Speech Perception Performance in Cochlear Implant Recipients Correlates to the Number and Synchrony of Excited Auditory Nerve Fibers Derived From Electrically Evoked Compound Action Potentials
Yu Dong,Jeroen J Briaire,H Christiaan Stronks,Johan H M Frijns
DOI: https://doi.org/10.1097/AUD.0000000000001279
Abstract:Objectives: Many studies have assessed the performance of individuals with cochlear implants (CIs) with electrically evoked compound action potentials (eCAPs). These eCAP-based studies have focused on the amplitude information of the response, without considering the temporal firing properties of the excited auditory nerve fibers (ANFs), such as neural latency and synchrony. These temporal features have been associated with neural health in animal studies and, consequently, could be of importance to clinical CI outcomes. With a deconvolution method, combined with a unitary response, the eCAP can be mathematically unraveled into the compound discharge latency distribution (CDLD). The CDLD reflects both the number and the temporal firing properties of excited ANFs. The present study aimed to determine to what extent the CDLD derived from intraoperatively recorded eCAPs is related to speech perception in individuals with CIs. Design: This retrospective study acquired data on monosyllabic word recognition scores and intraoperative eCAP amplitude growth functions from 124 adult patients with postlingual deafness that received the Advanced Bionics HiRes 90K device. The CDLD was determined for each recorded eCAP waveform by deconvolution. Each of the two Gaussian components of the CDLD was described by three parameters: the amplitude, the firing latency (the average latency of each component of the CDLD), and the variance of the CDLD components (an indication of the synchronicity of excited ANFs). Apart from these six CDLD parameters, the area under the CDLD curve (AUCD) and the slope of the AUCD growth function were determined as well. The AUCD was indicative of the total number of excited ANFs over time. The slope of the AUCD growth function indicated the increases in the number of excited ANFs with stimulus level. Associations between speech perception and each of these eight CDLD-related parameters were investigated with linear mixed modeling. Results: In individuals with CIs, larger amplitudes of the two CDLD components, greater AUCD, and steeper slopes of the AUCD growth function were all significantly associated with better speech perception. In addition, a smaller latency variance in the early CDLD component, but not in the late, was significantly associated with better speech recognition scores. Speech recognition was not significantly dependent on CDLD latencies. The AUCD and the slope of the AUCD growth function provided a similar explanation of the variance in speech perception (R 2 ) as the eCAP amplitude, the slope of the amplitude growth function, the amplitude, and variance of the first CDLD component. Conclusion: The results demonstrate that both the number and the neural synchrony of excited ANFs, as revealed by CDLDs, are indicative of postimplantation speech perception in individuals with a CI. Because the CDLD-based parameters yielded a higher significance than the eCAP amplitude or the AGF slope, the authors conclude that CDLDs can serve as a clinical predictor of the survival of ANFs and that they have predictive value for postoperative speech perception performance. Thus, it would be worthwhile to incorporate the CDLD into eCAP measures in future clinical applications.