Acoustic Change Complexes Recorded in Adult Cochlear Implant Listeners

Abstract

Objectives: 

The objectives of this study were to determine: 1) whether the acoustic change complex (ACC) could be reliably recorded in cochlear implant listeners and, 2) whether different speech sounds evoke distinct ACC patterns.

Design: 

Eight adults wearing the Nucleus-24 cochlear implant (CI) were tested using naturally produced speech tokens /si/ and /ʃi/. Stimuli were tokens from the standardized UCLA version of the Nonsense Syllable Test. Using a repeated-measures design, participants were tested and retested within a 3-wk period.

Results: 

Intraclass correlation coefficients for grand mean and individual-response waveforms recorded from the syllables /si/ and /ʃi/ ranged from 0.63 to 0.89 from test to retest. Also, ACC latencies signaling the onset of a vowel in /ʃi/ were significantly earlier than those evoked by /si/.

Conclusions: 

The ACC can be reliably recorded in individuals wearing CI. Furthermore, the naturally produced speech syllables /si/ and /ʃi/ evoke distinct ACC patterns. Because of its good stability and the ease with which it can be recorded in individual CI listeners, the ACC can be evoked using complex signals (such as naturally produced speech syllables) when studying central auditory function in CI listeners.

In Brief

Most cortical evoked potential studies typically use only one sound (short duration clicks, tones, and synthesized speech tokens (e.g., /da/)) to test all subjects. A limitation of this approach is that /da/ might not be representative of the types of perceptual errors made by all individuals tested. Consequently, when group comparisons are made, between normal hearing and cochlear implant (CI) listeners for example, individual variability is not reflected. We therefore set out to determine if different naturally produced speech sounds evoked distinct neural response patterns in individual CI listeners, and if the evoked patterns can be reliably recorded, from one session to another. Using a repeated measures design, acoustic change (P1-N1-P2) responses were recorded from participants wearing Nucleus-24 devices. CI listeners were tested and retested within a 3-wk period using two different naturally produced speech syllables. Results demonstrate that naturally produced speech stimuli evoke distinct neural response patterns in individual CI listeners. Moreover, these unique patterns can be reliably recorded in individuals, from one session to another. For these reasons, we believe acoustic change responses have the potential to be used to examine the neural detection of time-varying cues in individuals who wear cochlear implants.