Automatic characterization of Agatston score from coronary computed tomography
Automatic characterization of the Agatston score from coronary computed tomography (CT) is provided. In various embodiments, a plurality of coronary computed tomography images are segmented into a plurality of segments corresponding to features of coronary anatomy. A plurality of calcium candidates are extracted from the plurality of coronary computed tomography images by thresholding. Coronary calcification is located in the coronary computed tomography images by applying a trained classifier to the plurality of calcium candidates. An Agatson score is computed from the located calcification.
1. A method comprising:
segmenting a plurality of coronary computed tomography images into a plurality of segments corresponding to features of coronary anatomy, wherein segmenting the plurality of coronary computed tomography images comprises applying an atlas;
based on the plurality of segments, extracting a plurality of calcium candidates from the plurality of coronary computed tomography images by thresholding;
locating coronary calcification in the coronary computed tomography images by applying a trained classifier to the plurality of calcium candidates;
computing an Agatson score from the located calcification.
2. The method of claim 1 , wherein the features of coronary anatomy comprise the aorta, the coronary artery, the pulmonary artery, the four chambers of the heart, or the myocardium.
3. The method of claim 1 , wherein applying the atlas comprises performing joint atlas label fusion.
4. The method of claim 1 , wherein the trained classifier is a random forest classifier.
5. The method of claim 1 , wherein applying the trained classifier comprises computing a plurality of features of the plurality of calcium candidates and supplying the plurality of features to the trained classifier.
6. The method of claim 5 , wherein the plurality of features comprises position relative to the features of coronary anatomy, shape, size, or texture.
7. The method of claim 1 , wherein the plurality of coronary computed tomography images are without contrast.
8. The method of claim 1 , wherein segmenting the plurality of coronary computed tomography images comprises cropping the plurality of coronary computed tomography images based on the location of a lung.
9. The method of claim 1 , wherein segmenting the plurality of coronary computed tomography images comprises rigid registration.
10. The method of claim 1 , wherein segmenting the plurality of coronary computed tomography images comprises non-rigid registration.
11. The method of claim 1 , wherein segmenting the plurality of coronary computed tomography images comprises smoothing a preliminary segmentation.
12. The method of claim 1 , wherein extracting the plurality of calcium candidates comprises independent connected component analysis.
13. The method of claim 1 , wherein extracting the plurality of calcium candidates comprises excluding candidates having a volume below a predetermined threshold.
14. The method of claim 1 , wherein extracting the plurality of calcium candidates comprises excluding candidates having a volume above a predetermined threshold.
15. The method of claim 1 , wherein applying the trained classifier comprises computing a plurality of features of a neighborhood of calcium candidates and supplying the plurality of features to the trained classifier.
16. The method of claim 15 , wherein the neighborhood of calcium candidates comprises a predetermined volume around each candidate.
17. The method of claim 5 , wherein the plurality of features comprises average, minimum or maximum intensity, standard deviation of intensities, volume, blobness, tublarness, compactness, displacement from a heart center, spatial location position, or anatomy probability.
18. A system comprising:
a computing node comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor of the computing node to cause the processor to perform a method comprising:
segmenting a plurality of coronary computed tomography images into a plurality of segments corresponding to features of coronary anatomy, wherein segmenting the plurality of coronary computed tomography images comprises applying an atlas;
based on the plurality of segments, extracting a plurality of calcium candidates from the plurality of coronary computed tomography images by thresholding;
locating coronary calcification in the coronary computed tomography images by applying a trained classifier to the plurality of calcium candidates;
computing an Agatson score from the located calcification.
19. A computer program product for automatic calcium detection, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to perform a method comprising:
segmenting a plurality of coronary computed tomography images into a plurality of segments corresponding to features of coronary anatomy, wherein segmenting the plurality of coronary computed tomography images comprises applying an atlas;
based on the plurality of segments, extracting a plurality of calcium candidates from the plurality of coronary computed tomography images by thresholding;
locating coronary calcification in the coronary computed tomography images by applying a trained classifier to the plurality of calcium candidates;
computing an Agatson score from the located calcification.
20. The system of claim 18 , wherein applying the atlas comprises performing joint atlas label fusion.