Semi-automatic extraction of vascular networks in angiograms

A semi-automatic algorithm for the segmentation of angiographic images is proposed. First, the digital angiogram is smoothed by an image structure preserving technique which is based on homogeneity testing and anisotropic diffusion. Then, the watershed transform is applied to the smoothed image gradient magnitude and the resulting initial image partition is input to a fast hierarchical region merging process. Finally, the vessel regions are extracted from the segmented image by a simple point-and-click interactive procedure. Experimental results on Digital Subtracted Angiograms (DSA) are presented.

[1]  Milan Sonka,et al.  Robust simultaneous detection of coronary borders in complex images , 1995, IEEE Trans. Medical Imaging.

[2]  Jitendra Malik,et al.  Scale-Space and Edge Detection Using Anisotropic Diffusion , 1990, IEEE Trans. Pattern Anal. Mach. Intell..

[3]  A. Ross,et al.  Interpretation of coronary angiograms. , 1990, Journal of the American College of Cardiology.

[4]  Luc Vincent,et al.  Watersheds in Digital Spaces: An Efficient Algorithm Based on Immersion Simulations , 1991, IEEE Trans. Pattern Anal. Mach. Intell..

[5]  Nicos Maglaveras,et al.  Hybrid image segmentation using watersheds , 1996, Other Conferences.

[6]  Zhenyu Wu Homogeneity Testing for Unlabeled Data: A Performance Evaluation , 1993, CVGIP Graph. Model. Image Process..

[7]  José M. N. Leitão,et al.  A nonsmoothing approach to the estimation of vessel contours in angiograms , 1995, IEEE Trans. Medical Imaging.

[8]  Philippe Saint-Marc,et al.  Adaptive Smoothing: A General Tool for Early Vision , 1991, IEEE Trans. Pattern Anal. Mach. Intell..