Speech recognition under conditions of frequency-place compression and expansion

In cochlear implants, the length and the insertion depth of the electrode array determine the cochlear tonotopic range of stimulation, and the speech processor controls the mapping of acoustic frequency information onto this range. Conventional electrode arrays, 16 mm in length, stimulate a cochlear region corresponding to an acoustic frequency range of 500–6000 Hz. However, some implant speech processors map an acoustic frequency range from 150 Hz to 10,000 Hz onto these electrodes. While this mapping preserves the entire range of acoustic frequency information, it also results in a compression of the tonotopic pattern of speech information delivered to the brain. The present study measured the effects of such a compression of frequency-to-cochlear-place mapping on speech recognition, as well as the effects of an expansion. Such an expanded representation of speech in the cochlea might improve speech recognition by improving the relative spatial (tonotopic) resolution, like an “acoustic fovea.” Phoneme and sentence recognition scores were measured as a function of compression and expansion with normal-hearing listeners using a noise-band vocoder, and with implant users by changing the programs in their implant processors. The analysis frequency range was either compressed or expanded relative to the cochlear tonotopic range while the tonotopic range was held constant. Speech recognition in matched conditions was generally better than compression and expansion, even when the matched condition eliminated a considerable amount of acoustic information. It was also more beneficial to match the frequencies that contribute more to speech information. The results suggest that speech recognition, at least without training, is dependent on the mapping of acoustic frequency information onto the appropriate cochlear place. Further experiments detailed the trade-off between information loss and the accuracy of the cochlear location where the acoustic information is presented. A minor modification of the classic Speech Intelligibility Index model was able to account for the drop in scores due to the mismatch of the frequency to place. Overall results provide guidelines for achieving an optimal frequency-place map for implant users.