Pitch matching of electric and acoustic stimuli

Abstract

In the electric coding of speech for multiple-electrode cochlear implants, acoustic frequency ranges are mapped onto electrodes. The question arises as to whether the pitches of the electrically evoked hearing sensations are similar to those evoked by the corresponding acoustic stimuli in normal-hearing listeners. Obviously, the sensations are similar enough for many postlingually deaf implant users to understand speech with a minimum of retraining, but it is unlikely that the electric signals sound identical to the acoustic ones. There will also be differences between implant users arising from the variable insertion depth of the electrode array, the number of electrodes in use, and the frequency-to-electrode mapping. The most direct method of determining pitch is to ask implant users to compare electric and acoustic stimuli, but studies of this sort have been hampered by the fact that very few implant users have usable hearing for acoustic signals. In 1978, Eddington et al 1 reported pitch-matching results for one unilaterally deaf volunteer. They concluded that pitch matching was "roughly consistent with electrode position and tonotopic maps of the cochlea derived from basilar membrane motion and hearing loss measurements."

Several other studies 2-6 have investigated the relative pitch of electric signals using identification, scaling, and discrimination paradigms. These studies have established that electrode placement, electrode configuration, and rate of stimulation all affect the perceived pitch, and that the pitch increases tonotopically from apical to basal electrode positions. They have not determined the pitch of electric stimuli in an absolute fashion that can be compared with acoustic stimuli, however. A knowledge of the absolute pitch of electric stimuli for individuals, or as a function of position in the cochlea, would be very useful in optimizing the frequency mapping for cochlear implants.

The present study directly compared the pitch of acoustic pure tones in one ear with electric signals in the other ear of implant users with some residual hearing in the nonimplanted ear. The main questions addressed were whether the pulse rate of a matched electric stimulus would be equal to the frequency of the acoustic tone, and whether the electrode used in the matched stimulus would correspond in position to the place of maximum basilar membrane motion produced by the acoustic tone in a normal cochlea.