Pediatric Epilepsy Assessment Based on EEG Analysis

Epilepsy is one of the most common chronic neurological diseases, observed in all age groups and in all countries regardless of the level of the health system and the standard of living of the population. Epilepsy is caused by outbursts of electrical activity. The inability to detect seizures before they occur lead to the development of algorithms for detecting seizures. In this paper, an automated method of detection is presented. Filters are used to analyze the EEG signal into its frequency bands and then, features are extracted from the frequency and time domain. The extracted features are filtered to reweight the instances in the data and then the feature vector is used as input to train a Random Forests classifier, reaching a detection accuracy of 94.04%.