[Modification of speech encoding process of single channel cochlear implant utilizing digital delay circuit].

Speech perception by single channel cochlear implant can be improved to some extent by modifying the encoding process of the speech signals. The principle of our encoding strategy is to present speech signals with appropriate delay time for lower frequency components of the speech sound, while without delay for higher frequency components. This is based on a cochlear physiology that higher frequency sounds are perceived earlier than lower frequency sounds, since the former is perceived at the basal turn and the latter at the apical turn. The following tests were performed without the aid of lip reading in the two patients implanted with a single channel cochlear implant of 3M-House design. Lower frequency components of the speech signals were digitally delayed with a Sony SDP-777 ES surround processor. The results indicated that the optimal delay time for identification of /i/, /e/ and /u/ needed longer delay time, while that of /a/ and /o/ need shorter delay time, corresponding to the difference between the first and second formant of each vowel. Comparison of vowel perception with and without delay coding showed that vowel identification with delay coding was superior to that without it; 30% vs. 16% in case 1, and 48% vs. 30% in case 2.