神经核团异常放电的数值模拟与FPGA实现-《中国医学物理学杂志》

文章信息/Info

Title:: Numerical simulation of abnormal discharge of nuclei and FPGA implementation

Author(s):: WANG Chao; LU Mai; Key Laboratory of Optoelectronic Technology and Intelligent Control, Ministry of Education, Lanzhou Jiaotong University, Lanzhou 730070, China

Keywords:: Keywords: basal ganglia network of nuclei FPGA Izhkevich model numerical simulation

摘要:: 进行神经核团网络异常放电的数值模拟以及硬件实现。方法：选用Izhikevich模型模拟单个神经元放电，通过化学突触连接各神经元搭建神经核团网络，进而在FPGA上实现神经核团放电。对比Izhikevich模型在Simulink和DSP Builder两种软件中建模的相对标准误差，并分析各神经核团的放电特性。结果：两种软件中各神经核团的仿真结果相对标准误差均小于0.1，验证了建模的一致性。通过计算丘脑的中继可靠性指标RI=0.3<0.6，证明搭建的神经核团网络可以模拟帕金森病的一种放电状态。结论：模拟神经核团的异常放电对替代动物活体实验、探究神经性疾病的治疗方法和脑机接口研究等具有重要意义。

Abstract:: Abstract: Objective To realize the numerical simulation of abnormal discharge of nuclei and its hardware implementation. Methods Izhikevich model was used to simulate single neuron discharge, and neurons were connected via chemical synapses to build the network of nuclei, and finally, the discharge of nuclei was realized on FPGA. The root mean squared error was used to compare the results of Izhikevich modeling by Simulink versus DSP Builder, and the discharge characteristics of different nuclei were analyzed. Results The root mean squared error of simulation results between Simulink and DSP Builder was all less than 0.1, which verifies the consistency of the modeling. The calculated reliability index of the thalamic cortex was 0.3, less than 0.6, which proved that a discharge state of Parkinsons disease could by simulated by the established network of nuclei. Conclusion Simulating the abnormal discharge of nuclei is of significance for replacing the experiments in vivo, exploring the treatment of neurological diseases and studying brain-machine interface.