Amplitude mapping and phoneme recognition cochlear implant listeners

Objective: Speech and other environmental sounds must be compressed to acco mmodate the small electric dynamic range in cochlear implant listeners. The objective of this paper is to study whether and how amplitude compression and dynamic range reduction affect phoneme recognition in quiet and in nois e for cochlear implant listeners. Design: Four implant listeners using the Nucleus-22 SPEAK speech processor participated in this study. The amount of compression was varied by manipul ating the Q-value in the SPEAK processor. The size of the dynamic range was systematically reduced by increasing the threshold level and decreasing th e comfortable level in the processor. Both female- and. male-talker vowel a nd consonant materials were used to evaluate speech recognition performance in quiet and in noise. Speech-spectrum-shaped noise was mixed with the spe ech signal and presented continuously to the speech processor through a dir ect electric connection. Signal to noise ratios were changed over a 30 to 4 0 dB range, within which phoneme recognition increased from chance to asymp totic performance. Phoneme recognition scores were obtained as the number o f active electrodes was reduced from 20 to 10 to 4. For purposes of compari son, phoneme recognition data also were collected in four normal-hearing li steners under comparable laboratory conditions. Results: In both quiet and noise, the amount of amplitude compression did n ot significantly affect phoneme recognition. The reduction of dynamic range marginally affected phoneme recognition in quiet, but significantly degrad ed phoneme recognition in noise. Generally, the 20- and 10-electrode proces sors produced similar performance, whereas the 4-electrode processor produc ed significantly poorer performance. Compared with normal-hearing listeners , cochlear-implant listeners required higher signal to noise ratios to achi eve comparable recognition performance and produced significantly lower rec ognition scores at the same signal to noise ratios. Conclusions: The amount of amplitude compression does not significantly aff ect phoneme recognition, whereas reducing dynamic range significantly lower s phoneme recognition, particularly in noise and for vowels. Because the SP EAK processor extracts mostly spectral peaks, the present conclusions may n ot be applied to other types of processors extracting temporal envelope cue s. The present results also suggest that more than four electrodes are requ ired to optimize speech recognition in multiple-talker and noise conditions . A significant performance gap in speech recognition still remains between cochlear implant and normal-hearing listeners at the same signal to noise ratios. Improved cochlear implant designs and fitting procedures are requir ed to narrow and, hopefully, close this performance gap.