Journal of Neuroscience Methods 153 (2006) 163–182 EEG analysis using wavelet-based information tools O.A. Rosso a,∗ , M.T. Martin b , A. Figliola c , K. Keller d , A. Plastino b a Chaos and Biology Group, Instituto de C´ alculo, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires. Pabell´ on II, Ciudad Universitaria, 1428 Ciudad de Buenos Aires, Argentina b Instituto de F´ ısica, Facultad de Ciencias Exactas, Universidad Nacional de La Plata and Argentina’s National Research Council (CONICET), C.C. 727, 1900 La Plata, Argentina c Chaos and Biology Group, Instituto de Desarrollo Humano, Universidad Nacional de General Sarmiento. Campus Universitario, Modulo 5, Juan Maria Gutierrez 1150, Los Polvorines, Pcia. de Buenos Aires, Argentina d Institut f ¨ ur Mathematik, Universit¨ at zu L ¨ ubeck, Wallstrasse 40, D-23560 L¨ ubeck, Germany Received 26 May 2005; received in revised form 15 October 2005; accepted 16 October 2005 Abstract Wavelet-based informational tools for quantitative electroencephalogram (EEG) record analysis are reviewed. Relative wavelet energies, wavelet entropies and wavelet statistical complexities are used in the characterization of scalp EEG records corresponding to secondary generalized tonic–clonic epileptic seizures. In particular, we show that the epileptic recruitment rhythm observed during seizure development is well described in terms of the relative wavelet energies. In addition, during the concomitant time-period the entropy diminishes while complexity grows. This is construed as evidence supporting the conjecture that an epileptic focus, for this kind of seizures, triggers a self-organized brain state characterized by both order and maximal complexity. © 2006 Elsevier B.V. All rights reserved. PACS: 87.80.Tq; 05.45.Tp; 05.20.-y Keywords: EEG; Epileptic seizures; Information theory; Wavelet analysis; Signal entropy; Statistical complexity 1. Introduction Synchronous neuronal discharges create rhythmic potential fluctuations, which can be recorded from the scalp through elec- troencephalography. The electroencephalogram (EEG) can be roughly defined as the mean brain electrical activity measured at different sites of the head. The traditional way of analyz- ing brain electrical activity, on the basis of EEG records, relies mainly on visual inspection and years of training. Although it is quite useful, of course, one has to acknowledge its sub- jective nature that hardly allows for a systematic protocol. In order to overcome this undesirable feature, quantitative EEG analysis (qEEG) tools have been developed over the years intro- ducing in this way objective measures. We must remark, how- ever, that these methods have not been developed to substitute ∗ Corresponding author. Tel.: +54 11 4576 3375; fax: +54 11 4576 3375. E-mail addresses: oarosso@fibertel.com.ar (O.A. Rosso), mtmartin@venus.unlp.edu.ar (M.T. Martin), afigliol@ungs.edu.ar (A. Figliola), keller@math.uni-luebeck.de (K. Keller), plastino@venus.unlp.edu.ar (A. Plastino). for traditional EEG visual analysis, but rather to complement them. The EEG reflects not only on characteristics of the brain activity itself but also yields clues concerning the underlying associated neural dynamics. The processing of information by the brain is reflected in dynamical changes in the electrical activ- ity in time, frequency, and space. Therefore, the concomitant studies require methods capable of describing the qualitative and quantitative variation of the signal in both time and fre- quency. An attractive property for possible EEG quantifiers will be that they are related to physical properties; therefore their interpretation and the implications of their results is straight- forward. As a result, to understand the associated dynamics of the EEG time series, one can study the temporal evolution of these associated quantifiers and reach conclusions about their behavior under different pathologies and diseases. According to information theory, entropy is a relevant mea- sure of order and disorder for many systems, including dynam- ical ones (Cover and Thomas, 1991; Shannon, 1948). Indeed, Kolmogorov and Sinai converted Shannon’s information the- ory into a powerful tool for the study of dynamical systems 0165-0270/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jneumeth.2005.10.009