Measuring input synchrony in the Ornstein–Uhlenbeck neuronal model through input parameter estimation

Abstract

We present a method of estimating the input parameters and through them, the input synchrony, of a stochastic leaky integrate-and-fire neuronal model based on the Ornstein–Uhlenbeck process when it is driven by time-dependent sinusoidal input signal and noise. By driving the neuron using sinusoidal inputs, we simulate the effects of periodic synchrony on the membrane voltage and the firing of the neuron, where the peaks of the sine wave represent volleys of synchronised input spikes. Our estimation methods allow us to measure the degree of synchrony driving the neuron in terms of the input sine wave parameters, using the output spikes of the model and the membrane potential. In particular, by estimating the frequency of the synchronous input volleys and averaging the estimates of the level of input activity at corresponding intervals of the input signal, we obtain fairly accurate estimates of the baseline and peak activity of the input, which in turn define the degrees of synchrony. The same procedure is also successfully applied in estimating the baseline and peak activity of the noise.This article is part of a Special Issue entitled Neural Coding 2012.