Reaction time as a measure of spectrotemporal modulation sensitivity in Nucleus cochlear implant users

Cochlear implant users can achieve high levels of speech recognition despite severe hearing loss. However, outcomes vary considerably, and some users continue to struggle. This study assessed Nucleus cochlear implant users’ sensitivity to spectrotemporal modulations, important acoustic features for speech recognition, using a reaction-time task with ripple sounds. Ripples contain similar spectrotemporal dynamics as speech, without resembling words, and allow precise control over spectral and temporal parameters. Cochlear implant users’ reaction times to ripple onsets varied systematically with modulation characteristics: temporal modulations around 16 Hz consistently yielded the shortest reaction times, while reaction times increased with increasing spectral density. These effects of modulation were nearly identical across participants, indicating a similar tuning to the tested spectrotemporal features. In contrast, the average reaction time, across all modulations, differed substantially between participants, ranging from 200 ms to about 1 second. These individual differences correlated modestly with phoneme recognition scores (r = 0.52), indicating that overall reaction times for the tested modulations relates to speech processing ability. These two findings (consistent modulation-dependent reaction times across participants, and their modest correlation with phoneme recognition) indicate that reaction times reflect underlying auditory sensitivity to spectrotemporal modulations. The substantial unexplained variance in phoneme recognition, however, suggests that modulation sensitivity alone is not sufficient to account for individual differences in speech understanding. This points to an additional role for non-acoustic, cognitive factors. Further, modeling results also suggest that individual differences in reaction times can be attributed to a combination of both sensory factors (i.e., modulation sensitivity) and non-auditory influences, such as decision-making processes.