鸽子运动意图的最大似然估计解码算法

Decoding of Pigeon’s Movement Intentions Based on Maximum Likelihood Estimation

神经信息解码是目前植入式脑机接口(brain-computer interface,BCI)神经信息处理研究中的难点和重点;解码效果的优劣以及解码算法的效率直接决定了脑机接口应用的有效性和实用性。为了解码十字迷宫内鸽子运动转向信息,利用高斯分布模型对神经元锋电位发放率的概率密度函数进行建模;并结合最大似然估计(maximum likelihood estimation,MLE)算法对鸽子的转向意图进行了预测;并将其结果与(support vector machine,SVM)法和群矢量(population vector,PV)法进行了比较。结果表明,MLE算法能够有效地解码鸽子的运动意图,解码正确率显著高于SVM法和PV法。这一结果也为进一步分析鸽子运动意图神经信息处理机制奠定了基础。

Neural information decoding is the difficulty and keystone in neural information processing of implanted brain-computer interface(BCI).The advantages and disadvantages of decoding results and the efficiency of decoding algorithm directly determine the validity and practicability of BCI.In order to decode the movement intentions in the cross maze,the Gaussian distribution model was used to model the probability density function of spike s firing rates,and the maximum likelihood(ML)estimation algorithm was used to predict the pigeon s movement intentions.And its result was compared with the result of both support vector machine(SVM)algorithm and the population vector(PV)algorithm.The results show that the ML method can effectively decode the pigeon s motion intention,and the decoding accuracy is significantly higher than both the SVM method and the PV method.This results also lays a foundation for further analysis of the neural information processing mechanism of pigeon movement intention.