Characterization of state transitions in spatially distributed, chaotic, nonlinear, dynamical systems in cerebral cortex.

Abstract

The neurons of cerebral cortex are largely autonomous and generate activity that is manifested in trains of microscopic axonal action potentials. The neurons interact by sparse but numerous synaptic connections to generate macroscopic dendritic activity patterns that are observed in electroencephalographic (EEG) waves. The macroscopic patterns are constructed by the populations and they shape the output of cortical neurons in parallel arrays. Sensory cortexes receive sensory information in the form of microscopic action potentials, which induce state transitions in population dynamics. Each state transition transforms sensory information to perceptual meaning. The EEG reflects both kinds of activity. The sensory input is accessed by time ensemble averaging, whereas the perceptual output is found by spatial ensemble averaging. Spatial phase gradients in the EEG are useful for identifying EEG segments in a sequence of state transitions in response to sensory input. The rapidity and flexibility with which they take place give strong reason to postulate that the mechanism for the construction of these sequences of patterns is a dynamical system operating in a chaotic domain.