无监督的猕猴运动皮层锋电位信号CKF解码

发布时间：2019-05-22 22:07

【摘要】：如何通过猕猴运动皮层的神经元锋电位信号估计其手指移动位置是一神经解码问题,现存方法解决该问题大多采用有监督训练,需要通过训练数据得到神经元锋电位信号与手指移动位置的关系,因此其估计性能依赖于训练数据.本文提出了一种无监督解码方法,该方法基于状态空间模型(State space model,SSM),利用神经网络得到神经元锋电位数与手指移动位置的关系权值,再用逐次状态估计方法去估计手指移动的位置.为减少训练的复杂度和提高估计准确度,采用一种非线性的积分卡尔曼滤波(Cubature Kalman filtering,CKF)来完成神经网络的训练和手指位置的逐次状态估计.与传统方法相比,该方法的最大特点是无监督,可以由神经元锋电位簇向量直接估计手指移动位置,而无需有监督训练.实验结果显示,当采用较少的有监督数据,现存方法与本文方法相比有较大的估计误差;当采用较多的有监督数据,现存方法才具有与本文方法相近似的估计误差.
[Abstract]:How to estimate the finger moving position through the neuron front potential signal of the motor cortex of rhesus monkeys is a neural decoding problem, and most of the existing methods use supervised training to solve this problem. The relationship between the neuron front potential signal and the finger moving position needs to be obtained from the training data, so its estimation performance depends on the training data. In this paper, an unsupervised decoding method is proposed, which is based on the state space model (State space model,SSM). The neural network is used to obtain the relationship between the number of neuron front potentials and the position of finger movement. Then the successive state estimation method is used to estimate the position of finger movement. In order to reduce the complexity of training and improve the accuracy of estimation, a nonlinear integral Kalman filter (Cubature Kalman filtering,CKF) is used to complete the training of neural network and the successive state estimation of finger position. Compared with the traditional method, the biggest characteristic of this method is that it can estimate the finger moving position directly from the neuron front potential cluster vector without supervised training. The experimental results show that when less supervised data are used, the existing methods have larger estimation errors than the present method, and when more supervised data are used, the existing methods have estimation errors similar to those of this method.
【作者单位】：云南民族大学电气信息工程学院;
【基金】：国家自然科学基金(61262091) 云南省第17批中青年学术和技术带头人资助项目(2014HB019) 云南省教育厅科学基金重点项目(2014Z093) 云南民族大学研究生创新基金项目(2016YJCXS03)资助~~
【分类号】：TN713;R338

【二级参考文献】