基于窄带思想的虚拟人肺组织形状演化分割方法  

Method of virtual human lung shape propagation segmentation based on narrowband thinking

作　　者：徐婧[1] 吴健[1] 叶峰[1] 崔志明[1] 

机构地区：[1]苏州大学智能信息处理及应用研究所,江苏苏州215006

出　　处：《计算机工程与设计》2009年第4期931-933,共3页Computer Engineering and Design

基　　金：江苏省高校科研成果产业推进工程基金项目(JHB06-26);苏州市科技攻关(工业)基金项目(SGR0703)

摘　　要：由于肺组织形状的广泛相异性,且肺组织与其周围的组织具有相似的像素分布,致使肺组织边界较为模糊,使得医学图像中对肺器官的自动分割实现比较困难。提出了一种基于窄带思想的肺组织形状演化分割方法,将前一帧图像通过水平集算法分割所得的肺组织轮廓线经过基于窄带思想的形状演化后,作为下一帧肺组织图像分割的初始种子,肺组织分割结束后用可视化技术重建人体肺组织。实验结果表明,该方法对肺组织分割行之有效,与同类算法相比,不仅减少了人工干预量,而且分割效率明显提高。Automatic lung segmentation is difficult because of the wide range of human variations in the shapes of the lung. In addition, nearby organs and tissues have similar intensity distributions to the lung, making the lung＇s boundaries ambiguous. In this study, a method of lung shape propagation segmentation based on the narrowband thinking is proposed, the contour lines of the lung which are adopted by level-set method through shape propagation based on the narrowband, as the initial seed of the next slice. After the lung segmentation, visualization technology is used for tissue reconstruction. The method is proved for efficient lung segmentation which has high efficiency compared with congener algorithm and decreases the manual work successfully.

关 键 词：窄带思想 形状演化 肺组织分割 水平集 可视化 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]