Applying the Deconvolution Method in Nucleus CIs to Better Characterize Spread of Excitation

Background: Spread-of-excitation (SOE) in cochlear implants can be estimated objectively by measuring electrically evoked compound action potentials (eCAPs) with forward masking. Biesheuvel et al. (2016, Ear Hear, 37, 572‑581) proposed that the amplitude of these eCAPs is the result of the convolution of the excitation pattern of the fixed probe with that of the roving masker. We report preliminary findings using the deconvolution technique developed by Biesheuvel et al. to analyze SOE data collected with Nucleus CI recipients to produce excitation density profiles (EDP).

Methods: Subjects were Nucleus CI recipients in which SOE patterns were acquired. Both masker and probe stimuli were presented at all odd-numbered electrode contacts. Probe and masker current levels (CL) were set at the same value for all recordings starting with 225 CL units, going to 235 CL based on the amplitude of the response. This produced an 11 by 11 matrix of N₁‑P₁ eCAP amplitudes for analysis using the method described by Biesheuvel et al.

Results: Preliminary data with five cases indicates that the deconvolution technique is applicable to Nucleus CIs. The EDPs were relatively heterogeneous across electrodes and across patients: three cases showed higher amplitudes with wider EDPs for more apical electrodes, and one case the reverse. Further data will be presented at the time of the conference.

Conclusions: Preliminary results indicate that the deconvolution technique is practical and may be applied using Nucleus CIs. The technique opens the way to a better characterization of neural responses and the potential for improved diagnostics.