Design And Analysis Of Closed-loop DBS For Parkinsonian State

Parkinson's disease (PD) is one of the neurological diseases,which seriously affects human health.Deep brain stimulation (DBS) works by applying continuous high frequency pulses to the focal zone such that the PD related symptoms are relieved.Currently used DBS is "open-loop",which delievers preset stimulation to the brain,and cannot change stimulation parameters with clinical state,and cannot relalize the personalized therapy.How to realize the realtime parameters modulation according to the generation mechanism of PD is really challenging.This paper constructs the biophysical model of the PD state,and describes the relationship between physiological signal and PD state.The closed-loop DBS system is constructed based on this model,which can realize the automatic modulation of PD state with optimal energy consumption,and stimulation parameters.This paper constructs the mathematical model of the PD focal zone (the basal-ganglia-thalamus-cortex network) to describe the PD related neurological signals.By analyzing the characteristics of these signals,we demonstrate the abnormal spike pattern of neurons and the increase of beta oscillation activity can be used to describe the PD state.Then,the synaptic block experiment is constructed,results show that when the synaptic input from the Globus Pallidus externa to other nucleus is blocked the oscillation and spiking activities of the whole network are significantly changed.According to above analysis,the"push-pull"control theory is introduced to apply stimulation sequences to the Globus Pallidus externa,and the abnormal beta oscillation of the whole network can be inhibited.The building blocks of the basal-ganglia-thalamus-cortex network (the excitatory-inhibitory motif) are extracted.Then,we prove that the network structure,single neuron characteristics and synaptic connection can directly afrect the controllability and observability of the motif network.When the structure of different motifs is the same,inhibitory motif is more controllable and observable than excitatory motif.Above analyses demonstrate the rationality of closed-loop DBS design.Based on the basal ganglia-thalamus-cortex model,the"on-off" DBS is constructed to lower the energy consumption of DBS.The Subthalamic Nucleus is chose as the stimulation target,and the spike time of Globus Pallidus internal neuron is chosen as the feedback signal.The "on-off" DBS experiment for the PD rat is constructed,and by analyzing the neurological recording data and movement behavior,this"on-off" closed-loop method is able to modulate the pathological activities and lower energy consumption.Adjusting the DBS parameters according to the variation of clinical state is able to realize individualized treatment.This paper proposes the adaptive DBS method,where the basal ganglia-thalamus-cortex model is used as a "virtual patient"experimental platform to generate physiological data.The system identification and nonlinear predictive control algorithms are introduced to modulate the closed-loop DBS amplitude and frequency parameters.Simulation results prove that this closed-loop system is able to adjust the DBS parameters according to the variation of the local field potential signal.The proposed mathematical model construction and clos ed-loop DBS methods can help profoundly understand the construction and analysis process of closed-loop DBS system,which lays a theoretical foundation for the clinical realization of closed-loop DBS.