A comparison analysis of embedding dimensions between normal and epileptic EEG time series.

The embedding dimensions of normal and epileptic electroencephalogram (EEG) time series are analyzed by two different methods, Cao's method and differential entropy method. The results of the two methods indicate consistently that the embedding dimensions of EEG signals during seizure will change and become different from that of normal EEG signals, and the embedding dimensions will vary intensively during seizure, whereas the embedding dimensions of normal EEG signals basically maintains stability. The embedding dimension results also reflect the variation of freedom degree of the human brain nonlinear dynamic system (NDS) during seizure. And based on the results of Cao's method, it is also found that normal EEG signals are of some degree of randomness, whereas epileptic EEG signals have determinism.