Automated detection of a preseizure state based on a decrease in synchronization in intracranial electroencephalogram recordings from epilepsy patients

The question whether information extracted from the electroencephalogram ~EEG! of epilepsy patients canbe used for the prediction of seizures has recently attracted much attention. Several studies have reportedevidence for the existence of a preseizure state that can be detected using different measures derived from thetheory of dynamical systems. Most of these studies, however, have neglected to sufficiently investigate thespecificity of the observed effects or suffer from other methodological shortcomings. In this paper we presentan automated technique for the detection of a preseizure state from EEG recordings using two differentmeasures for synchronization between recording sites, namely, the mean phase coherence as a measure forphase synchronization and the maximum linear cross correlation as a measure for lag synchronization. Basedon the observation of characteristic drops in synchronization prior to seizure onset, we used this phenomenonfor the characterization of a preseizure state and its distinction from the remaining seizure-free interval. Afteroptimizing our technique on a group of 10 patients with temporal lobe epilepsy we obtained a successfuldetection of a preseizure state prior to 12 out of 14 analyzed seizures for both measures at a very highspecificity as tested on recordings from the seizure-free interval. After checking for in-sample overtraining viacross validation, we applied a surrogate test to validate the observed predictability. Based on our results, wediscuss the differences of the two synchronization measures in terms of the dynamics underlying seizuregeneration in focal epilepsies.