Analysis of EEG signals during micro-sleeps

In the paper the design and testing of efficiency in recognizing vigilance drifts and EEG-patterns, which precede micro-sleep onset, is described. These algorithms, which are based on Gabor filtering and other approaches, are discussed. In the analysis of EEG signals pseudo-spectra usually four frequency bands are considered. Brain waves change during different states of somnolence. The typical patterns were analyzed and the results of this analysis are presented.