Comparison of cochlear implant performance to pitch shift with audiovocal feed-back in postlingual cochlear implant patients

Auditory feedback requires normal auditory and laryngeal system in normal hearing subjects. Various subjective and objective test are used for determining auditory performance of postlingual cochlear implant patients. Objective tests are generally focused on cortical potentials, but they do not show the integrity of the auditory tractus. In this study, audio-vocal feed-back is analyzed in cochlear implant patients and the difference in frequency response shifts in control and patient group is investigated. It has also been studied, whether there is a correlation between frequency shifts and speech discrimination scores. The aim was to establish a test battery which will anticipate implant performance and show the real condition of central auditory processing systems. Nine postlingually implanted patients and nine controls of similar ages were included in the study. The subjects' /ah/ vocalizations were recorded and processed for through custom software. The recorded voices were pitch-shifted for one octave and half an octave both to lower and higher frequencies and were presented to the subject with a 5 sec delay. Fundamental frequency (F0) and strongest formant (Fx) frequency values before and after pitch-shifting was analyzed and compared between two groups. F0 response to half octave pitch-shifted voice was the only statistically different parameter between two groups (p=0,040). Fx response to half octave pitch-shifted voice was closer to siginificant values than other measurements (p=0,11). It has also been investigated whether there is a difference in auditory feed-back responses with regard to speech discrimination scores (SDS). In patients with SDS over 70%, F0 and Fx responses to minus half octave and minus one octave pitch shifting were siginificantly different than controls. The responses to plus half octave and plus one octave pitch shifted voice were similar to the control group. However, the responses to plus half octave and plus one octave pitch shifted voice were siginificantly different than control group in patients with low SDS. According to these results, it is hypothetised that if responses to pitch-shifted voice in both ways is taken into consideration during preoperative evaluation, more significant predictive results can be obtained. However, changes in formant frequencies should also be evaluated in time axis in miliseconds. Objective data collected in terms of cortical potentials, activated zones in fMRI, etc. and in synchronization with voice analysis will determine the origin of voluntary audio-vocal response more accurately.