Signal selective amplification for below-the-threshold stimulus in Fitzhugh-Nagumo neuronal model

TL;DR

This paper demonstrates that noise shaping in the FitzHugh-Nagumo neuronal model enables efficient amplification of subthreshold signals and spike excitation, revealing a novel selective amplification mechanism driven by complex variable interactions.

Contribution

It introduces the first demonstration of noise shaping enabling subthreshold signal amplification and spike excitation in the FitzHugh-Nagumo model, highlighting a new amplification process.

Findings

Noise shaping enables efficient subthreshold signal amplification.

Complex interplay between FHN variables leads to noise accumulation at spiking frequencies.

Tails follow a power-law frequency scaling.

Abstract

Brain operates at remarkably low signal power. It has been noted that noise may play a constructive role in neural networks and facilitate the subthreshold signaling. The process of spiking pattern excitation at the characteristic neuronal spiking frequency from the random noisy stimulus remains unexplained. Furthermore, recent research indicates that neuronal processing enables signal selective amplification. Also, there is a growing number of studies indicating the role of coloured noise for spike generation, while noise shaping effect have been recently observed in neuronal networks as well as in single neurons. We demonstrate for the first time that noise shaping enables efficient signal amplification and spike excitation for below-the-threshold stimulus. On the example of the seminal FitzHugh-Nagumo (FHN) model we reveal that the complex interplay between the fast and slow FHN variables leads to noise accumulation at the FHN characteristic spiking frequencies, while tails follow a power-law frequency scaling. We demonstrate that the discovered effect allows selective amplification of signal waveforms presenting a novel type of amplification process.