基于贝叶斯估计的初级视皮层光栅朝向编码研究

Orientation Encoding of Primary Visual Cortex Based on Bayesian Estimation Theory

为了认知初级视皮层(V1区)的朝向编码机制,提出一种基于贝叶斯估计理论综合大鼠V1区神经元发放率和发放时间间隔(Inter-Spike-Interval,ISI)特性的复合编码方法.设计不同朝向的光栅对LE大鼠进行刺激,使用微电极阵列采集初级视觉皮层的锋电位序列,采用贝叶斯估计理论,综合V1区神经元的发放率和ISI特性对光栅朝向进行编码.将其编码结果与单独基于发放率的编码结果进行了对比分析.结果表明:复合编码方式对朝向编码的正确率高于发放率编码方式的正确率,也证明了神经元发放时间间隔与发放率共同参与了对光栅朝向信息的编码.

In order to recognize the orientation encoding mechanism in the primary visual cortex（V1）,a hybrid encoding method integrating firing rates and the inter-spike-intervals of neurons in rats＇ primary visual cortex was proposed based on Bayesian estimation theory here.The gratings with different orientations were used to stimulate the anesthetic LE rats,and their spike sequences were collected by the microelectrode array,the firing rates and inter-spike-intervals were integrated to encode the orientation using Bayesian estimation theory.And the results were compared with the encoding ones with firing rates simply.The results showed that the encoding precision of the hybrid method was higher than the precision of the firing rate method.It is also proved that inter-spike-intervals and the firing rates in rats＇ area V1 encode the orientation together.