Electroencephalogram Signals Processing for the Diagnosis of Petit mal and Grand mal Epilepsies Using an Artificial Neural Network

Abstract

In this study, a novel wavelet transform?neural network method is presented. The presented method is used for the classification of grand mal (clonic stage) and petit mal (absence) epilepsies into healthy, ictal and interictal (EEGs). Preprocessing is included to remove an artifact occurred by blinking and a wandering baseline (electrodes movement) as well as an eyeball movement artifact using the Discrete Wavelet Transformation (DWT). Denoising EEG signals from the AC power supply frequency with a suitable notch filter is another job of preprocessing. The preprocessing enhanced speed and accuracy of the processing stage (wavelet transform and neural network). The EEGs signals are categorized into normal and petit mal and clonic epilepsy by an expert neurologist. The categorization is confirmed by the Fast Fourier Transform (FFT) analysis. The dataset includes waves such as sharp, spike and spike?slow wave. Through the Countinous Wavelet Transform (CWT) of EEG records, transient features are accurately captured and separated and used as classifier input. We introduce a two?stage classifier based on the Learning Vector Quantization (LVQ) neural network localized in both time and frequency contexts. The particular coefficients of the Continuous Wavelet Transform (CWT) are networks. The simulation results are very promising and the accuracy of the proposed method obtained is of about 80%.