Large-scale neural network model for functional networks of the human cortex

TL;DR

This paper models large-scale human cortical networks using FitzHugh-Nagumo oscillators to understand how indirect connections, distance, and collective effects influence resting state functional connectivity observed in fMRI data.

Contribution

It introduces a dynamic neural network model incorporating time delays and noise to simulate and analyze resting state functional networks of the human cortex.

Findings

Model reproduces empirical resting state networks

Functional connectivity depends on signal speed and correlation threshold

Simulation results align with observed fMRI data

Abstract

We investigate the influence of indirect connections, interregional distance and collective effects on the large-scale functional networks of the human cortex. We study topologies of empirically derived resting state networks (RSNs), extracted from fMRI data, and model dynamics on the obtained networks. The RSNs are calculated from mean time-series of blood-oxygen-level-dependent (BOLD) activity of distinct cortical regions via Pearson correlation coefficients. We compare functional-connectivity networks of simulated BOLD activity as a function of coupling strength and correlation threshold. Neural network dynamics underpinning the BOLD signal fluctuations are modelled as excitable FitzHugh-Nagumo oscillators subject to uncorrelated white Gaussian noise and time-delayed interactions to account for the finite speed of the signal propagation along the axons. We discuss the functional connectivity of simulated BOLD activity in dependence on the signal speed and correlation threshold and compare it to the empirical data.