Signal Dynamics Analysis for Epileptic Seizure Classification on EEG Signals

Epilepsy is the most common form of neurological disease. Patients with epilepsy may experience seizures a certain duration or without provocation. analysis can be done an electroencephalogram (EEG) examination. Observation qualitative EEG signals generates high cost and often confuses due to nature non-linear signal noise. In this study, we proposed processing system for seizure detection. The dynamics approach normal signals' characterization became main focus study. Spectral Entropy (SpecEn) fractal are used estimate as feature sets. method validated using public dataset, which included preictal, ictal, interictal stages Naïve Bayes classifier. test results showed that able generate ictal detection accuracy up 100%. It hoped considered in on long-term recording. Thus it simplify diagnosis epilepsy.