Epileptic Eeg Automatic Identification And Chaos Analysis

The diagnosis of epilepsy relies on an identification of epileptic spikes in EEG However, between seizures it is difficult to identify such spikes in the EEG of an epileptic patient. Doctors have to supervise a long term of EEG to detect those spikes. So people advance a lot of automatic epilepsy detection techniques.In this paper, we introduce some present techniques for automatically detecting multichannel epileptiform activity and conclude a "Wavelet pretreatment--BP Artificial neural network automatically detecting—Chaos anlysis"studying theory. And we describe a method and theory which processes the original data of EEG by analysis, reducing noise, restraining noise in wavelet analysis, transform. After processing the original based on the method, An artificial neural network(BP) is also trained to process the data of EEG, and we get a good result in 0.9 by regressive anlysis betweem network imitative result and aim result, in order to deeply explain why and what arouse epilepsy for pathology's studying, especially for directing clinic, we deeply discuss and study epileptiform movement by the nonlinear chaotic method for the brain wave based on Synergetics,and open out brain activity is dominated by few parameter,and little epileptiform activity is a Chaos.