EEG Oscillatory Power and Complexity for Epileptic Seizure Detection

Monitoring patients at risk of epileptic seizure is critical for optimal treatment and ensuing the reduction complications. In general, detection done manually in hospitals involves time-consuming visual inspection interpretation by experts electroencephalography (EEG) recordings. The purpose this study to investigate pertinence band-limited spectral power signal complexity order discriminate between seizure-free EEG brain activity. are evaluated five frequency intervals, namely, delta, theta, alpha, beta, gamma bands, be used as feature representation. Classification data was performed prevalent potent classifiers. Substantial comparative performance evaluation experiments were on a large record 341 Temple University Hospital database. Based statistically validated criteria, results show efficiency when using random forest gradient-boosting decision tree classifiers (95% area under curve (AUC) 91% both F-measure accuracy). These support use these automatic classification schemes assist practicing neurologist interpret records more accurately without tedious inspection.