Robust segmentation of blood vessels from angiographic images of the human heart

Although a variety of edge detection algorithms are available, they do not always lead to acceptable results in extracting various features in an image. In this paper, an algorithm for detecting blood vessels in angiographic images of human heart is presented. The gray-level profile of the cross section of a blood vessel is approximated by a Gaussian shaped curve. Blood vessels usually have poor local contrast and the existing edge detection algorithms do not give better results. Now the proposed algorithm using morphology works well for extracting blood vessels which are linear in nature in angiographic images. Results on various medical data from a normal heart and from a set of abnormalities are presented and show that this algorithm can be used as a robust segmentation tool.