基于Spiking神经网络的脑运动分类研究  

Studies on Cortical Neural Signals Classification Based on Spiking Neural Networks

作　　者：王博[1] 胡成玉[2] 方慧娟[2] 王永骥[2] 

机构地区：[1]上海交通大学自动化系,上海200084 [2]华中科技大学图像信息处理与智能控制教育部重点实验室,湖北武汉430074

出　　处：《计算技术与自动化》2008年第4期102-104,110,共4页Computing Technology and Automation

基　　金：国家自然科学基金(60674105;60340420431);教育部博士点基金(20050487013)

摘　　要：针对大脑运动皮层群体神经元信号与运动行为关系的分析,提出一种Spiking神经网络(SNN)的分类算法。SNN的网络连接权值与突触连接的延时参数采用改进的粒子群优化方法(PSO)进行训练。仿真结果表明SNN分类效果优于群体向量法(PV)分类效果,有利于实现性能更高的用于神经康复的脑机接口系统。In this paper, a Spiking neural network （SNN） based classifier is used to analyze motor cortical neural signals. The neural ensemble data were recorded simultaneously with kinematics of arm movement while the monkey performed reaching tasks from the center position to eight peripheral targets in a three- dimensional virtual environment. The weighting and synapse time delay of SNN were updated by an improved particle swarm optimization （PSO） algorithm. The classification results of the SNN were compared with that of the population vector algorithm classifier. The results showed that the SNN classifier is better than PV classifier and SNN method holds hope for a possibly more accurate brain- computer interface for neural prosthesis.

关 键 词：脉冲神经网络 分类 粒子群优化 群体向量法 脑机接口 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]