Comparison of the F0F2 and F0F1F2 processing strategies for the Cochlear Corporation cochlear implant

In April 1985, an updated processing strategy became available for the Cochlear Corporation Nucleus cochlear implant. Whereas the original strategy codes only fundamental frequency, amplitude, and information in the second formant region, the newer strategy also codes frequencies in the first formant region. This investigation evaluated the speech recognition skills of five subjects who were experienced with both designs. On average, the addition of first formant information improved word identification in an audition-only condition and improved spondee recognition in noise. Scores for the NU 6 Monosyllabic Word Test and the Sentence Test Without Context improved from 8% (2-12%) to 28% words correct (10-42%), and from 31% (10-45%) to 64% words correct (39-84%), respectively. Scores for the Four-Choice Spondee Test in noise improved from 37% (25-50%) to 75% (45-90%). The percent correct scores for the Iowa 14-Item Consonant Confusion Test in an audition-only and vision-plus-audition condition did not vary with the change in processing strategy. However, an information transfer analysis performed on the responses to the consonant test in a vision-only and a vision-plus-audition condition suggested that the newer strategy enhances the transmission of the voicing, duration, and envelope features.