Collective dynamics in spiking neural networks: A systematic review

TL;DR

This systematic review analyzes recent research on collective behaviors in excitatory-inhibitory spiking neural networks, identifying key dynamical states and factors influencing network dynamics, with a focus on the role of Quadratic Integrate-and-Fire neurons.

Contribution

It provides a comprehensive synthesis of recent findings on collective dynamics in E-I spiking neural networks using PRISMA methodology, highlighting the significance of QIF neurons.

Findings

Identified four dynamical states: synchrony, irregular, stationary, oscillatory.

Collective dynamics depend on intrinsic properties, balance, and stimuli.

QIF neurons are prominently used for large-scale network studies.

Abstract

This study aims to review recent research on the collective behaviour of excitatory and inhibitory (E-I) spiking neural networks. The research methodology used is Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) procedures, comprising three primary stages: an initial search for literature in the SCOPUS database, a screening process based on specific inclusion and exclusion criteria, and a review of the selected literatures. Out of 491 documents from 2014 to 2024, 6 research papers are qualified for review. Four distinct dynamical states have been identified: synchrony, irregular behaviour, stationary state, and oscillatory dynamics. Our review findings suggest that the collective dynamics of E-I spiking neurons stem from the interaction of intrinsic neuronal characteristics, balance mechanisms, and the type of external stimuli. Additionally, the widespread use of Quadratic Integrate-and-Fire (QIF) neurons in the literature highlights its significance as a robust candidate for exploring collective behaviours in large-scale neuronal networks.