Translator Disclaimer
9 May 2002 Vessel-valley-course generation algorithm for quantitative analysis of angiograms
Author Affiliations +
Three dimensional reconstruction and quantitative analysis of angiograms require vessel centerline determination and tracking. In a vessel profile, it is more straightforward to locate the valley point (local minimum) than the center. Therefore, vessel valley courses offer advantages over centerlines in terms of natural features and easy-to-locate. We propose a 'star' scan technique to generate a valley map, which is then traced to determine the valley courses. The angiogram is scanned along the horizontal, vertical, diagonal and anti-diagonal directions. The scan pattern resembles a 'star'; therefore, it is referred to as a 'star' scan. The scanning along each direction provides an image consisting of scan profiles, which may be multi-modal functions. We then detect and record the local minimum locations, thereby generating a valley map. By searching over the valley map, we can generate valley courses, which can be used for vessel quantitative analysis and 3-D reconstruction. Using the valley course, it is a straightforward process to generate centerlines. This is a robust and easily implementable algorithm for quantitative analysis of angiograms. Experimental validation of the algorithm will be reported using coronary angiograms and phantom images.
© (2002) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Zikuan Chen and Sabee Y. Molloi "Vessel-valley-course generation algorithm for quantitative analysis of angiograms", Proc. SPIE 4684, Medical Imaging 2002: Image Processing, (9 May 2002);

Back to Top