项目利用基底神经节的直接通路和间接通路的神经闭环结构，经过神经突触的STDP学习和神经团块的奖励强化学习，建立具有稀疏网络特性的基底神经节闭环动作选择模式。.（1）直接通路和间接通路的神经闭环，添加神经元之间的突触STDP学习，得到具有稀疏特性的神经回路，模型解释大脑神经系统中，神经突触的稀疏连接问题，稀疏连接的原因-突触的STDP学习..（2）建立基底神经节直接通路和间接通路的神经闭环动作选择模型，模型给出稀疏基底神经节闭环，皮层输入-稀疏闭环-丘脑输出的对应选择模式，提出稀疏闭环的动作选择特性通过神经信号传递的时间确定，优先-突出的闭环通道信息，被用于基底神经节闭环的动作选择模式。.最后，利用稀疏基底神经节闭环动作选择模式，研究帕金森病脑深度刺激的治疗机理和帕金森病脑深度刺激的最优靶点选择.

The project uses nerve closed-loop structure of basal ganglia direct pathway and indirect pathway, through STDP learning of neurosynaptic and reward reinforcement learning of nerve block,establish the basal ganglia closed-loop action selection mode of a sparse network..(1) Neuronal closed-loop of direct pathways and indirect pathways, adding synaptic STDP learning between neurons, the project obtains neural circuits with sparse properties and models explain the sparse connectivity problems of neuronal synapses in the brain's nervous system, the reason for sparse connections - synapses STDP learning..(2) The model of neural closed-loop motion selection in the direct and indirect pathway of basal ganglia was established. The model of closed basal ganglia closed loop is given by the project. The corresponding of action selection mode is cortical input-sparse closed-loop-thalamus output. Proposed the action selection of sparse closed is time-determining, priority-salient closed-loop channel information is used in basal ganglia motor selection mode..In the final the sparse basal ganglia closed-loop motion selection mode is used to study the deep brain stimulation treatment mechanism of Parkinson's disease and the deep brain stimulation optimal target selection for Parkinson's disease treatment.