Simulating the Effects of Neural Pathology on Cochlear Implant Responses

Abstract

Cochlear implants (CIs) provide a strategy for treating severe to profound sensorineural hearing loss (SNHL) that cannot be adequately managed by amplifying acoustic signals. There is considerable variability of CI effectiveness among bilaterally implanted patients with similar pre-operative audiograms, and the factors underlying this variability are not completely understood (Blamey et al., 2012; Lazard et al., 2012; Orabi et al., 2006). Some individuals achieve near perfect performance on tasks requiring speech recognition in quiet while others receive far less benefit (Lazard et al., 2012). Binaural hearing-dependent tasks, such as localizing sound sources based on interaural timing and level cues or discerning speech in noise, are particularly variable even in the research setting with optimized delivery of fine temporal structure information (Kan and Litovsky, 2014; Litovsky et al., 2012). Understanding the biological mechanisms of this variability may enable the design of individually-tailored stimulation strategies and improve prognostics. Temporal bone studies in humans and animal models describe both significant, diameter-dependent degeneration of spiral ganglion neurons (SGNs) and demyelination of remaining peripheral axons (Incesulu and Nadol, 1998; Leake and Hradek, 1988; Seyyedi et al., 2014; Shepherd et al., 2004). The present work explores the question of whether pathological demyelination and fiber diameter-dependent degeneration of spiral ganglion neuron processes following loss of afferent input from hair cells can explain particular features of experimental inter-animal and patient-to-patient variability in outcomes to electrical stimulation. These pathological manifestations were incorporated into population-scale biophysical computer simulations of implant stimulation and shown capable of reproducing key dimensions of observed variability at multiple scales, including model animal single-unit response timing, human inter-aural timing difference detection, and electrically-evoked, intracochlearly-recorded potentials. These findings suggest that both asymmetric loss and microarchitectural changes of SGNs may shape the temporal dynamics of population responses and may place fundamental restrictions on the performance of implants in some individuals.