A WITHIN-SUBJECT COMPARISON OF ADULT PATIENTS USING THE NUCLEUS F0F1F2 AND F0F1F2B3B4B5 SPEECH PROCESSING STRATEGIES

This study compares the Nucleus F0F1F2 and F0F1F2B3B4B5 (also known as ''Multipeak'' or ''Mpeak'') processing schemes in 17 patients wearing the Mini Speech Processor. All patients had at least 18 months implan t experience using the F0F1F2 processing strategy. For this study, the y were switched to the F0F1F2B3B4B5 processing strategy for 3 months. They then returned to using the F0F1F2 strategy for 3 months, then use d the F0F1F2B3B4B5 strategy again for 3 months, and lastly used the F0 F1F2 strategy for 3 months. Performance was evaluated with both scheme s after each interval, using speech recognition tests and subjective r atings. Overall, differences between the results for the two processin g schemes were not large. Average performance was somewhat better for the F0F1F2B3B4B5 strategy for word and sentence identification, but no t for any of the other speech measures. Superior performance was obser ved in 8 patients with the F0F1F2B3B4B5 strategy. However, 6 of the 8 individuals were significantly better on only one of the six speech me asures in the test battery. The other 2 patients performed better on t wo of the speech measures. Superior performance was also observed in 3 patients with the F0F1F2 strategy for consonant recognition. For the remaining patients, there was little difference in their performance w ith the two strategies. Information transmission analyses indicated th at the F0F1F2B3B4B5 strategy transmitted consonant duration and fricat ion cues more efficiently than F0F1F2. Experience with one strategy ap peared to benefit performance with the other strategy.