基于区域的主动轮廓模型的医学图像应用

Application of Region-based Active Contour Model for Medical Image

对脑部MRI中各个大脑组织的准确分割在临床医学上具有重要的意义。该文综述了MR图像不均匀性校正方法及近几年来脑部MRI的分割方法的研究新进展,包括聚类算法、基于形变模型的分割方法、图割算法、形态学分水岭算法和基于图谱的MRI分割算法等。综合多种方法的优点,取长补短,在算法中引入图像的空间结构信息、纹理信息以及一些先验知识,有望在新理论技术上和新工具方面有所突破。

Accurate segmentation of brain MRI in various brain tissue has important significance in clinical medicine.This paper reviews the research progress of MR image intensity inhomogeneities correction method and in recent years the brain MRI segmentation methods,including clustering and segmentation method based on deformable model,the graph cut algorithm,Morphological Watershed Algorithm and MRI segmentation algorithm based on map etc..Combined with advantages of various methods to complement each other,introducing image spatial information and texture information in the algorithm as well as some prior knowledge,breakthrough is expected in new theory and new technology tool.