基于MDL形状模型的医学图像分割

Medical image segmentation based on MDL shape model

主动形状模型是一种有效的图像分割方法,但在建立模型的过程中需要从训练集中建立满足对应关系的轮廓采样点集合。传统采用手动标定的方法工作耗时,且充满主观性,不能向三维空间拓展。文章使用了一种自动的方法,通过构造基于最小描述长度准则的目标函数,获得最优化意义上的点分布模型,然后建立主动形状模型,该方法提高了建模的效率并具有客观可重复性。通过对于腰部与膝关节MRI图像的分割实验证明,该文方法可获得较好的图像分割效果。

Active shape model（ASM） is an efficient method of image segmentation.One key factor in building this model is to obtain correspondent landmarks among the whole shape dataset.Traditional manual landmarking is time-consuming and subjective,and cannot be conducted in three-dimensional space.This paper adopts a multi-scale parameterization approach which allows minimum description length（MDL） based optimization on landmark correspondence.This achieves an effective,objective and repeatable ASM building.The segmentation test on MRI images of waist and knee joint verifies the good performance of the proposed method.