Subthreshold signal encoding in coupled FitzHugh-Nagumo neurons

TL;DR

This study demonstrates that coupled FitzHugh-Nagumo neurons can robustly encode subthreshold signals through spike pattern probabilities, with a second neuron enhancing detection by lowering firing thresholds.

Contribution

It reveals how neuronal coupling preserves and enhances subthreshold signal encoding, extending understanding beyond isolated neuron responses.

Findings

Signal encoding is robust in coupled neurons.

The second neuron facilitates detection by lowering firing thresholds.

Pattern probabilities vary with signal period, peaking near the mean firing period.

Abstract

Despite intensive research, the mechanisms underlying how neurons encode external inputs remain poorly understood. Recent work has focused on the response of a single neuron to a weak, subthreshold periodic signal. By simulating the FitzHugh-Nagumo stochastic model and then using a symbolic method to analyze the firing activity of the neuron, preferred and infrequent spike patterns (defined by the relative timing of the spikes) were detected, whose probabilities encode information about the signal. As not individual neurons in isolation but neuronal populations are responsible for the emergence of complex behaviors, a relevant question is whether this coding mechanism is robust when the neuron is not isolated. We study how a second neuron, which does not perceive the subthreshold signal, affects the detection and the encoding of the signal, done by the first neuron. Through simulations of two coupled FitzHugh-Nagumo neurons we show that the coding mechanism is indeed robust, as the neuron that perceives the signal fires a spike train that has symbolic patterns whose probabilities depend on the features of the signal. Moreover, we show that the second neuron facilitates the detection of the signal, by lowering the firing threshold of the first neuron. This in turn decreases the internal noise level need to fire the spikes that encode the signal. We also show that the probabilities of the symbolic patterns achieve maximum or minimum values when the period of the external signal is close to (or is half of) the mean firing period of the neuron.