Representation of epileptic discharge dynamics in EEGs on the basis of multiscale correlative signal dynamics analysis
Viacheslav E. Antsiperov, V. A. Morozov, Yu. V. Obukhov · Pattern Recognition and Image Analysis · 2008
This paper presents the latest results in the application of developed multiscale correlative signal dynamics analysis to epileptic discharge investigations. The basic principle of the approach proposed is in detecting the epileptic discharge time intervals and controlling the signal dynamics in such intervals by means of some signal quasi-periodicity features. These features are determined by the structure of a measure of self-similarity calculated for different scales. The graphical temporal representation based on this measure is at the center of the discussion. Interpretations of typical epileptic discharge dynamics patterns in the representation proposed are put into practice; typical patterns are illustrated by experimental representation examples.