Leslie M. Collins
Dept. of Elec. Eng. and Comput. Sci., 1301 Beal Ave., Univ. of Michigan, Ann Arbor, MI 48109
Teresa A. Zwolan
Gregory H. Wakefield
Univ. of Michigan, Ann Arbor, MI 48109
Results from previous experiments with multichannel cochlear implant subjects [Collins et al., Midwinter Research Meeting, Association for Research in Otolaryngology, February, 1994] have shown that removal of indiscriminable electrodes from subjects' maps resulted in significant improvements in speech perception scores in some of the better subjects. The procedure used to evaluate electrode discrimination in this previous study lacked clinical applicability due to the amount of time required for data collection. In the present paper several measures of electrode discrimination, including an adaptive discrimination procedure, a pitch ranking procedure, and a paired comparison procedure are evaluated and compared. The clinical applicability of each procedure will be discussed as it relates to both efficiency and impact on mapping for improved speech perception, as measured by both open and closed set tests. In order to evaluate the effect of remapping on speech perception, information transmission analyses were performed on vowel and consonant confusion data from each subject. These data were obtained while subjects used a standard map employing all active electrodes, and while they used maps based on the results of each electrode discrimination procedure. [Research supported by NIH.]