Effects of electrode-to-fiber distance on temporal neural response with electrical stimulation

This paper presents an analysis of the effects of the electrode-to-fiber distance on the temporal response properties of an auditory nerve fiber stimulated by electric current pulses. This analysis was based upon results from a computational model of a mammalian auditory nerve fiber axon having 50 nodes of Ranvier, each consisting of 130 stochastic sodium channels and 50 stochastic potassium channels, making it possible to represent the temporal fluctuations of action potential initiation and conduction. A monopolar stimulus electrode was located above a central (26th) node at electrode-to-fiber distances of 1, 4, and 7 mm, while the recording electrode was located at the 36th node. Action potentials (spikes) were generated by the biophysical model using the Crank-Nicholson method to solve a diffusive partial differential equation. By observing the occurrence times of spikes in response to 2000 cathodic monophasic stimulus pulses, temporal jitter (i.e., the standard deviation of spike times) was calculated and the poststimulus time (PST) histogram was generated as well. Furthermore, by computing the PST histogram for each initiation node as functions of space (node number) and time (PST), it was shown that spike initiation was distributed not only spatially but also temporally for stimulus levels producing firing efficiencies (FEs) near 0.5. However, at levels producing FEs near 0.99, while temporal variations approached zero, the spatial distribution of initiating nodes was comparable to that observed for the FE near 0.5. As temporal fluctuations are important for speech coding in cochlear implants, we conclude that spatial characteristics of the electrode-auditory nerve fiber interface may play a significant role in influencing these stochastic temporal processes.