Enhancement of spike coherence by the departure from Gaussian noise in a Hodgkin-Huxley neuron

Abstract

Experimental study has shown that non-Gaussian noise exists in sensory systems like neurons. The departure from Gaussian behavior is a characteristic parameter of non-Gaussian noise. In this paper, we have numerically studied the effect of a particular kind of non-Gaussian colored noise (NGN), especially its departure q from Gaussian noise (q = 1), on the spiking activity in a deterministic HodgkinHuxley (HH) neuron driven by sub-threshold periodic stimulus. Simulation results show that the departure q can affect the spiking activity induced by noise intensity D. For smaller q values, the minimum in the variation coefficient (CV) as a function of noise intensity (D) becomes smaller, showing that D-induced stochastic resonance (SR) becomes strengthened. Meanwhile, depending on the value of D, q can either enhance or reduce the spiking regularity. Interestingly, CV changes non-monotonously with varying q and passes through a minimum at an intermediate q, representing the presence of "departure-induced SR". This result shows that appropriate departures of the NGN can enhance the spike coherence in the HH neuron. Since the departure of the NGN determines the probability distribution and hence may denote the type of the noise, "departure-induced SR" shows that different types of noise can enhance the spike coherence, and hence may improve the timing precision of sub-threshold signal encoding in the HH neuron.neuron, non-Gaussian colored noise, spike coherence, stochastic resonance, stochastic dynamics