Segmentation method based on transition region extraction for coronary angiograms

Aiming at the complex background of coronary angiograms, weak contrast between the coronary arteries and the background, a new segmentation method based on transition region extraction is proposed. Firstly, we construct 6 different Gaussian templates that are used to enhance the coronary angiograms. Then the transition region is extracted by using local entropy-based on transition region extraction method. Finally, determine the segmentation threshold and the vessels are obtained. The experiments indicate that the proposed method outperforms the method using morphological operator on the small vessels extraction, connectivity and effectiveness. In addition, the method is indeed valuable for diagnosis and the quantitative analysis of coronary arteries.