TY - JOUR
T1 - Spatial release from masking in simulations of cochlear implants for single-sided deafness
AU - Bernstein, Joshua G.
AU - Iyer, Nandini
AU - Brungart, Douglas
PY - 2013
Y1 - 2013
N2 - Previous studies have shown that single-sided-deaf (SSD) patients implanted with cochlear implants (CIs) can receive spatial release-from-masking (SRM), likely due to head-shadowing effects. This study investigated the possibility that SSD-CI patients might also obtain SRM benefits from binaural-integration cues when the target speech is masked by a spatially-separated fluctuating masker. Experiment 1 measured psychometric functions for word-recognition performance in the presence of stationary noise, modulated noise, and one or two same- or opposite-gender interfering talkers. The first ear received an unprocessed mixture containing the target and masker(s). The second ear received no signal (NH/Deaf), an unprocessed mixture containing just the maskers (NH/NH), or a mixture containing just the maskers that was processed with an eight-channel noise vocoder (NH/Vocoder). The results show that SRM occurred in the NH/NH condition for all masker types, but that it only occurred in the SSD/Vocoder condition with same-gender interfering talkers. Experiment 2 revealed that SRM occurred in the NH/Vocoder condition with as few as two vocoder channels, and that maximum performance occurred with six or more channels. These results suggest that CIs for SSD have the potential to produce SRM in situations where monaural cues are insufficient for concurrent speech-stream segregation.
AB - Previous studies have shown that single-sided-deaf (SSD) patients implanted with cochlear implants (CIs) can receive spatial release-from-masking (SRM), likely due to head-shadowing effects. This study investigated the possibility that SSD-CI patients might also obtain SRM benefits from binaural-integration cues when the target speech is masked by a spatially-separated fluctuating masker. Experiment 1 measured psychometric functions for word-recognition performance in the presence of stationary noise, modulated noise, and one or two same- or opposite-gender interfering talkers. The first ear received an unprocessed mixture containing the target and masker(s). The second ear received no signal (NH/Deaf), an unprocessed mixture containing just the maskers (NH/NH), or a mixture containing just the maskers that was processed with an eight-channel noise vocoder (NH/Vocoder). The results show that SRM occurred in the NH/NH condition for all masker types, but that it only occurred in the SSD/Vocoder condition with same-gender interfering talkers. Experiment 2 revealed that SRM occurred in the NH/Vocoder condition with as few as two vocoder channels, and that maximum performance occurred with six or more channels. These results suggest that CIs for SSD have the potential to produce SRM in situations where monaural cues are insufficient for concurrent speech-stream segregation.
UR - http://www.scopus.com/inward/record.url?scp=84878985531&partnerID=8YFLogxK
U2 - 10.1121/1.4800211
DO - 10.1121/1.4800211
M3 - Conference article
AN - SCOPUS:84878985531
SN - 1939-800X
VL - 19
JO - Proceedings of Meetings on Acoustics
JF - Proceedings of Meetings on Acoustics
M1 - 050087
T2 - 21st International Congress on Acoustics, ICA 2013 - 165th Meeting of the Acoustical Society of America
Y2 - 2 June 2013 through 7 June 2013
ER -