Information on mean, fluctuation and synchrony conveyed by a population of firing neurons

TL;DR

This study investigates how a population of firing neurons encodes information through mean rate, fluctuation, and synchrony, using a generalized rate-code model with noise, and shows these components can carry independent information.

Contribution

The paper introduces an augmented moment method to analyze neuronal ensemble responses, demonstrating the independence of information carried by mean rate, fluctuation, and synchrony.

Findings

AMM results agree with direct simulations

ICA shows independent information channels

Higher sensitivity of mean rate to multiplicative noise

Abstract

A population of firing neurons is expected to carry not only mean firing rate but also its fluctuation and synchrony among neurons. In order to examine this possibility, we have studied responses of neuronal ensembles to three kinds of inputs: mean-, fluctuation- and synchrony-driven inputs. The generalized rate-code model including additive and multiplicative noise (H. Hasegawa, Phys. Rev. E {\bf 75}, 051904 (2007)) has been studied by direct simulations (DSs) and the augmented moment method (AMM) in which equations of motion for mean firing rate, fluctuation and synchrony are derived. Results calculated by the AMM are in good agreement with those by DSs. The independent component analysis (ICA) of our results has shown that mean firing rate, fluctuation (or variability) and synchrony may carry independent information in the population rate-code model. The input-output relation of mean firing rates is shown to have higher sensitivity for larger multiplicative noise, as recently observed in prefrontal cortex. A comparison is made between results obtained by the integrate-and-fire (IF) model and our rate-code model. The relevance of our results to experimentally obtained data is also discussed.