Perception of Spectrally Enhanced Speech through Companding in Individuals with Cochlear Hearing Loss

Abstract

The present study aimed to assess the benefit of spectral contrast enhancement using companding strategy in individuals with cochlear hearing loss. T1venty adult subjects having normal hearing and cochlear hearing loss participated in the study. Consonant recognition scores (quiet, 15 dB SNR, 10 dB SNR mrd 0 dB SNR) and sentence recognition threshold in noise was obtained in unprocessed (original) and processed (spectrally enhanced using companding strategy) condition. Numbers of correctly recognised consonants were calculated for the consonant recognition task. The SNR level at which50 % of the correct responses were obtained using sentences (SNR-50), across different SN Rs (+20 dB to -10 dB SNR) was found. Results showed that processed condition of consonant recognition scores lead to higher performance than unprocessed in. both the groups. Significant improvement was found at 0 dB SNR (12%) for normal hearing individuals and at 15 dB (4.5 o/o ), 10 dB (5.25%) and 0 dB SNR ( 18. 75 o/o) for those with cochlear hearing loss. In sentence recognition threshold in noise (SNR-50) task, both the subjects performed at lower SNR levels in processed than unprocessed condition. improvement found was about -3.75 dB SNR for normal hearing individuals and -5 dB SNR for those with cochlear hearing loss. Thus, it can be concluded that spectrally enhanced speech through companding strategy improves the speech perception in noise in individuals with cochlear hearing loss to a much greater extent than do for normal hearing individuals.