Nonlinear dynamical techniques for analysis and modeling of EEG data

Electroencephalogram (EEG), the record of electrical potential of the brain, is now analysed as a chaotic signal rather than a stochastic signal in the light of new developments in nonlinear dynanmics and  chaos. The paper reviews the application of nonlinear dynamical techniques to analyse EEG data. Earlier studies mainly concerned wllh calculating the characterstics of the system like correlatlon dimension and Lyapunov exponents and use them for various applications like study of different sleep stages, epileptic seizures, depths of anaesthesia, etc. One major direction is to develop chaotic, realistic models for EEG generation. Their efficiency is demonstrated by data compression as an application.