Algorithm:GATech:Centerline Generation for Vessels

Back to NA-MIC_Collaborations

Objective:

The goal of this work is to generate centerlines from segmented 3D surfaces of blood vessels using a harmonic skeletonization technique. The generated centerlines are used as a guide to visualize and evaluate stenoses in human coronary arteries. A harmonic skeleton is the center line of a multi-branched tubular surface extracted based on a harmonic function, which is the solution of the Laplace equation. This skeletonization method guarantees smoothness and connectivity and provides a fast and straightforward way to calculate local cross-sectional areas of the arteries, and thus provides the possibility to localize and evaluate coronary artery stenosis, which is a commonly seen pathology in coronary artery disease.

Progress:

File:Skeletons.TIF

Figure 1 shows the centerline extraction results on two coronary artery tress. Left: Coronaries and the skeleton of a healthy volunteer. Middle: Coronaries and the skeleton of a patient with plaques in LAD. Right: The skeleton of the middle coronaries.

File:Areas.TIF

File:Stenosis quan.TIF

Cross-sectional areas of vessels can be calculated using the centerlines as a guide. Figure 3 shows the areas at different locations along the vessels by specifying a u value of the harmonic function (solution to the Laplace equation). We can also show a continuous variation of the cross-sectional area along a certain vessel, as shown in Figure 4, where the areas along the left anterior descending (LAD) coronary artery of the second case in Figure 1 are measured. Figure 5 shows a stenosis detected in this vessel.

Ongoing Work

• This method is being tested and evaluated for its accuracy based on more clinical data sets.
• The algorithms used in this work are currently implemented in MATLAB and C.

References:

[1] Y. Yang, L. Zhu, S. Haker, A. Tannenbaum, and D. Giddens. [/Wiki/images/8/83/YangMICCAI2005.pdf Harmonic Skeleton Guided Evaluation of Stenoses in Human Coronary Arteries]. In Proc. MICCAI 2005: International Conference Medical Image Computing and Computer-Assisted Intervention, Lecture Notes in Computer Science (3749). Springer-Verlag, pp.490-497, 2005

Links:

• Stony Brook Algorithms
• NA-MIC Collaborations