Computer simulation of neuronal circuit models of rhythmic behavior in the electroencephalogram

A computer program for modeling some features of the electroencephalogram (EEG) recorded by scalp electrodes in terms of the time- and space-dependent interactions of populations of neurons in different hypothetical brain configurations was developed. The input of the model consists of: (1) geometric and anatomic data characterizing the brain configuration; (2) physiologic features characterizing neurons; and (3) functions describing the time-dependent afferent impulses to the brain configuration under study. The output of the model consists of plots of selected intracellular and extracellular potentials as a function of time. In application of the model to various brain configurations, some configurations were sufficiently compact spatially that propagation times of action potentials between elements were either taken to be constant or were nearly independent of distance within the accuracy of the calculations. Other configurations represented cerebral cortex alone or a combined thalamocortical system in which many elements interacted via global interconnections. The basic requirement for rhythmic behavior is the existence of circuits containing inhibitory neurons. The characteristic frequencies of rhythmic oscillations are primarily determined by the local circuit parameters and are relatively independent of global circuit parameters.