基于局部C-V水平集的CT肝脏病灶提取  

Segmentation of Focal Liver Lesions in CT Images Based on Local C-V Level Set

作　　者：罗朝辉[1,2] 吴效明[1] 欧陕兴[3] 岑人经[1] 

机构地区：[1]华南理工大学生物医学工程系,510006 [2]湖南科技大学信息与电气工程学院,411201 [3]广州军区总医院放射科,510010

出　　处：《微计算机信息》2010年第1期4-5,127,共3页Control & Automation

摘　　要：本文针对肝脏CT图像的特点,提出一种局部C-V模型水平集算法对肝脏病灶进行分割。该算法首先在肝脏内部选择包含病灶的局部图像,再采用局部最大类间方差法进行预分割,根据最佳阈值确定初始水平集,最后采用局部C-V模型对初始轮廓曲线进行演化。实验结果表明该方法能较好地提取出肝脏病灶。In this paper,we propose a level set algorithm based on local C-V model to segment CT focal lesions in accordance with the features of CT abdominal CT images.We first select a region of interest in the liver region comprising the focal lesion to create a local image,then by using Maximum variance between clusters algorithm we automatically get the threshold to pre-segment the local image and get the initial level set,at last we use the local C-V model level set to get the final contour.The experiments showed that the algorithm can efficiently segment the focal liver lesion.

关 键 词：腹部CT图像 肝脏病灶 水平集 局部C-V模型 最大类间方差 

分 类 号：R318.103[医药卫生—生物医学工程]