Epileptic seizure detection by exploiting EEG signals using different decomposition techniques and machine learning approaches

Abstract

In recent years, detecting epileptic seizure has gained a high demand in the field of research. It is such a common and high talked brain disorder, since more than 65 million individuals worldwide are affected by this very disease. Electroencephalogram (EEG) signals is widely used for identifying brain diseases like epileptic seizure. In this thesis, two features are extracted based on short-time fourier transform(STFT) and pseudo-wigner distribution (PWD) and these features are then used to classify seizure and non-seizure EEG signals using support vector machine (SVM). Experimental results show that our proposed approach achieved high classification accuracy (i.e.,92.4%) considering five groups of people. Key-words: EEG, Epilepsy, Seizure, SVM, STFT, PWD.