Sound-processing strategy for cochlear implants

Abstract

A sound processing method for auditory prostheses, such as cochlear implants, which is adapted to improve the perception of loudness by users, and to improve speech perception. The overall contribution of stimuli to simulated loudness is compared with an estimate of acoustic loudness for a normally hearing listener based on the input sound signal. A weighting is applied to the filter channels to emphasize those frequencies which are most important to speech perception for normal hearing listeners when selecting channels as a basis for stimulation.

Description

TECHNICAL FIELD [0001] The present invention relates to a sound processing strategy for use in hearing prosthesis systems, with particular application to cochlear implant systems. BACKGROUND ART [0002] In cases where individuals have experienced sensorineural deafness, the restoration of hearing sensations to such individuals has been achieved through the use of hearing aids and cochlear implants. Cochlear implants in particular have been in clinical use for many years. A wide variety of different speech processing strategies have been employed in order to process a sound signal into a basis for electrical stimulation via implanted electrode arrays. Some systems have focused upon extracting particular acoustic components of the detected sound signal, which are important to the user's understanding of speech, for example the amplitudes and frequencies of formants, and using these as a basis for generating stimuli. Other approaches have also attempted to utilise the generally tonotopi...

AI Technical Summary

Benefits of technology

[0010] According to another aspect, the input sound signal is processed to determine an overall loudness estimate for a hypothetical listener with normal hearing. After the parameters of the electrical stimuli are determined according to the stimulation scheme employed, the loudness perceived by the implant user with the proposed stimuli is estimated, based on parameters including the previous stimuli applied, and relevant characteristics of the patient's auditory perception with electric stimulation which have been previously determined clinically. If the loudness of the proposed stimuli is not the same as that for a normally hearing person within a predetermined range, then the stimuli are adjusted and the loudness estimated again, until the range is met. This may be termed normalising the overall loudness as perceived by a listener using electric hearing. Preferably, the input values for the electrical stimuli are initially determined using an established sound processing scheme for cochlear implants, such as the SPEAK scheme described above. However other schemes may also be employed, including schemes which generate simultaneous or analog patterns of stimulation, rather than stimulation using sequences of rectangular pulses which do not overlap in time. The intention of this approach is to adjust the electric stimulation so that the overall loudness of the user's percept is comparable to that of a normal hearing listener for the same input sound signal, including taking account of the specific user's characteristics. This approach is particularly applicable when it is desired to produce overall loudness which is not identical to that perceived by a hypothetical normal hearing listener, but has a predetermined relationship to normal loudness. For example, it may be desirable to compress the range of loudness levels perceived using the implant compared with the normal range so as to reduce the effects of background noise, or to enhance speech intelligibility. This approach also provides signals which better emphasise the signals known to be most important to speech perception in normally hearing people.

[0011] The present invention also attempts to provide a scheme that can improve the control of not just the overall loudness of signals perceived by cochlear implant users but also to improve the control of the relative loudness of signals presented to implant users, particularly the relative loudness of different components of speech signals, such as phonemes. This is done by using a version of the present scheme which controls the distribution of loudness contributions across frequency or across cochlear position, rather than controlling the overall loudness. This aspect will be described in more detail below.

Problems solved by technology

The problem with this method of loudness-mapping is that is does not take into consideration the effects of loudness summation when multiple electrodes are activated in quick succession, as they generally are in the output of speech processors.

This loudness summation leads to the situation that the output of the processor is too loud, even though the individual levels on each electrode do not exceed a comfortable loudness.

These methods, although alleviating the discomfort of implant users for loud sounds, do not address a second important issue, and that is the impact of loudness summation on speech perception.

The present loudness coding methods, whereby the acoustic output of a filter is mapped to a fixed range of electrical levels (however determined) on its corresponding electrode, lead inevitably to a perceptual distortion of the amplitude envelope shape because these methods do not take into account the variations from moment to moment of important aspects such as the number of electrodes activated in each stimulus cycle, and the relative loudness contributions from these other electrodes.

In summary, the relative loudness of electrically stimulated hearing using present approaches does not accurately convey the relative loudness that a normally-hearing person would hear for the same acoustic input.

As well as distorting the perception of the amplitude envelope of the acoustic signal, this effect will lead to narrow-band signals being masked by lower-level broad-band noise, thus disrupting the ability of implantees to understand speech in background noise.

Images