Use of Cellular Automata-Like Models to Understand Cortical Very Fast Oscillations

by Prof. Roger D. Traub (IBM T.J. Watson Research Center and Columbia University, New York)

Tuesday 30 Oct 2012, 10:00 → 11:00 Europe/Stockholm

RB35

Cellular automata (CA) were introduced in the 1940s in order to model the statistical behavior of gases; they have since been studied as objects of mathematical interest in themselves, and for modeling a variety of processes in physics, computing and biology. Networks of electrically coupled axons lend themselves to CA models because - to first approximation, when electrical coupling is strong - the axon exists in an "on" or "off" state, with brief refractoriness, and with changes of state induced by recent states of nearby other axons. CA models are suitable for large, fast simulations, and for illuminating physical principles. For example, they have been used to show that very fast oscillations (VFO, >100 Hz) can be generated in two distinctive ways in an axonal plexus: the first is by a series of propagating, non-intersecting, randomly initiated waves; the second, by re-entry around particular "minimal" cycles, whose length is just longer than that needed to overcome axonal refractoriness. Prof. Traub propose that the first mechanism may account for ripples, at up to ~250 Hz; and the second may account for so-called fast ripples, >250 Hz, and likely is of epilepsy-related significance.