一种优化树状模型的肝脏自动分割方法  

An Automatic Segmentation Method of Liver with Optimized Tree Model

作　　者：周丽芳[1,2] 王璐 李伟生 雷帮军[2] 许志[1] ZHOU Lifang;WANG Lu;LI Weisheng;LEI Bangjun;XU Zhi(College of Software Engineering,Chongqing University of Posts and Telecommunications,Chongqing 400065,China;College of Computer Science and Technology, Chongqing University of Posts and Telecommunications,Chongqing 400065,China;College of Computer and Information Technology,China Three Gorges University,Yichang,Hubei 443002,China)

机构地区：[1]重庆邮电大学软件工程学院,重庆400065 [2]三峡大学计算机与信息学院,湖北宜昌443002 [3]重庆邮电大学计算机科学与技术学院,重庆400065

出　　处：《计算机工程》2019年第2期226-232,共7页Computer Engineering

基　　金：国家自然科学基金(61100114;U1401252);重庆市自然科学基金(cstc2015jcyjA40011);重庆市研究生科研创新项目(CYS18247);水电工程智能视觉监测重点实验室开放基金(2017SDSJ02)

摘　　要：针对CT图像的复杂性与肝脏形态的多样性,提出一种优化的肝脏自动分割方法。通过基于部件共享池的混合树状模型(TSPM)捕获肝脏边界的拓扑形态变化,利用凹凸点算法根据肝脏形变自动筛选TSPM中的关键点,避免冗余点对肝脏边界的错误定位,并将不同颜色空间应用于肝脏图像分割中提高分割精度。实验结果表明,与现有分割方法相比,优化方法可获得更准确的肝脏分割结果。Aiming at the complexity of CT images and the diversity of liver morphology,an automatic optimized liver segmentation method is proposed.Through a hybrid tree model based on the shared pool of components,called TSPM,the topological changes of the liver boundary can be captured.The proposed Convex Concave Point(CCP) algorithm can automatically screen out key points to avoid the misalignment of redundant points on the liver boundary.It applys different color spaces to liver image segmentation to improve segmentation accuracy.Experimental results show that compared with the existing segmentation method,the optimized method can obtain more accurate liver segmentation results.

关 键 词：肝脏分割 可变形部件模型 树状模型 凸凹点算法 颜色空间 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]