The Relationship Between Speech Discrimination in Noise and Spectral Resolution, Temporal Resolution, and Melodic Contour Identification Abilities in Cochlear Implant Users.

Purpose: Speech-in-noise performance of cochlear implant (CI) users varies considerably, and understanding speech in a complex auditory environment remains challenging. It is still unclear which auditory skill is causing this difficulty. This study aimed to evaluate spectral resolution, temporal resolution, and melodic contour identification (MCI) skills to determine which of these skills is most closely related to speech understanding in noise and to investigate whether these three skills differ among CI users with varying performances in speech-in-noise tasks.

Method: The three-syllable word identification test (in quiet), the Turkish Matrix Test (in quiet, at fixed +8 dB SNR, and adaptively in noise) were administered to evaluate speech perception to 24 adult CI users (18-55 years old). To evaluate auditory abilities, tests involving spectral-temporally modulated ripple, gap detection, and MCI were administered. CI recipients were separated into two groups (good and low performers) based on their adaptive matrix speech reception threshold (SRT) in noise.

Results: The abilities of temporal and spectral resolution were correlated with speech understanding at +8 dB SNR. However, the effect of spectral resolution on understanding at +8 dB SN was not significant. The skill most associated with the Matrix SRT in noise was spectral resolution. The groups differed significantly in both the speech tests and spectral resolution skills.

Conclusions: Varying spectral resolution skills among CI users may affect their speech perception performance in noise. The effects of spectral and temporal resolution on speech perception in noise varied across noise conditions. Thus, there may be an increased requirement for spectral resolution in more adverse listening environments.