AN IMPROVED RANDOM WALK SEGMENTATION ON THE LUNG NODULES  

作　　者：LI GUO YUNTING ZHANG ZEWEI ZHANG DONGYUE Li 

机构地区：[1]School of Medical Imaging, Tianjin Medical University Tianjin 300203, P. R. China [2]General Hospital, Radiology, Tianjin Medical University Tianjin 300203, P. R. China

出　　处：《International Journal of Biomathematics》2013年第6期105-120,共16页生物数学学报（英文版）

基　　金：Acknowledgments This work was supported by the National Natural Science Foundation of China （Project Nos. 81000639 and 31000450）, China Postdoctoral Science Foundation （Project Nos. 20100470791 and 201104307）, and Program of the Pearl River Young Talents of Science and Technology in Guangzhou （No. 2012J2200041）.

摘　　要：In this paper, we proposed a semi-automatic technique with a marker indicating the target to locate and segment nodules. For the lung nodule detection, we develop a Gabor texture feature by FCM （Fuzzy C Means） segmentation. Given a marker indicating a rough location of the nodules, a decision process is followed by applying an ellipse fitting algorithm. From the ellipse mask, the foreground and background seeds for the random walk segmentation can be automatically obtained. Finally, the edge of the nodules is obtained by the random walk algorithm. The feasibility and effectiveness of the proposed method are evaluated with the various types of the nodules to identify the edges, so that it can be used to locate the nodule edge and its growth rate.

关 键 词：Lung nodules GABOR FCM ellipse fitting random walk. 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] S851.347[自动化与计算机技术—计算机科学与技术]