Voice gender release from masking in cochlear implant users is correlated with binaural pitch fusion

Abstract

Spatial and voice gender separation of target from masking speech leads to substantial release from masking in normal-hearing listeners. However, binaural pitch fusion is often broad in cochlear implant (CI) listeners, such that dichotic stimuli with pitches differing by up to 3-4 octaves are fused (Reiss et al., 2014). We hypothesized that broad binaural fusion could reduce a listener’s ability to separate competing speech streams with different voice pitches, and thus reduce the voice gender as well as spatial benefit for speech perception in noise. Speech reception thresholds were measured in both bilateral and bimodal CI users, using male and female target talkers at two spatial configurations (co-location and 60-degrees of target-masker separation). Binaural pitch fusion was also measured. Different-gender maskers improved target detection performance in bimodal CI users, and performance was better with female than male targets in bilateral CI users. No spatial benefit was seen in either CI group. As hypothesized, voice gender masking release was strongly correlated with binaural fusion range in bimodal CI users. These results suggest that sharp binaural fusion is necessary for maximal speech perception in noise in bimodal CI users, but does not benefit bilateral CI users. [Work supported by NIH-NIDCD grant R01 DC01337.]Spatial and voice gender separation of target from masking speech leads to substantial release from masking in normal-hearing listeners. However, binaural pitch fusion is often broad in cochlear implant (CI) listeners, such that dichotic stimuli with pitches differing by up to 3-4 octaves are fused (Reiss et al., 2014). We hypothesized that broad binaural fusion could reduce a listener’s ability to separate competing speech streams with different voice pitches, and thus reduce the voice gender as well as spatial benefit for speech perception in noise. Speech reception thresholds were measured in both bilateral and bimodal CI users, using male and female target talkers at two spatial configurations (co-location and 60-degrees of target-masker separation). Binaural pitch fusion was also measured. Different-gender maskers improved target detection performance in bimodal CI users, and performance was better with female than male targets in bilateral CI users. No spatial benefit was seen in either CI group. As...