Performance Evaluation of Epileptic Seizure Prediction Using Time, Frequency, and Time–Frequency Domain Measures

The prediction of epileptic seizures is crucial to aid patients in gaining early warning and taking effective intervention. Several features have been explored predict the onset via electroencephalography signals, which are typically non-stationary, dynamic, varying from person-to-person. In former literature, applied classification shared similar contributions all patients. Therefore, this paper, we analyze impact specific combination feature channel time, frequency, time–frequency domains on performance disparate Based minimal-redundancy-maximal-relevance criterion, proposed framework uses a sequential forward selection approach individually find optimal channels. Trained models could discriminate pre-ictal inter-ictal with sensitivity 90.2% false rate 0.096/h. We also present comparison between accuracy obtained by features, several summarized complete set three domains. results indicate that various patient interpretations certain specificity feature-channel. Furthermore, detailed list proffered for reference those who research corresponding database.