Modeling And Closed-loop Control Of Parkinsonian State

With its application and development in the treatment of Parkinson's Disease(PD),the action mechanism and performance optimization of Deep Brain Stimulation(DBS)has become a hot topic of current research.Generally,current clinical DBS uses open-loop technology,with its stimulator continuously transferring high-frequency and constant-amplitude electrical pulses to the subcortical region of the brain.Without any feedback information,the stimulus signal cannot provide a personalized,adaptable and energy-saving treatment.Thus,this work proposes closed-loop DBS optimization scheme,which focuses on establishing the computational model,exploring the stimulus mechanism and optimizing the energy expenditure of the stimulation.Cortico-basal ganglia-thalamus neural circuits are the main nuclei which are associated with the motor function.Their neurodegeneration changes the firing modes of the neural nuclei,hinders the normal physiological function of the basal ganglia(BG),and thus results in Parkinsonian state.According to the anatomy structure and the dynamic behavior characteristics of the BG,this work establishes the computational models from the levels of the individual neurons and neural network,characterizes the discharge behaviors of the BG under different conditions,describes the impacts of the lack of dopamine on their discharge behavior,and reveals the biophysical mechanism of the Parkinsonian state.By applying various types of DBS to the established computational model,this work studies the various firing modes of the cortico-basal ganglia-thalamus neural network,establishes the relationship between DBS and firing mode of the BG,obtains the optimized performance index of DBS and explores the action mechanism of DBS.It is shown that DBS can modulate the discharge behaviors of the BG with the various ways of local inhibition,distal effect,synchronization and resonance.In order to explore the characteristics of the DBS,the thalamus is selected as the research subject.The relay reliability of the thalamus to response to the sensorimotor information is one of the quantitative standards to describe the Parkinsonian state.Based on this,this work proposes a closed-loop control strategy to explore the mechanism of the closed-loop DBS.Based on the computational model of the thalamus under the DBS condition,this work proposes a closed-loop control strategy based on thalamic neuron's slow variables feedback,which achieves fast track to a desired waveform with zero error and demonstrates that it can robustly suppress the real-time changes of the internal parameters of the nervous system.The low-amplitude and smoothly varying-waveform of DBS can demonstrate the effectiveness of the closed-loop control algorithm.In order to deal with the disadvantages of the open-loop DBS such as high energy expenditure and non-specificity,this work proposes the closed-loop DBS with local field potential feedback,based on the established relationship between DBS and the response of the BG.Generalized predictive and neuron adaptive control algorithms are used to achieve a closed-loop modulation of the Parkinsonian state.Simulation results show that the proposed scheme can automatically adjust the waveform of DBS in real time,which reduces the energy expenditure of the stimulation compared with traditional open-loop DBS,and thus extends the battery life of the DBS,so as to reduce side effects and risks of the frequent battery replacement surgery.The proposed models and algorithms in this work can be applied to clinical research to provide a new idea for the treatment of neurological disorders using DBS.