Advanced Models of Cortical Dynamics in Perception

Abstract

A phenomenon of interest is the flash of recognition and accompanying emotion one experiences when one re-ceives a familiar stimulus. Here we explain the speed and richness of the event by postulating phase transi-tions in cortical neuropil: condensation from a gas-like phase to a liquid-like phase followed by evaporation. We model the process with a Carnot-like thermodynam-ic cycle at three successive levels of complexity: prima-ry sensory cortices; limbic system; global neocortex. We replace the thermodynamic state variables of pressure, volume and temperature with neurodynamic variables, respectively mean beta-gamma power, pattern stability (negentropy), and neural feedback gain (mean interac-tion strength). We cite evidence that all sensory cortices use this cycle, necessarily so for two reasons. They all evolved from same primordial forebrain of vertebrates dominated by olfaction, and they all transmit the same form of perceptual information, the wave packet, so that signals in all modalities can be integrated by linear summation.