplest realization (% sound output) would result in a good frequency match of the tinnitus.

[0067] Before starting with the first step, the patient may select, if the assessment is done for the left or right ear alone, or for both ears at the same time. Then, the patient moves a button on a two-dimensional plane (e.g. by moving the patient's finger on a touch screen). A first direction (e.g. x-direction) indicates the intensity of the tone, a second direction (e.g. y-direction) indicates the frequency of the first test tone. As output, either sinusoidal tones, or narrow bandwidth tones, or peaked noisy tones are used. The minimal and maximal intensity of the first test tones should be chosen to maximally support the patient in this process. A simple choice would be 0% intensity on the left boundary and 100% intensity on the right boundary (cf. FIG. 3).

[0068] A more complex, but better choice would be to use the audiogram-like test result to define the boundaries. E.g. the left (minimal value) boundary could be represented by threshold minus 5 dB sound intensity and the right (maximal value) boundary could be represented by the threshold plus 20 dB sound intensity. The boundaries may be defined for each frequency step independently, using the audiogram-like test information and interpolation. Referring again to FIG. 2, a determination of the first frequency and intensity range can be based on the threshold of perception determined for different frequencies. The x-axis of the graph on the left illustrates the intensity in terms of the amplitude (as shown in FIG. 2), whereas the x-axis of the 2D plane on the right illustrates the perceived intensity. This means, the frequencies lying on a vertical line are perceived by the patient of the same intensity.

[0069] For frequencies above 10.000 Hz it might be advantageous to use 100% as maximal output, therefore, the right boundary may smoothly increase from the 10.000 Hz maximal boundary (threshold plus 20 dB) to 100% output. If the test is performed for the left and right ear at the same time (stereo), the intensity boundaries are independently determined and the first intensity range is independently defined for the left ear and the right ear, resulting in the left and right ear to be balanced in intensity.

[0070] In an alternative approach, the psychoacoustic loudness scaling is envisaged. For example, if the hearing threshold of left and right differs more than a certain value to one frequency or few frequencies, e.g. by 15 dB, the pitch matching for both ears at the same time is not available and separate pitch matching has to be done for left and right.

[0071] Alternatively, the psychoacoustic loudness scaling is considered to adjust the upper intensity boundary in such a way, that the left and right tone is perceived as of same intensity even if there is a significant difference in the hearing threshold.

[0072] With respect to the first frequency range, the minimal and maximal frequency is chosen to cover the possible tinnitus frequencies, for example 200 Hz could be chosen as lower limit (bottom) and 14.000 Hz could be chosen as upper limit (top) (cf. FIG. 3).

[0073] The intensity scale may be a linear scale but is preferably a logarithmic scale, such that similar distances on the two-dimensional plane would be perceived as the same intensity changes. The frequency scale may be a linear scale but is preferably a logarithmic scale, such that similar distances on the two-dimensional plane would be perceived as the same frequency changes.

[0074] When the patient takes his finger from the button (e.g. by detaching the finger from the touch screen), the tone will stop. The tone will start again, when the patient touches the button. The start and stop of the tone may be smoothed by a short fade out and fade in. This function will allow the patient to concentrate on his tinnitus for few seconds, before the comparison with the test tone is continued.

[0075] The output of the first step is a first information about the tinnitus frequency and tinnitus intensity, herein referred to as the preliminary estimate of the patient's tinnitus tone.

[0076] In a second step, the frequency limits are chosen to be around the frequency result of the first step (cf. FIG. 4).

[0077] For example, two octaves below and two octaves above the first frequency information can be chosen as lower and upper frequency limit, respectively. Alternatively, one octave below and one octave above the first frequency information can be chosen as lower and upper frequency limit, respectively. More narrow arrangements around the first frequency information might be possible as well. The intensity boundaries might be the same as for the second step, or could be narrowed around the first intensity information as well. For example, minus and plus 5 dB around the first intensity information could be chosen. The new boundaries can be slightly modified from being symmetrically arranged around the first frequency and intensity information (cf. FIG. 3) to avoid that the first information (frequency/intensity) would correspond to the middle position of the two-dimensional plane. With such variation, it could be avoided, that the patient could manipulate the results.

[0078] FIG. 5 is intended to illustrate a general sequence according to one embodiment of the invention. According to this sequence, the masking levels are determined, and the first intensity and frequency ranges are determined based on the masking levels. The first intensity and frequency range are depicted on a screen as two axes in a first 2-D plane, on which the patient can select the tone which comes closest to the patient's tinnitus tone. This tone is then used to define second intensity and frequency ranges, which are then depicted in a second 2-D plane, on which a further round of selection is made.

