Problem of automated EEG signal analysis and extraction of normal and proepilptic activity was studied. Method based on wavelet analysis and empirical mode decomposition was proposed. Analysis of experimental EEG signals was performed, proposed method allows to detect and distinguish basic types of patterns.
$^1$Scientific educational center «Nonlinear dynamics of complex systems», Saratov State Technical University. Saratov 410054, Russia
$^2$Faculty of Nonlinear Processes, Saratov State University. Saratov 410012, Russia
$^3$Laboratory of Ontogenesis, Institute of Nigher Nervous Activity and Neurophysiology Moscow 117485, Russia