EEG Analysis Based on the Empirical Mode Decomposition for Detection of Mental Activity

The article describes a new approach to the analysis of frequency components of electroencephalograms for the determination of human psychic activity. To solve the set problems, we used the method of empirical mode decomposition implemented in Matlab and R environment. To assess a human mental state (before/after mental impact, or stress), the estimation of signal empirical modes is used as a decisive rule for setting a diagnosis. The research results of more than 400 signals of electroencephalograms showed a high accuracy of a signal ratio to a mental disorder: an error of the 1st kind is 2%, an error of the second kind is 5%.