DISCRIMINATION OF MULTICHANNEL-COMPRESSED SPEECH IN NOISE - LONG-TERMLEARNING IN HEARING-IMPAIRED SUBJECTS

Objective: It takes time for an individual to obtain optimal benefit f rom a new hearing aid. This research examines the possibility that sim ilar longterm learning can be seen in consecutive laboratory studies o f multichannel compression (MCC). Design: Three studies of different p arameters of MCC processing, carried out over the period of 1 yr, incl uded the same 15 hearing-impaired subjects and one identical MCC-proce ssing condition. The full-range MCC had 8, 12, or 16 independent frequ ency channels, using a Robinson-Huntington compression algorithm. The City University of New York nonsense syllable test was modified to fac ilitate digital signal processing and control of the experiments. The subjects discriminated nonsense syllables (a female and a male voice) in speech spectrum noise at -5 to 15 dB signal-to-noise ratios (S/N). Conditions were not ideal for learning: subjects' experience with MCC- processed speech was limited to the laboratory and no trial-by-trial f eedback was provided. Percent correct syllable discrimination and cons onant confusion matrices were compared across experiments to observe t he subjects' learning to listen with MCC processing. Results: All subj ects combined, and 14 of 15 individual subjects, showed significant im provement across experiments. For the subject showing the maximum lear ning, the percent correct difference between the first and third exper iments was equivalent to a 9.8 dB increase in S/N. The average learnin g for all subjects was equivalent to +3 dB S/N. The difference between the consonant-confusion matrices for the first and third experiments indicated that improved discrimination occurred for both manner and pl ace information. The pattern of changes in the confusion matrices was consistent with improved use of the high-frequency information supplie d by the MCC signal processing. A brief comparison of the results of t he first experiment with a fourth experiment indicated that the learni ng was specific to MCC processing because it did not generalize to fre quency-shaped linear amplification which was also studied in those two experiments. Conclusions: These results indicate that specific long-t erm learning occurred for hearing-impaired subjects listening to nonse nse syllables in noise with 8- to 16-channel MCC processing. Since pre vious experiments have provided subjects with much less listening expe rience, the results suggest that MCC with large numbers of channels ma y be much more beneficial for the hearing-impaired individuals than th e results of previous experiments had indicated.