Inhibitory connections enhance pattern recurrence in networks of neocortical pyramidal cells

Abstract We consider biological neural networks of pyramidal cells in a quasistatic approximation. We argue that they can be treated as a coupled map lattice of inhibitory and excitatory site maps, where both maps are derived from perturbation response of rat neocortical pyramidal cells. Inhibitory site maps generate chaotic spike patterns on an open parameter set of positive measure [R. Stoop, K. Schindler, L.A. Bunimovich, submitted], excitatory site maps are nonchaotic. Our network simulations show that local chaos by inhibition may be used to synchronize cortical networks.