Binaural intelligibility level difference with a mixed-rate strategy simulation

Abstract

In noisy situations, bilateral cochlear implant (BiCI) listeners demonstrate poorer spatial release from masking compared to normal hearing (NH) listeners. One reason for this difference could be the limited access for BiCI listeners to fine structure interaural time differences (ITDs) which are crucial for binaural unmasking. Clinical processors convey speech envelopes using pulsatile high-rate (~1000 Hz) carriers, and discard the original temporal fine structure. At high rates, BiCI listeners are not sensitive to ITDs, whereas low rates (< 300 Hz) result in better ITD sensitivity. To preserve both ITD sensitivity and well-sampled speech envelopes, high and low rates are needed. The present study examines whether a mixed rate strategy will enable binaural unmasking in NH listeners using a 16-channel mixed-rate vocoder. Binaural intelligibility level differences were calculated from speech reception thresholds measurements in five mixed-rate configurations. Targets had a 0-μs ITD and masking speech-shaped noises were co-located or separated by applying a 400-μs ITD to the low-rate channels. Results from this experiment will help determine the potential benefits of a mixed-rate strategy to restore spatial unmasking abilities of CI listeners in noisy situations.