一种从多维Spike Train数据中进行神经解码的方法

A method on neuron decoding from multi-dimension Spike Train data

神经元集群编码与解码是神经信息处理的关键问题。本文首先引入放电率与峰电位计数率度量神经元对外界刺激的响应,分析了根据实验记录的脑信号电波获得神经电位活动的规律。利用微电极阵列记录并分离脑电信号,然后借助现代密码学的思想,将之与多维随机数列进行比较,得到有效信息量从而在Spike Train中进行神经解码。最后用一个实例来分析了发现模式的具体步骤。

The coding and decoding of neuron clustering is a critical issue in neural information processing.In this paper,the discharge rate and spike potential counting rate are introduced to measure the response of neuron for external stimulus,and the regulation of neuron potential according to brain signal wave from experimental records is analysed.Then,recording and separating brain wave signal with micro-electrode array are done,and compared them with multi-dimension random numbers by means of theory of modern cryptology to obtain the effective information.Consequently,decoding is performed in the Spike Train.Finally,the specific steps of discovery pattern is analysed with a case.