Chaos analysis in a computer simulation of the piriform cortex

In this paper, we study the level of chaos (using the largest Lyapunov exponent measure) and the geometry of the attractors (using fractal dimension) for a biologically realistic neural network model of the piriform cortex. The analyzed signals which describe the model are the EEGs, recorded during simulations obtained with the GENESIS system (general neural simulation system). We show experimentally that modifying the structure of the model by increasing the density of the neurons leads to a decrease in the level of chaos. We also present the evolution of the geometry of the attractor (the correlation dimension) function of the level of excitation or inhibition.