Use of data driven optimal filter to obtain significant trend present in frequency domain parameters for scalp EEG captured during meditation

The Scalp EEG is a large-scale & robust information source about neocortical dynamic functions. In this paper, we analyze a scalp Electro Encephalogram (EEG) database of 33 human subjects during the cognitive activity of Meditation, specifically Kriya Yoga. The information measures such as Renyi, Shannon entropies and Relative Energy of the different EEG Bands such as Alpha, Beta, & delta of scalp EEG captured at specific electrodes are calculated for all subjects for the entire duration of Meditation. These frequency domain parameters are obtained as sequences corresponding to the dynamical activity of Meditation and are found to have a hidden dominant trend with many variations present which make the problem of Identification of the dominant trend a difficult problem. Here use of a data driven optimal filter has been employed to find out the dominant trend, and found to yield a clear monotonic change in the frequency parameters. This monotonic sequence can easily assumed to be corresponding to the dynamic activity during Meditation.