Neural correlates of speech in noise perception differences between combined electric-acoustic stimulation and standard cochlear implants

Abstract

Cochlear implants have evolved to utilize residual acoustic hearing that combines to electric stimulation, known as electric-acoustic stimulation (EAS). However, there are mixed expectations about the benefits of EAS. A positive perspective expects that contributions from residual hearing provide better access to acoustic cues that is helpful for speech-in-noise perception. An opposing view concerns potentially poorer spectral resolution of EAS electrodes’ stimulation as those electrodes are often inserted close to lateral wall, which may cause poorer speech-in-noise perception. This study aimed to directly compare neural processes of speech-in-noise perception between EAS and standard CI users to provide an answer to the above alternative expectations. We used 64-channel EEG to measure cortical evoked responses to (1) background noise and (2) target word while listeners perform a word-in-noise task. Then, we compared the amplitude ratio of evoked responses to the target word and background noise, referred to as “internal SNR,” which reflects how well target sound is unmasked from the mixture of speech and noise. Based on the comparison of 55 EAS and 22 standard CI users, internal SNR was significantly larger in EAS CI users. This result indicates that EAS provides enhanced neural processes for speech unmasking.