Segmentation of liver metastasis on CT images using the marker-controlled watershed and fuzzy connectedness algorithms
Jiayong Yan, John Q. Fang · 2015
Accurate quantification for the size of liver metastasis on CT images is critical to surgery/treatment planning and therapy response assessment. To date, there are still no practical methods for automatic or semiautomatic segmentation of liver metastases. This paper presents a method, which combines the marker-controlled watershed transform and fuzzy connectedness algorithm for semiautomatic delineation of liver metastases on contrast-enhanced sequential CT images. The key to successful use of marker-controlled watershed transform is to reliably determine internal and external markers, which is also the focus of this work. With the fuzzy connectedness technique, we propose a practical method to determine the internal and external markers for the liver metastasis in CT images. The performance of the proposed method was evaluated over 30 liver metastases from 10 patients. The results manually delineated by a radiologist served as the “gold standard” for comparison. The preliminary results have shown the potential of this algorithm for the segmentation of liver metastases on CT images.