European Journal of Molecular & Clinical Medicine

Abstract

Epilepsy causes sudden and unforeseen seizures in patients. Electroencephalogram (EEG) is a non invasive procedure used in brain control and epilepsy diagonsis. The non-stationary nature of EEG signals poses difficulty in extracting features of diagnostic importance by simple analysis method. For epileptic seizure detection from EEG signals,a simple classification methods is suggested in this paper. Publicly available dataset is chosen for testing the method..The Fourier transformation is used to convert EEG input signals from the time domain to the frequency domain for deeper analysis. Butter-worth band-pass filter is applied to get improved signal-to-noise ratio and decomposition is performs Discrete Wavelet Transform (DWT) for better spatial resolution. Based on the interval and extracted spike based features the normal and epileptic signals are classified using one of the machine learning algorithms known as Support Vector Machine (SVM).