Research On Closed-loop DBS Control Of The Basal Ganglia In Parkinson's Disease

Parkinson's Disease(PD)is a degenerative disease of the nervous system that most often affects the elderly.Deep Brain Stimulation(DBS)is a neurosurgical treatment of PD.The open-loop DBS method currently used in clinics cannot adaptively adjust the stimulus according to the symptoms of PD and can easily cause damage to the stimulus target and excessive energy consumption of the system.Compared with open-loop DBS,closed-loop DBS has a variety of stimulation methods such as on-demand stimulation and adaptive stimulation,which can alleviate the symptoms of PD while reducing side effects and energy consumption.Therefore,the closed-loop DBS control research has become a hot spot in PD research,in which the study of models and control algorithms are the most important.However,there is currently no intelligent low-power control algorithm for PD patients closed-loop DBS.In response to these problems,the details have the following four aspects:Firstly,the open-loop DBS and closed-loop DBS are analyzed and compared in terms of stimulation methods,side effects,energy consumption,etc.The computational model is introduced as a virtual patient to verify the effectiveness of the closed-loop DBS control method,and the key technologies in closed-loop DBS control based on the computational model are discussed.The physiological structure of the Basal Ganglia(BG)network is elaborated in detail to explain that the incidence of PD is related to the lesion of the BG network.Secondly,building a BG network computational model based on the physiological structure of the BG network for making rhythm analysis,evaluation,and verification under normal state,PD state,and DBS state.And some biomarker signals in the closed-loop DBS are explained.In the model-based DBS control,some acquisition methods of feedback signals are analyzed,and the evaluation indicators of the closed-loop DBS control system are explained.Next,making waveform optimization based on the BG network computational model for open-loop DBS and closed-loop DBS control.A closed-loop DBS control method based on neural network predictive control is proposed,and it is compared with closed-loop proportional integral differential control based on thalamus relay accuracy parameters and synaptic conductance in terms of model error and stimulation energy consumption.Finally,the results show that,compared with the open-loop waveform optimization using genetic algorithm and particle swarm algorithm for optimal parameter search,closed-loop DBS has a better improvement effect on PD state and lower energy consumption.In the closed-loop DBS control method,the proposed closed-loop DBS control frequency based on neural network predictive control has the lowest model error and stimulation energy consumption,which is more suitable for subsequent use in closed-loop DBS.