Automated Extraction of the Coronary Tree by Integrating Localized Aorta-Based Intensity Distribution Statistics in Active Contour Segmentation

State-of-the-art Computed Tomography Angiography (CTA) scanners are capable of acquiring rigorous 3D vasculature information. Blood filled vessels are extracted from the data cloud for pathological analysis on the basis of intensity value, measured in Hounsfield units. Setting a hard threshold in CTA images for differentiating coronaries from fatty muscles of heart could be misleading as it lacks behavioural information of the contrast agent in the respective CTA volume. It is common for under-or over-segmentation to occur due to the improper diffusion of the contrast agent in the different branches. This problem motivates research to determine an optimal threshold for volumes individually by examining the behaviour of the contrast agent. In this work, intensity distribution statistics (extracted from the segmented aorta through an examination of the initial CTA axial slices) is integrated in the curve evolution process to track the progression of coronary arteries. Optimal threshold value is obtained individually for 12 clinical volumes by Gaussian fitting of the aorta intensity histogram. The obtained range is validated by comparing the intensity values of manually selected coronary segments for each volume at 50 random points. The automatic segmentation process starts with the detection of a coronary seed point based on geometric analysis of the aorta. In the subsequent stages, the derived intensity threshold value and seed point are used in localized active contour-based segmentation for precise delineation of vessel boundaries. Initial visual results appear promising and validate the standard anatomical structure of coronary trees, whereas statistical quantification is in process.

[1]  Allen R. Tannenbaum,et al.  Localizing Region-Based Active Contours , 2008, IEEE Transactions on Image Processing.

[2]  Chunming Li,et al.  Implicit Active Contours Driven by Local Binary Fitting Energy , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[3]  Anthony J. Yezzi,et al.  A statistical approach to snakes for bimodal and trimodal imagery , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[4]  Bram van Ginneken,et al.  Automatic detection of calcifications in the aorta from CT scans of the abdomen1 , 2004 .

[5]  M. Reiser,et al.  Coronary artery calcium: absolute quantification in nonenhanced and contrast-enhanced multi-detector row CT studies. , 2002, Radiology.

[6]  Rachid Deriche,et al.  Adaptive Segmentation of Vector Valued Images , 2003 .

[7]  Joachim Hornegger,et al.  Semi-automatic level-set based segmentation and stenosis quantification of the internal carotid artery in 3D CTA data sets , 2007, Medical Image Anal..

[8]  J. Sethian,et al.  Fronts propagating with curvature-dependent speed: algorithms based on Hamilton-Jacobi formulations , 1988 .

[9]  S. Osher,et al.  Algorithms Based on Hamilton-Jacobi Formulations , 1988 .

[10]  Don P. Giddens,et al.  AUTOMATIC SEGMENTATION OF CORONARY ARTERIES USING BAYESIAN DRIVEN IMPLICIT SURFACES , 2007, 2007 4th IEEE International Symposium on Biomedical Imaging: From Nano to Macro.

[11]  Alejandro F. Frangi,et al.  Muliscale Vessel Enhancement Filtering , 1998, MICCAI.

[12]  Isabelle Bloch,et al.  A review of 3D vessel lumen segmentation techniques: Models, features and extraction schemes , 2009, Medical Image Anal..

[13]  Tony F. Chan,et al.  Active contours without edges , 2001, IEEE Trans. Image Process..

[14]  Demetri Terzopoulos,et al.  Snakes: Active contour models , 2004, International Journal of Computer Vision.