Epilepsy Classification And Modulation Based On Sparse Representation And Prescribed Time Control

Epilepsy is a brain disorder caused by abnormal discharge of brain nerve cells,with repeated,sudden and transient,specific performance of mental,sensory,consciousness,movement and other aspects of the dysfunction,so that patients in the physical and mental are suffering,but also caused a huge burden on the family and society.Epilepsy is also one of the most common neurological disorders in the world.The World Health Organization estimates that there are about 50 million epilepsy patients worldwide.Therefore,the diagnosis,prevention and treatment of epilepsy is still one of the challenging issues faced by human beings around the world,which plays a crucial role in improving the quality of life and improving the mental health of patients.EEG signals are some spontaneous neural oscillation activities,which contain a variety of physiological and pathological information.They are not only used for basic theoretical research in brain science,but also provide an important basis for the diagnosis and treatment of neuropsychiatric diseases such as epilepsy.In this study,machine learning theory,signal analysis and processing technology and methods in modern control theory are used to study the classification and simulation of epileptic EEG signals and the suppression of Seizure.The main research work is as follows.First of all,in view of the problems of slow training speed,weak generalization ability and easy to be interfered by noise in the classification method of epileptic EEG signals based on statistical learning theory,a classification method of epileptic EEG signals based on sparse representation theory and compressed sampling matching pursuit algorithm is proposed.At the same time,the Bonn epilepsy database is used to verify the given method and compare it with the traditional classification method of epileptic EEG signals.Secondly,based on the dynamic characteristics and power spectrum density characteristics contained in epileptic EEG data,the type and parameters of neural group models are determined to simulate the generation of these data,and the correlation coefficients are used as the selection criteria for model optimization parameters.Exploring the generation mechanism of different types of data by analyzing the relationship between the dynamic characteristics of the model and important parameters.Finally,based on the neural group model,a novel predetermined time controller is designed to suppress epileptic spike dynamics and track the expected dynamics within the specified time,and the effectiveness of the designed controller is verified through simulink simulation.The above research involves the intersection of traditional learning theory,signal analysis and processing technology,control theory,brain science and other disciplines,which is expected to provide new ideas,new methods and new theoretical basis for the diagnosis,prevention and suppression of seizure.