Classification of epilepsy EEG signals using DWT-based envelope analysis and neural network ensemble

Abstract

Epilepsy is a neurological disorder of brain which is characterized by recurrent disorders. And people with epilepsy and their families frequently suffer from stigma and discrimination. Hence, seizure identification has great significance in clinical therapy of epileptic patients. Electroencephalogram (EEG) is most commonly used in epilepsy detection, since it contains valuable physiological information of the brain. However, it could be a challenge to reveal the subtle but critical changes contained in EEG signals. In this paper, we propose a novel method for detecting normal, interictal and epileptic signals using wavelet-based envelope analysis (EA) neural network ensemble (NNE). The discrete wavelet transform (DWT) in combination with EA method is developed to extract significant features from the EEG signals. Moreover, an effective network model called NNE is designed specifically to the task of epilepsy detection. For the purpose of evaluating the performance of presented algorithm effectively, different classifiers and feature extracting techniques have been considered in this work. The experimental results have shown that the introduced scheme achieved a satisfying recognition accuracy of 98.78%, which is able to be a valuable method for practical applications treating with epileptics.