The Modeling And Control Of Epileptic Discharge State

Electrical stimulation is an effective therapy for epilepsy,which has a successful clinical application.But its underlying mechanism has not been fully understood.Currently,the majority of applications in clinic are open-loop stimulation,resulting in a poor flexibility and adaptability.A closed-loop electrical stimulation strategy,which is based on the model,is proposed in this paper,and the research will be developed from three aspects,namely the mechanism,modeling and control:Firstly,this paper builds the physiological model of seizures area,analyzes the mechanism and achieves the suppression of epileptic discharge.With the neural mass modeling method,through analyzing the dynamic properties and the influence of key parameters on neural mass' s firing states and waves' propagation to explore the mechanism of epilepsy,obtaining the characteristics of seizures' output.The key parameters include excitatory synaptic gain and coupling strength.Meanwhile,the strategy,which combines the feedback linearization control(FLC)with unscented Kalman filter(UKF),is applied to achieve the suppression of seizures.Secondly,the Volterra model representing the normal state and epileptic state is established and the estimation of model parameters is achieved.By defining the identify performance indicators and traversing all the parameters,this paper selects the most appropriate model order,identification input and model structure parameters,which make the identification performance most optimal.Given the model structure and parameters,the states of normal and epilepsy can be modeled based on the input-output data generated by NMM.At the same time,this paper makes the modeling of dynamic properties of multi-coupled NMM,laying foundation for the subsequent design of controller.Finally,according to the identified Volterra model,the predictive control method is employed to suppress the epilepsy state.As the process model of model predictive control(MPC),the identified Volterra model is used to predict the process output.Then,the control signal is designed with MPC and imposed on NMM to realize the real-time closed-control of epileptic discharge,so as to restrain from its further propagating.Besides,this paper makes comparison with FLC to prove the effectiveness of predictive control method.The model-based closed-loop electrical stimulation strategy proposed in this paper is not dependent on any accurate model,but only several input-output data,the effective control of epilepsy state can be achieved.The simulation results have validated the effectiveness and potential advantages of this means.Moreover,the study of this paper lays foundation for the further research on electrical stimulation therapy of epilepsy.On the other hand,this thesis provides the theory basis for the hardware implement of epilepsy control system and the important theoretical value for the in vivo studies and clinical researches on neural system disease.