A Unique Approach of Noise Elimination from Electroencephalography Signals between Normal and Meditation State

Abstract

In this paper, unique approach is presented for the electroencephalography (EEG) signals analysis. This is based on Eigen values distribution of a matrix which is called as scaled Hankel matrix. This gives us a way to find out the number of Eigen values essential for noise reduction and extraction of signal in singular spectrum analysis. This paper gives us an approach to classify the EEG signals between normal condition (Controlled) and meditation condition, the extraction of various patterns, the EEG signal filtering and the noise removal from the signals. Different parameters are used as features for classification during subject's normal EEG segments and at the time of practicing Meditation. The results showed positive approach for noise removal in both EEG signals.