Difference between revisions of "Projects:BloodVesselSegmentation"

From NAMIC Wiki
Jump to: navigation, search
Line 65: Line 65:
  
 
Fig. 9.
 
Fig. 9.
 +
 +
= Vessel segmentation using Tubular Surface Extraction framework =
 +
In this work, we are extending the Tubular Surface Extraction framework of Mohan et al towards segmenting vessel structures. The blood vessel is modeled as a tube with a center-line and a radius function associated with each point. Further, the work is also being extended to accomodate evolution of end points. This allows a segmentation framework where a portion of the main branch of the vessel tree can be selected as input and the framework evolves this to capture the entire vessel tree. Fig. 10 shows the results from the application of the fixed end points version of this framework.
 +
 +
[[Image:Mohan_Tubular_Vessels_1.PNG | Figure 10]]
 +
 +
 +
  
  
Line 70: Line 78:
 
= Key Investigators =
 
= Key Investigators =
  
* Georgia Tech Algorithms:Yan Yang, Ponnappan Arumuganainar, Vandana Mohan, Allen Tannenbaum  
+
* Georgia Tech Algorithms:Vandana Mohan, Shawn Lankton, Yan Yang, Ponnappan Arumuganainar, Allen Tannenbaum  
  
 
= Publications =
 
= Publications =

Revision as of 18:05, 10 December 2008

Home < Projects:BloodVesselSegmentation
Back to Georgia Tech Algorithms

Blood Vessel Segmentation

Atherosclerosis is a systematic disease of the vessel wall that occurs in the aorta, carotid, coronary and peripheral arteries. Atherosclerotic plaques in coronary arteries may cause stenosis (narrowing) or complete occlusion of the arteries and lead to serious results such as heart attacks. Imaging techniques have greatly assisted the diagnoses and treatment procedures of atherosclerosis. Three dimensional imaging such as CTA for coronary arteries is a relatively new approach but has great potentials for detecting and evaluating coronary calcification and stenosis. Fig. 1 (b) shows an example of the 3D reconstruction of coronary arteries and the aorta.

Description

Figure 1

A novel image segmentation approach is proposed combining Bayesian pixel classification method and the active surface model in a level set formulation to extract coronary arteries from CT angiography images. Fig. (2) shows the reconstructed coronary arteries from three different patients, and Fig. (3) are sample slices showing the original images and the delineated vessels as cross-sections.

Figure 2

Once the surface of the coronaries are reconstructed, further shape analysis and measurements can be conducted based on it. Fig. (4) shows the results of performing centerline extraction using a hamonic skeletonization technique [3]. The skeletons can then serve as a guide for finding the perpendicular planes to the arteries, and these planes are used to intersect with the vessel in order to measure the local cross-sectional areas, as shown in Fig. (5).

Figure 3

Figure 4

Figure 5

Soft Plaque Detection and Segmentation

Recent studies have shown that the soft plaque is more vulnerable to rupture than the hard plaque. Hence it becomes necessary to develop methods to detect and segment the soft plaque automatically. The soft plaque has an intensity that lies between the intensities of the blood lumen and the cardiac muscle, thus making it difficult to be detected using the energy calculated globally. In the following section, we briefly discuss a local region based energy and the evolution based on that energy that was used to detect the soft plaque.

Local Region based energy

For the purpose of soft plaque detection we used a hybrid energy scheme proposed by Lankton et al. This scheme blends the benefits of the geodesic active contours and the region based active contours and is accomplished by forming a geodesic energy from local regions around the curve (zero level set). The energy schemes of geodesic active contours, region based active contours (Chan-Vese) and a hybrid scheme (using Chan-Vese) are given below.

Figure 6

Fig. 6. Here, C represents the evolving curve, I represents the image data, f is any positive and decreasing function of the image data, u and v represent the mean image intensities inside and outside the curve. In the hybrid energy, the image values and averages from a local region around the curve (zero level set) are used.

The resulting flow is more robust to initial curve placement and image noise, like the region-based flows, but also capable of finding significant local minima and partitioning the image without making global assumptions about its makeup. This makes it ideal for soft plaque detection. The results obtained by Lanton et al while segmenting the Putamen from MRI volume is given below. The Hybrid flow seems to give a better result.

Figure 7

Fig. 7. A 2D slice of a 3D MRI image of the putamen being segmented by several methods.

(a) The initial contour.

(b) Attempted segmentation using the Chan-Vese region-based active contour.

(c) Attempted segmentation using edge-based geodesic active contours.

(d) Correct segmentation using the presented hybrid flow.


Results of local region based evolution in 2D

The results of the local region based evolution on a 2D slice of CTA volume is given below. The Red contour is the output of the vessel segmentation algorithm discussed in the previous section. The Blue contour is the result of the local region based evolution using Chan-Vese energy. The region between the Red and the Blue contour is the soft plaque and the wall of the coronary vessel.

Figure 8

Fig. 8.


Results of local region based evolution in 3D

The results of the local region based evolution in 3D is given below. The local evolution was done for one of the two arteries visible in the slices shown below.


Figure 9

Fig. 9.

Vessel segmentation using Tubular Surface Extraction framework

In this work, we are extending the Tubular Surface Extraction framework of Mohan et al towards segmenting vessel structures. The blood vessel is modeled as a tube with a center-line and a radius function associated with each point. Further, the work is also being extended to accomodate evolution of end points. This allows a segmentation framework where a portion of the main branch of the vessel tree can be selected as input and the framework evolves this to capture the entire vessel tree. Fig. 10 shows the results from the application of the fixed end points version of this framework.

Figure 10




Key Investigators

  • Georgia Tech Algorithms:Vandana Mohan, Shawn Lankton, Yan Yang, Ponnappan Arumuganainar, Allen Tannenbaum

Publications

In Print