[0079] The patient is then asked in the third step to find its dominating tinnitus frequency on this modified two-dimensional plane. The outcome of the third step will be a second frequency/intensity information (i.e. final estimate of the patient's tinnitus tone). It is assumed, that the second frequency/intensity information is potentially more precise, since the frequency and intensity boundaries of the two-dimensional plane are usually narrower and thus allow a finer selection. Therefore, the second frequency/intensity information will be used as the final estimate.

[0080] In some embodiments, a quality of the result is determined. For instance, the deviation of the first and second information may be used to define the quality of the result. If both information (preliminary and final estimate) is very similar, the information is considered of high quality, if the information differs strongly, the quality is considered of low quality.

[0081] Quality can be measured on different scales, e.g. on a scale from 0% quality to 100% quality. Different implementations of the quality measure are possible and may be tested clinically. As one example, the quality could be defined by the sum of the factorized deviation in the frequency and intensity space, using the following equation:

Q=100%?Difference(F1,F2)?0.5*Difference(I1,I2), [0082] where the difference is defined as Difference(Z,W)=(Z?W)/((Z?FW)/2)) [%].

[0083] The quality may be further based on the patient's evaluation of the quality of the pitch matching. The evaluation may be conducted by asking the patient to answer one or more questions, for example, how good the patient considers the degree of match of the final estimate to the patient's tinnitus tone. As a non-limiting example, the factor reflecting the patient's evaluation may be defined by the following equation:

(?12.5*Z+50), [0084] where Z=1, 2, 3 or 4 depending on the patient's answer (1: Quality was rated as very good, 4: Quality was rated as poor.)

[0085] The overall quality according to the above example could then be calculated as follows:

Q=100%?Difference(F1,F2)?0.5*Difference(I1,I2)?(?12.5*Z+50), [0086] where the parameters are defined as above.

[0087] In a further step, the resulting information (i.e. the final estimate of the patient's tinnitus tone) can be compared with several other test tones. Therefore, a pairwise comparison may be done. The tones that are compared to the final estimate resulting from step three could be: one octave up and one octave down (to test for octave confusion), 1.4 times the second frequency, 0.766 times the second frequency, 1.1 times the second frequency, 0.9 times the second frequency, 0.95 times the second frequency or 1.05 times the second frequency. Other test tones might be possible as well. The corresponding intensities are chosen such that the test tones and the tinnitus tone are perceived as of similar intensity (e.g. based on the audiogram-like test information). The patient may be allowed to listen to pairs of tones, one is a test tone and the other one is the result from the third step (i.e. the final estimate). After listening to both tones, the patient identifies the tone that fits his tinnitus best. Only if the frequency and intensity testing of the third step was properly done and resulted in a frequency/intensity that corresponds to the patient's tinnitus tone, the patient is assumed to be able to identify this tone as the tone that fits the tinnitus best in all the pair-wise comparisons.

[0088] The further step may serve as a validation step. For example, only if all pair-wise comparisons resulted in the frequency and intensity testing of step three, this step would be registered as passed. It is also possible to define the validation to be passed if only one pair-wise comparison testing failed (not resulting in frequency and intensity testing of step three).

[0089] After the validation step, the final result may be displayed and stored in the device. The final result may consist of the tinnitus information, namely frequency and intensity, the quality of the measurement (in %) and/or the result of the pair-wise comparison test (failed or passed).

[0090] Referring to FIG. 6, functionalities of the present invention according to various embodiments of the invention are illustrated. The functionalities may be part of a method as disclosed herein implemented as a computer program (app) executed by a digital processing device (smartphone). When the app is started, the patient may be guided through some functionalities along a predetermined path, whereas other functionalities may occur only when a patient selects them by pressing on a respective button on the screen. A first exemplary functionality may allow the patient to select between the pitch matching method as disclosed herein and an audiogram method, as illustrated in FIG. 6a. After selecting the pitch matching method, the patient may be asked to indicate, which ear(s) are to be analyzed (FIG. 6b). Before starting the pitch matching method as such, an instruction screen may pop up, briefly explaining the patient what to do (FIG. 6c). FIG. 6d shows an exemplary screen, on which the patient can vary the frequency and intensity of the first test tone in order to determine the preliminary estimate. FIG. 6e shows a corresponding exemplary screen for determining the final estimate. FIG. 6f shows an exemplary way that allows the patient to conduct the pair-wise testing. Afterwards, the results of the pitch matching method may be shown (FIG. 6g). FIG. 6h illustrates an exemplary screen that allows the patient to review previous pitch matching results.