Place-pitch and vowel-pitch comparisons in cochlear implant patients using the Melbourne-Nucleus cochlear implant.

Results of place-pitch and vowel-pitch comparisons are presented in 21 cochlear implant patients using the Melbourne-Nucleus cochlear implant. Vowel-pitch comparisons were also carried out in 10 normal hearing subjects. A technique for the place-pitch ranking test has been developed. A graphic representation of the results shows the well-ranked electrodes in sequential pitch-order, and reveals any indication of abnormal place-pitch perception. It aids the selection of correctly place-pitch ranked electrodes. The vowel-pitch comparisons showed that both normal hearing subjects and cochlear implant patients are able to rank vowels according to 'vowel-pitch'. In normal hearing subjects, three main types of vowel-pitch processing have been found. Results indicate that an information selection and reduction process occurs at higher levels along the auditory pathway. Cochlear implant patients test results showed the limited contribution of the first and the virtual lack of the second formant's contribution to pitch-ranking the voiced vowels. These results indicate that fundamental frequency converted to pulse rate may not be adequate at certain segments along the cochlear partition. Vowels are not perceived by cochlear implant patients according to their first or second formant frequency converted to place-pitch. There would seem to be a need for alternative speech processing strategies in the Melbourne-Nucleus implant.