Electric to acoustic pitch matching: a possible way to improve individual cochlear implant fitting.

Poor pitch resolution has been shown to have negative implications for speech and music perception in implanted patients. Surprisingly, works on the subject have not focused much on the impact that the non-correspondence between frequencies allocated to electrodes and perceived frequencies could have on speech and music perception. The aim of the present study is to investigate the correlation between pitch mismatch and speech performance with the implant, and to ascertain the effects of mismatch correction through a mapping function making a personalized frequency reallocation possible. We studied ten postlingually deaf adult patients with detectable bilateral residual hearing, implanted in our Clinic with Cochlear Nucleus devices. In each test session, we asked the patients to find the best match between the pitch elicited by the residual ipsilateral and contralateral pure tones and the pitch elicited by stimulation of electrodes. We also assessed patients' vowel and consonant recognition performance. Finally, in the only implanted patient in our clinic who had bilateral residual hearing and used a Digisonic DX10/C device, which makes manual electrode-by-electrode frequency reallocation possible, we modified electrode-assigned frequency ranges on the basis of the pitch matching test results. We found that in none of the studied patients, the electric-to-acoustic pitch matching corresponds to the theoretical assignment pattern. A very strong correlation was detected between the electric-to-acoustic pitch mismatch and patient's speech performance. In the Digisonic patient, a remarkable improvement in all phoneme recognition scores was obtained 1 month after frequency reallocation. In the light of our results, we propose to assess, whenever possible, any frequency-to-electrode mismatch in all implanted patients, and correct it through mapping programs allowing manual frequency reallocation for the pitch-matched electrodes, and automated allocation of the non-tested electrodes. Cochlear implantation should therefore be proposed when residuals for all frequencies are still present, at least in one ear, so as to allow optimal alignment between allocated and subjectively perceived frequencies.