Distributed Delay and Desynchronization in a Brain Network Model

Isam Al-Darabsah, Sue Ann Campbell, Bootan Rahman

Published 2023-11-26Version 1

We consider a neural field model which consists of a network of an arbitrary number of Wilson-Cowan nodes with homeostatic adjustment of the inhibitory coupling strength and time delayed, excitatory coupling. We extend previous work on this model to include distributed time delays with commonly used kernel distributions: delta function, uniform distribution and gamma distribution. Focussing on networks which satisfy a constant row sum condition, we show how each eigenvalue of the connectivity matrix may be related to a Hopf bifurcation and that the eigenvalue determines whether the bifurcation leads to synchronized or desynchronized oscillatory behaviour. We consider two example networks, one with all real eigenvalues (bi-directional ring) and one with some complex eigenvalues (uni-directional ring). In bi-directional rings, the Hopf curves are organized so that only the synchronized Hopf leads to asymptotically stable behaviour. Thus the behaviour in the network is always synchronous. In the uni-directional ring networks, however, intersection points of asynchronous and synchronous Hopf curves may occur resulting in double Hopf bifurcation points. Thus asymptotically stable synchronous and asynchronous limit cycles can occur as well as torus-like solutions which combine synchronous and asynchronous behaviour. Increasing the size of the network or the mean time delay makes these intersection points, and the associated asynchronous behaviour, more likely to occur. Numerical approaches are used to confirm the findings, with Hopf bifurcation curves plotted using Wolfram Mathematica. These insights offer a deeper understanding of the mechanisms underlying desynchronization in large networks of oscillators.