Performance Analysis of Lifting based DWT and MLPNN for Epilepsy seizure From EEG

EEG recording are used to analyze the electrical signals generated by the brain. It is used in diagnosing and monitoring process of neurological disorder such as Epilepsy. Epilepsy cannot be controlled by available medical treatments. Its major manifestation is epilepsy seizure. Lifting Based Discrete Wavelet Transform (LBDWT) an efficient toll for representing electroencephalogram signals. EEG changes will be classified by Multilayer perceptron Neural Network (MLPNN). The classification rules were extracted from EEG that were reordered from healthy volunteers, epilepsy patients during seizure free interval and epilepsy patients during epileptic seizure. EEG signals were used as input of the MLPNNs trained with Back propagation and Levenberg-Marquadrant algorithm. Decision making was done in two stages: feature extraction by using LBDWT and classification using MLPNNs trained with the BP and LM algorithms. In this paper, we present an algorithm for classification of EEG (normal and Epilepsy) signals based on lifting based Discrete Wavelet Transformation and patterns recognize techniques.