Nonlinear Analysis Of The Cerebral Electric Information Of Epilepsy

Epilepsy is a serious disease for human health, and the typical features of epileptic seizure are abruptness, transitoriness and repeat. Epilepsy was caused by excess firing of neurons, which resulted in repeated and transient abnormality of center nerve system. The main treatment of epilepsy is medical therapy and operation. But medication is only efficient to proportion, and has some reverse-effect. Operation does harm to brain and causes many complications.As a chaotic attractor, the cerebral electric activity has chaotic characteristic and strong unstable behavior. For the electric activity in neural network is the basis of thought, when the transition appeared from one steady structure to another steady structure in neural network, the electric signal of brain become chaotic behavior. The abnormal discharge in the epileptic ictal time or the epileptic interictal time was called epileptic discharge. The sudden appearance in the basic electric activity along with high amplitude and lowfrequency is one of the characteristic of epileptic discharge. If the change of electric signal of brain can be measured out of skull, epileptic seizure can be predicted, and epilepsy may be prevented and controlled when patients take medicine in a little dose of slight sedation drugs.Recently, the development of nonlinear dynamics has accelerated research of cerebral electric activity. It is reported that the time series of cerebral electric activity may be low dimension chaotic behavior because of some neurons excess and repeat firing when epilepsy outbreaks. So, it will be novel method to measure epilepsy which the characteristic parameters of chaotic behavior of cerebral electric activity can be measured with methods of nonlinear dynamics.Aim at measurement and prediction of epilepsy, the whole process of epileptic seizure was studied and analyzed in this research. Firstly, the information of biologic neural firing is analyzed using nonlinear methods. Then the animal model of rat's epilepsy is constructed and analyzed. We did the following work:(1) To study of algorithms: to understand and master the basic algorithms of nonlinear dynamics, especially to skillfully make use of correlation dimension, nonlinear prediction, unstable periodic orbits and approximate entropy.(2) To analyze time series of biologic neural firing in order to distinguish and master various algorithms.(3) To construct the animal model of rat's epilepsy. We realize that the law of complexity change of cerebral electric signal through analyzing the data of cerebral electric activity in order to predict seizure of epilepsy.Main results:(1) To analyze the data of neurons firing with different nonlinear methods, it was presented that correlation dimension and nonlinear prediction areonly used in low dimensional system. The method of unstable periodic orbits can identify determination of anomalous rhythm in depth of high dynamic levels and it has more high reliability.(2) In the whole epileptic process, the complexity of cerebral electric activity changes from high to low at first, whereafter it changes from low to high. The preictal time is a special time. The complexity in preictal time differs from complexity in nonictal time, ictal time and postictal time. We find out that this process of change does not abruptly appear. In epileptic preictal time, though the complexity of cerebral electric activity has declined, there is not any symptom of seizure. We realized that the complexity of cerebral electric signal begins to decline slowly in about 100 seconds in preictal time before symptom of seizure can be seen when real time to note cerebral electric signal of whole epileptic process. So by measuring cerebral electric signal of whole epileptic time and analyzing it with nonlinear dynamic methods, we can determine epileptic focus and predict seizure of epilepsy.We measure cerebral electric signal with scalp electrodes out of skull, which is harmless to human body. There is a virtue of little harm to body. In...