Click here to load whole tree
NITAAI-Veda.nyf > Soul Science God Philosophy > Science and Spiritual Quest > Section 4 Towards a New Biology > MATHEMATICAL TECHNIQUES FOR STUDY OF EEG > 3. Results

3. Results


During meditation there is an increase in the average fractal dimension value.


Four channels of unipolar EEG are acquired (F3, F4, 01 and 02 referenced to A2) from the subjects. The subjects are instructed to be restful for five minutes following which they meditated for some time. After meditation, the recording is continued for about 5 minutes. The EEG signals are digitized at a sampling rate of 128 Hz and off-line filtered through a band pass (0.5-32 Hz) fourth order Butterworth filter twice cascaded. The compressed spectral array (CSA) was obtained by taking 4 seconds of data blocks. The running fractal dimension was estimated using consecutive data blocks of 8 seconds duration. The running attractor dimension is obtained by taking 4096 data points and the data window was shifted across the entire record by 1024 points (8 sec). The spectral plots (CSA) yield some interesting features. During pre-meditation the activity range is that of a relaxed individual. During meditative phase, increased cerebral activity ensues and marked transitions are observed ( The running fractal plots (Fig. lb) show low average fractal dimension values during premeditative and post meditative periods.During meditation there is an increase in the average fractal dimension value. The attractor dimension values also show changes ( As the meditation progresses the average attractor dimension rises to a value which is more than that for the premeditative period. It shows a decline during some stages of meditation.Four-classification algorithms viz., KNNC, back propagation networks, RBF networks and support vector machines were tested for the classification of different states of meditation. Tables 1-3 shows the classification accuracy on the test sets of the algorithms with the expert set of inputs for 2, 3 and 4 category classification. The classification accuracies are presented for different sets of features to the algorithms.