Neurological Disorders

Epileptic seizure prediction using one channel EEG signal and 2 D-convolutional neural networks

Abstract

Golamali Alizadeh*, Tohidy Yousefi Rezaii and Saeed Meshgini

Purpose: Epilepsy is a neurological disorder affecting more than 55 million people worldwide. Predicting epileptic seizures will improve the lives of people with epileptic seizures. Currently, the diagnosis of an epileptic attack and the analysis of recorded brain activity is performed by a neurologist, which is often accompanied by human error. Recently, researchers have been looking to design and build an automatic system to detect and estimate the occurrence of epilepsy.

Methods: In the present study, two new-fangled methods were proposed based on brain signals and a Convolutional Neural Network (CNN). In this research, Pulse Width Modulation (PWM) is used to create features.

Results and conclusion: Numerous experiments were performed and the accuracy of estimating epilepsy of the proposed methods was achieved in 100%. The proposed methods were more accurate than the previous methods and can be employed as a physician's assistant once entering the field of operation.

HTML PDF