Comparing Bilateral and Single-Sided Deaf Cochlear Implant Recipients in a Novel Speech-in-Noise and Localization Task

Objective: This study investigated how individuals who have undergone bilateral cochlear implantations (BiCIs) use compensatory head movements to optimize sound localization and speech-in-noise (SIN) performance compared with single-sided deaf users of cochlear implant (SSDCI) and normal hearing controls.

Study Design: Nonrandomized, prospective human-subject study.

Setting: Tertiary academic medical center.

Methods: Subjects were presented with Harvard IEEE sentences at varying signal-to-noise ratios (SNRs) in a darkened, semianechoic chamber. An electromagnetic head-tracking system monitored head movement. Outcomes included head movement patterns, localization accuracy, and SIN performance.

Results: BiCI localized targets less accurately with higher SNR-dependent variability than controls, but with greater accuracy than SSDCI. For SIN, SSDCI consistently showed superior performance to the BiCI, with no statistical significance found at any noise condition between SSDCI and controls. Across SNR, BiCI consistently initiated head movements more promptly compared to SSDCI, with controls predictably exhibiting the least delay. All CI recipients exhibited greater absolute displacement overall compared to controls. Although BiCI showed greater displacement compared to SSDCI, their pattern resembled controls, with decreased movement necessary as SNR became more favorable.

Conclusion: When comparing performance for localization and SIN across groups, the same impaired group does not perform most poorly at both binaural tasks; although SSDCI users maintain better speech understanding in noise, BiCIs tend to perform better on locating stimuli in space. Preserved acoustic hearing in one ear does not provide a performance advantage across all tasks. The study underscores the intricate interplay of adaptive behaviors in CI users, showcasing successes and challenges in optimizing binaural hearing performance.