ARCHIVES

Abstract: Electroencephalographic record is one of the most important tools for the study of the brain electrical activity and for the diagnosis of neurological diseases.The electroencephalogram (EEG) is widely used clinically to investigate brain disorders. However, abnormalities in the EEG cannot be understood by simply looking at EEG signal waveform. Changes and abnormalities in EEG signal is too subtle to be detected using conventional techniques of frequency domain analysis. This paper compares the various methods for filtering of EEG signal into different frequency bands. It was concluded that if feature extraction of the EEG signal is done by using wavelet decomposition methods using Daubechies2 wavelet, the accuracy, sensitivity and specificity is the best.