Epileptic Seizures Detection on EEG Signal Using Deep Learning Techniques

Abstract

Epilepsy is a chronic brain disease that affects a large percentage of the psychiatric population, and its progression can be fatal. It is a major public health problem as well as a financial burden on affected families, health systems and the overall economy of the country. This highlights the role that neurologists can play in diagnosing epilepsy based on electroencephalography (EEG) signals. In this paper, we proposed a method for seizures detection using two models: Convolutional Neural Network (CNN) and Xception. The CHB-MIT dataset was used as a basic dataset for performance evaluation of considered classification methods. Experimental results shown that our proposed CNN model performs better than the Xception model. Accuracy, precision, recall and F1 score values are respectively equal 98.47%, 99.79%, 98.93%, and 98.51% for the CNN model.