Enhancement of Signal Detection and Information Transmission of Retinal Ganglion Cell Model with Stochastic Resonance

Jing Wu,Qingli Qiao,Menghua Jin
DOI: https://doi.org/10.16289/j.cnki.1002-0837.2017.05.001
2017-01-01
Abstract:Objective In order to apply stochastic resonance (SR) to epiretinal prosthesis stimulation,the effects of different noises on the responses of retinal ganglion cells (RGC) to electrical stimulation and the information transmission were explored.Methods The RGC was modeled as a multi-compartment model of extracellular stimulation.The model was stimulated by the sub-threshold stimulus with Ornstein-Uhlenbeck (OU) processes or Gaussian white noise.The detection of sub-threshold signal was determined by observing RGC firing.The power norm and mutual information rate (MIR) were used to characterize SR.Results When a certain intensity of noise was added to the sub-threshold signal,RGC could fire.With the increasing of noise intensity of both noises,the curve of power norm and MIR variation presented the inverted U-like graph.Compared to the Gaussian white noise,the optimal noise intensities in the OU process were smaller,but the maximum power norm and MIR were larger.Conclusion There are SR phenomena in the RGC.SR can improve and expand the detection and information transmission of sub-threshold signals in the RGC.It is easier to produce SR phenomenon in the RGC by OU process and the effect is better.These studies have a practical significance to reduce epiretinal prosthesis stimulation threshold.
What problem does this paper attempt to address?