Symbolic Dynamics Analysis And Its Application In EEG Signals Processing

Recently, with the improvement of modernization and civilization, the incidences of the human diseases are increasing significantly, so that World Health Organization (WHO) has ranked mental-health protection as one of the most important works. However, the symptoms and courses of these diseases are so various and the etiology is so complex that it is very difficult for doctors' diagnosis. To establish objective criterion for evaluation and fine quantitative analysis is a hot topic of research. Electroencephalography (EEG) is an external representation of human brain's behaviors, and also an important assistant tool for clinical diagnosis. But because of the limitation of the classical EEG analysis, most EEG analysis lacks catholicity. Attempts to use it as a clinical diagnostic tool has not yet succeed and then accepted by psychiatrist. Since brain wave has been proved to be typical nonlinear, non-stationary and chaotic, we think it is more appropriate to analyze EEGs' dynamical characteristics with nonlinear and non-stationary methods.In this paper, we first introduced the theory of symbolic dynamics. Then we proposed a new symbolic entropy based on sliding temporal window technique to analyze the complexities of the EEGs with evolution of time. By computing the temporal evolution of symbolic entropy, the abrupt changes in signals can be detected by significant larger values of the entropy. EEGs, including both normal group and epileptic patients were studied based on symbolic dynamics, including Lyapunov exponent analysis, the correlation dimension, nonlinear detection and entropy analysis. As we all know, human brain is a time-varying coupled neural network. Each function comes from the information transmission and exchange in cortex, so from the viewpoint of information, we proposed a new cross symbolic entropy to estimate the similarity of different brain areas under different cognitive processes. We also did some research on visual attention, and made the corresponding experiment.The research results of this paper prove that the complexity based on symbolic dynamics may be a new direction for EEGs' analysis. It is very simple and reliable. The symbolic dynamics especially that based on the sliding temporal window technique can describe the information of the system with the evolution of time in each parts of the brain more accurately and capture the abrupt changes in the signals. So it can characterize the features of the EEGs, which may help us to understand the advance neural functional electrophysiology base.