Abstracts

There are numerous analytical models of the mechanisms by which seizures are precipitated. Relatively little effort, by comparison, has gone into the quantification of the termination of seizures. Similarly, there is relatively little information on the molecular mechanisms involved in the termination of seizures. The purpose of this project was to to determine whether, at first approximation, the quantitative dynamics of entry and termination mirrored one another. We selected a range of linear and nonlinear dynamic systems measures to quantifiy the entry and termination processes, including Kolmogorov entropy, golobal nonlinearity, correlation dimension (two types) and eigenvalue. All of the selected seizures were obtained from the pediatric aged population and were predominantly but not solely neocortical events., Three seizures from each of ten pediatric aged subjects were consecutively selected from patients undergoing intracranial monitoring. [quot]Control[quot] segments, temporally distant from the seizure onsets, were also processed. Seizure onset was selected by visual analysis and the data sixty seconds before and sixty seconds after seizure cessation was used. A moving window approach (one second steps) was used with a window duration of thirty seconds. The seizure durations were variable but we always compared the thirty seonds preeceding seizure and thirty seconds following seizures, coupled with at least 15 seconds of seizure activity temporally adjacent to the end and beginning of the event. This was performed for multiple electrodes. The temporal waveforms generated within subjects for the multiple moving windows were correlated to determine replicability., In general, the within subject replicability was extremely high. Correlations between seizure moving window waveforms one and two and two and three were virtually always greater than .9 and oftern above .95. Entry into seizures is reflected in the decreased dimensionality but increased nonlinearity of the system. This is reflected in the decreased entropy associated with entry into seizures and the increase in entropy associated with termination. In most of the subjects there was a clear [quot]overshoot[quot] in entropy at termination but not so for the measures of global nonlinearity nor eigenvalue. The shape of the time course of entry and termination (termporal dynamics) was virtually never the same for eigenvalue, entropy nor global nonlinearity., The termination of seizures, as quantified by multiple nonlinear systems measures, does not mirror the entry into seizures. This suggests that the molecular mechanisms allowing termination are not the same as those operative in seizure initiation, or this reflects the impact of applying the same mechanisms to a different starting point. The within subject replicability of the moving window charactrerization is highly replicable., (Supported by Falk Medical Trust Foundation.)