A new seizure prediction method based on return map

Epilepsy is one of the most common neurological disorders and affects closely 50 million people worldwide. The unpredictability of seizures onset is a main agony for all patients who has epilepsy. In this article, a new seizure prediction method based on chaos theory and nonlinear dynamics are used to make a return map based on zero crossing to find the Preictal stage before Ictal stage in electroencephalographic (EEG) signal in order to seizure onset prediction. Our index is calculated for sliding windows and compared with adaptive threshold. Lastly our average prediction time is about 20±5 minutes with 86.6 % in sensitivity and 0.067 in FPR.