College of Automation, Guangdong University of Technology, Guangzhou510006, Guangdong, China.

---

Abstract—First, a region growing algorithm basedon Clifford algebra is put forward to segment blood vessels of abdominal aorta. Compared with the traditional region growing method, the proposed algorithm is easy to operate and good at robustness. Furthermore, it is available to segment both normal and abnormal blood vessels of abdominalaorta: the seed point can keep on growing even when there are some isolated points in region of blood vessels. The key point of this algorithm is that: the gray values of eight neighborhood points are composed as unit feature vectors of current seed points, and then Clifford algebra vector product representation will be applied. Second, for the purpose to segment thinner global abdominal blood vessels rapidly, a novel segmentation algorithm of blood vessels based on 2D Voronoi diagram is put forward as an attempt. Experiments demonstrate that it is potential to segment thinner global abdominal blood vessels fast. In the future, this algorithm is potential to segment blood vessels in abdominal organs.

Index Terms—Medical image segmentation, blood vessels, Clifford algebra, Voronoi diagram

[PDF]

Cite: Medical Image Segmentation of Blood Vessels Based on Clifford Algebra and Voronoi Diagram, "Medical Image Segmentation of Blood Vessels Based on Clifford Algebra and Voronoi Diagram," Journal of Software vol. 13, no. 6, pp. 页码, 2018.