Unravelling the Neural Firing Properties Underlying Human eCAPs through an Extended Deconvolution Method

Yu Dong Jeroen Briaire Johan Frijns
Department of Otorhinolaryngology, Leiden University Medical Center, Leiden, The Netherlands

Background:

The electrically evoked compound action potential (eCAP) has been widely investigated for its clinical value in cochlear implants (CIs). However, there is still much uncertainty about the firing properties underlying eCAPs. Strahl et al. (2016) developed a method to calculate the compound discharge latency distribution (CDLD, consisting of a two-component Gaussian mixture model) in humans and guinea pigs by deconvolution using a unitary response (UR) from a guinea pig. Here, an extended deconvolution method was used to further explore the firing properties of eCAPs.

Objective:

The goal of this research is to develop a new deconvolution method to explore the firing properties underlying human eCAPs.

Methods:

In our initial experiment, a human UR (5 additional parameters) and CDLD were derived simultaneously by deconvolution of intraoperatively measured eCAPs of 111 CI-recipients. Next, we deconvolved the recorded eCAPs by using this UR to obtain a parameterized CDLD. Then we investigated the firing properties of eCAPs from the CDLD.

Results:

With the human UR obtained in the initial experiment, consistent CDLDs could be estimated. The area under the CDLD, a measure of the number of excited nerve fibres, was significantly correlated with stimulus level (r=0.93, p<0.01) and eCAP amplitudes (r=0.515, p<0.01). CDLD latencies were negatively correlated with eCAP amplitudes (r1=-0.3, r2=-0.25; both p<0.01).

Conclusions:

With an improved version of Strahl’s deconvolution method, we derived a UR of human auditory nerve fibres and obtained a more realistic CDLD, reflecting the number of excited nerve fibres and their latencies.









Powered by Eventact EMS