Understanding the nonlinear dynamics of seizure and sleep EEG patterns generated using hierarchical chaotic neuronal network

Abstract

The purpose of this article is to describe how a chaotic biological neural network based on a mammalian olfactory system can be used to generate EEG patterns during seizures, REM and NREM sleep. The parameters governing the connection between each node at each layer of an olfactory system's K3 topology have been tuned to replicate low and high dimensional activities as well as periodic bursts matching to distinct brain states. The chaotic qualities of the simulated time series are evaluated against practical recordings of EEG patterns generated during distinct brain states by computing Hurst exponent, fractal dimension, and detrended fluctuation analysis. Our findings contribute to a better understanding of the complex cognitive tasks involved in various functional stages of the brain, as well as to the modelling of these activities using a biologically plausible hierarchical network of neurons.