Deterministic chaos and noise in three in vitro hippocampal models of epilepsy

Recent reports have suggested that chaos control techniques may be useful f or electrically manipulating epileptiform bursting behavior in neuronal ens embles. Because the dynamics of spontaneous in vitro bursting had not been well determined previously, analysis of this behavior in the rat hippocampu s was performed. Epileptiform bursting was induced in transverse rat hippoc ampal slices using three experimental methods. Slices were bathed in artifi cial cerebrospinal fluid containing: (1) elevated potassium ([K+](o) = 10.5 mM), (2) zero magnesium, or (3) the GABAA-receptor antagonists bicuculline (20 muM) and picrotoxin (250 muM). The existence of chaos and determinism was assessed using two different analytical techniques: unstable periodic o rbit (UPO) analysis and a new technique for estimating Lyapunov exponents. Significance of these results was assessed by comparing the calculations fo r each experiment with corresponding randomized surrogate data. UPOs of mul tiple periods were highly prevalent in experiments from all three epilepsy models: 73% of all experiments contained at least one statistically signifi cant period-1 or period-2 orbit. However, the expansion rate analysis did n ot provide any evidence of determinism in the data. This suggests that the system may be globally stochastic but contains local pockets of determinism . Thus, manipulation of bursting behavior using chaos control algorithms ma y yet hold promise for reverting or preventing epileptic seizures. (C) 2001 Biomedical Engineering Society.