Publication:
Segmentation of liver structure from computed tomogrhapy dataset

Loading...
Thumbnail Image
Date
2010-04-01
Authors
Thong, Kee Sin
Journal Title
Journal ISSN
Volume Title
Publisher
Research Projects
Organizational Units
Journal Issue
Abstract
Accurate segmentation of liver structure from CT images is an essential step in the medical field such as surgery planning and diagnosis of liver disease. Manual segmentation is done by manually tracing the liver contour on each image slice, which is a slow and time-consuming process. In this project, a new algorithm of the liver segmentation is proposed. The idea of the proposed algorithm is based on the combination of various image processing techniques. The algorithm’s process flow can be divided into four stages. In the first stage, a ROI is defined by using prior knowledge about the liver position to obtain the initial threshold values for automatic segmentation. The second stage integrates region-based segmentation with thresholding to detect coarse liver object. The third stage is post-processing that utilizes morphological operations iteratively with region-labeling. In order to perform accurate segmentation, the last stage is image refinement with 3D morphology. The implemented algorithm is tested on a given CT dataset to evaluate the effectiveness in the liver segmentation by comparing the results to ground truth. Overall, the proposed algorithm successfully segments the liver in most of the slices with satisfying results, although some of the results are under-segmented or over-segmented.
Description
Keywords
Citation