基于匹配小波变换的初级视皮层神经元锋电位分类

Neuronal Spike Sorting in Primary Visual Cortex Based on Matched Wavelet Transform

神经元锋电位分类是研究神经系统信息处理机制的关键.为了提高锋电位分类效果,将匹配小波应用到锋电位分类中来,提出了基于匹配小波变换的初级视皮层神经元锋电位分类方法.首先根据提取的锋电位信号波形特性和匹配准则,构造出匹配小波,然后基于匹配小波对锋电位信号进行小波变换,获取锋电位信号特征进行锋电位聚类,从而实现锋电位分类.将该方法用于由多电极阵列采集的大鼠初级视皮层神经元锋电位的分类,并与基于传统小波的分类方法进行聚类比较,该方法得到的散点图具有两个更加明显的聚类中心,聚类效果更好.再对信号进行加噪处理,当信噪比较低时,该方法仍保持较高的分类正确率,抗噪性能较强.

Sorting of neuronal spike is the key to research information processing mechanism of nervous system.To improve sorting effect,matched wavelet is applied and neuronal spike sorting in primary visual cortex based on matched wavelet transform is introduced.Firstly,matched wavelet is constructed according to character of spike signals and match criterion.Then matched wavelet transform is used to extract features of spike signals and clustering of spike signals is carried out.Sorting of spike signals is realized.The method proposed is applied to analyze neuronal spike signals of rat in the primary visual cortex,sampled by multi-electrode array.Compared with sorting method based on classical wavelet,scatter plot generated by new method has two clearer clustering centers.Clustering effect is better.When noise is added,correct rate of new sorting method has remained high under low signal noise ratio.New method has strong noise resistance.