Abstract
This article presents an algorithm for the segmentation of liver metastases in CT scans. It is a hybrid method that combines adaptive thresholding based on a gray value analysis of the ROI with model-based morphological processing. We show the results of the MICCAI liver tumor segmentation competition 2008 which were successful for all ten tumors.
Keywords
Source Code and Data
No source code files available for this publication.
Reviews
Xiang Deng
Friday 25 July 2008
This paper presents a semi-automatic live tumor segmentation technique.
The segmentation is initialized with a user-drawn stroke across the lesion. The tumors are then segmented using adaptive thresholding followed with refinement by morphological operation.
Detailed comments:
1) The proposed technique requires a user-drawn stroke to initialize the segmentation, and some manual editing in refinement. In my opinion, this algorithm should belong to "Interactive system".
2) It would be helpful if the authors could provide some information regarding the sensitivity of the proposed algorithm to the initial stroke, i.e., variation of segmentation against different positions of the initial stroke.
3) In Table 1, please use the provided tumor name, e.g., IMG05_L1.