BINAURAL DICHOTIC PRESENTATION FOR MODERATE BILATERAL SENSORINEURAL HEARING-IMPAIRED
Alice N. Cheeran, Prem C. Pandey · 2004
For persons with moderate bilateral sensorineural hearing loss, binaural dichotic presentation of speech has helped in improving speech perception by reduction the effect of increased spectral and temporal masking. Speech processing schemes, which split the speech into two for binaural dichotic presentation, are evaluated on normal subjects with simulated hearing loss, and on hearing-impaired subjects with moderate to severe sensorineural hearing loss. Spectral splitting with perceptually balanced comb filter, temporal splitting with trapezoidal fading functions of inter-aural switching periods of 20, 40, and 80 ms, combined splitting with cyclically swept comb filter of sweep cycle of 20, 40, 80 ms and 4, 8 and 16 shiftings is implemented and evaluated by conducting listening tests, which involved recognition of phonetically balanced monosyllables. After the initial round of experiments on seven normal hearing subjects with hearing loss simulated to different levels, processing conditions which provided least improvement were omitted for hearing-impaired subjects. For all the processing conditions considered, the speech was passed through an audiogram dependant filter with gain varying between +3 dB and −3 dB to partially compensate for the frequency dependant shifts in hearing threshold separately for right and left ears, before doing the processing for spectral, temporal, and combined splitting with different conditions, for hearing-impaired subjects. Interaural switching period of 20 ms gave better results in temporal splitting, while sweep cycle of 40 ms was preferred in combined splitting. More improvement was for spectral splitting for maximum number of subjects.