AN IMPROVED RANDOM WALK SEGMENTATION ON THE LUNG NODULES

作者：Guo, Li^*; Zhang, Yunting; Zhang, Zewei; Li, Dongyue; Li, Ying

来源：International Journal of Biomathematics, 2013, 6(6): 1350043.

DOI：10.1142/S1793524513500435

• Lung nodules
• Gabor
• FCM
• ellipse fitting
• random walk

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.

• 出版日期2013-11
• 单位天津医科大学