基于变分推断的磁共振图像群组配准
周勤,王远军

Groupwise Registration for Magnetic Resonance Image Based on Variational Inference
Qin ZHOU,Yuan-jun WANG
图3 概率生成模型图解.白圈表示隐含变量,灰圈表示观测值(输入图像),方块表示参数.虚线矩形框表示里面的变量xyz独立重复N次.配准先验由高斯参数的均值μz和协方差Σz定义
Fig.3 Probability generation model. White circle represents hidden variable, gray circle represents observed value (input image), and squares represent parameters. Dashed rectangle indicates the number of independent repeats N of the x, y, z variables inside. Registration prior is defined by the mean value μz and covariance Σz of Gaussian parameters