Characterization of complexity in electroencephalographic signals

Abstract

Based on the definition of statistical complexity introduced by Lopez-Ruiz, Mancini y Calbet (Phys. Lett. A 209:321, 1995), we present an efficient algorithm to calculate the complexity of a system from experimental data. By using this algorithm, the complexity of electroencephalographic signals is calculated. The data base consists of 10 healthy subjects and 30 epileptic patients. The Principal Component Analysis method has been employed to reduce the dimensionality of the signals. The values of the statistical complexity obtained in this way allow to characterize the two groups of individuals. The results show that the complexity of the collective brain states associated to the epileptic pathology is lower than that corresponding to healthy subjects.