Computer Methods and Programs in Biomedicine (2005) 80, 17—23 Characterization of EEG—–A comparative study N. Kannathal a,b,* , U. Rajendra Acharya b , C.M. Lim b , P.K. Sadasivan a a Department of ECE, National University of Singapore, Singapore b ECE Division, Ngee Ann Polytechnic, 535 Clementi Road, Singapore Received 13 May 2004; received in revised form 28 April 2005; accepted 7 June 2005 KEYWORDS Electroencephalogram; Epilepsy; Correlation dimension; Lyapunov exponent; Entropy; Hurst exponent Summary The Electroencephalogram (EEG) is a representative signal containing information about the condition of the brain. The shape of the wave may contain useful information about the state of the brain. However, the human observer cannot directly monitor these subtle details. Besides, since bio-signals are highly subjective, the symptoms may appear at random in the time scale. Therefore, the EEG signal parameters, extracted and analyzed using computers, are highly useful in diagnos- tics. Chaotic measures like correlation dimension (CD), largest Lyapunov exponent (LLE), Hurst exponent (H) and entropy are used to characterize the signal. Results indicate that these nonlinear measures are good discriminators of normal and epilep- tic EEG signals. These measures distinguish epileptic EEG and alcoholic from normal EEG with an accuracy of more than 90%. The dynamical behavior is less random for alcoholic and epileptic compared to normal. This indicates less of information processing in the brain due to the hyper-synchronization of the EEG. Hence, the application of nonlinear time series analysis to EEG signals offers insight into the dynamical nature and variability of the brain signals. As a pre-analysis step, the EEG data is tested for nonlinearity using surrogate data analysis and the results exhibited a significant difference in the correlation dimension measure of the actual data and the surrogate data. © 2005 Elsevier Ireland Ltd. All rights reserved. 1. Introduction Computer technology has an important role in structuring biological systems. The explosive growth of high performance computing techniques in recent years with regard to the development of good and accurate models of biological systems has contributed significantly to new approaches to fun- ∗ Corresponding author. E-mail address: kna2@np.edu.sg (N. Kannathal). damental problems of modeling transient behavior of biological system. The importance of the biological time series analysis, which exhibits typically complex dynam- ics, has long been recognized in the area of nonlin- ear analysis. Several features of these approaches have been proposed to detect the hidden impor- tant dynamical properties of the physiological phe- nomenon. The nonlinear dynamical techniques are based on the concept of chaos and it has been applied to many areas including the areas of medicine and biology. 0169-2607/$ — see front matter © 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.cmpb.2005.06.005