Automated Prediction of Epileptic Seizures in Rats with Recurrence Quantification Analysis

The prediction of epileptic seizures is a very important issue in the neural engineering. This is because it may improve the life quality of the patients who are suffering from uncontrolled epilepsy. In our earlier work, we found that the dynamical characteristics of EEG data with recurrence quantification analysis (RQA), also called complexity measure, can identify the differences among inter-ictal, pre-ictal and ictal phases. In this paper, we propose an automated technique with complexity measure of EEG recording to detect pre-ictal phase. Using the EEG recorded from rats with experimentally induced generalized epilepsy, it is found the method can detect the complexity changes of the neural activity prior to epileptic seizures. We suggest that the new method could be considered as an alternative of epileptic seizures prediction in practice