Effects of electrode location on speech recognition with the Nucleus-22 cochlear implant.

Speech recognition performance was measured as a function of electrode in two experiments with the Nucleus-22 cochlear implant using 4-electrode SPEAK speech processors. In experiment 1, the four stimulated electrode pairs were shifted in 0.75-mm steps over 3 mm in the apical-basal direction. In experiment 2, the four electrodes were closely spaced and positioned apically, medially, or basally. An additional condition spaced the four electrodes as widely as possible. In experiment 1, City University of New York sentence scores showed a significant decrease in performance as the electrodes were shifted basally; no other speech measures showed a significant change with electrode location. For experiment 2, all scores were the best with the processor that had the electrodes spaced as widely as possible. In both experiments, all 4-electrode SPEAK processors produced significantly poorer speech recognition than the subject's own 20-electrode processor. These results indicate that the location of electrodes is an important factor in implant performance.