Epilepsy seizure detection using EEG signals
Loading...
Authors
Lasefr, Zakareya
Ayyalasomayajula Venkata Naga, Raghavendra Sai Shiv
Elleithy, Khaled M.
Issue Date
2016-04-01
Type
Presentation
Language
en_US
Keywords
Detection , Electroencephalography (EEG) , Epilepsy , Signal processing
Alternative Title
Abstract
EEG signal processing involves multiple algorithms in which epileptic data is received in the MATLAB environment and needs to be processed in order to obtain a perfectly filtered waveform and process it in both the time and frequency domain. In our work we have shown the EEG signal in the frequency domain using Fast Fourier Transform and its absolute value. Using Wavelet decomposition technique we divide the EEG signal into different sub-level bands then the lowest frequency sub-band was selected to perform feature extraction. Discrete Wavelet Transform (DWT) was applied and Vector Analysis was used for feature extraction and then we have used Inverse Discrete Fourier Transform to transform from frequency to time domain so that frequency analysis of the feature extracted EEG signal could fetch the best results. We have used the lowest frequency band possible between 1 and 3.45 Hz which could be the smallest possible in order to either classify a signal or to apply threshold and compare the results. In order to verify our work, we are comparing our results with some of the mostly used classifiers results even though classifiers do not show frequency analysis.