System and method for coronary digital subtraction angiography
View Patent ↗A method and system for extracting motion-based layers from fluoroscopic image sequences are disclosed. Portions of multiple objects, such as anatomical structures, are detected in the fluoroscopic images. Motion of the objects is estimated between the images is the sequence of fluoroscopic images. The images in the fluoroscopic image sequence are then divided into layers based on the estimated motion. In a particular implementation, the coronary vessel tree and the diaphragm can be extracted in separate motion layers from coronary angiograph fluoroscopic image sequence.
1. A method of extracting motion-based layers from a sequence of fluoroscopic images, comprising:
detecting a portion of each of a plurality of objects in said sequence of fluoroscopic images;
estimating motion of the detected portion of each of said plurality of objects in said sequence of fluoroscopic images; and
extracting a plurality of image layers from said sequence of fluoroscopic images based on the motion estimated for said plurality of objects.
2. The method of claim 1 , wherein said step of detecting a portion of each of a plurality of objects in said sequence of fluoroscopic images comprises:
automatically detecting the portion of each of said plurality of objects using learning-based detection.
3. The method of claim 1 , wherein said step of detecting a portion of each of a plurality of objects in said sequence of fluoroscopic images comprises:
detecting the portion of said plurality of objects in each image in said sequence of fluoroscopic images.
4. The method of claim 1 , wherein said step of estimating motion of the detected portion of each of said plurality of objects comprises:
for an image in said sequence of fluoroscopic images, estimating motion of the detected portion of each of said plurality of objects between the image and another image in said sequence of fluoroscopic images.
5. The method of claim 4 , wherein said step of estimating motion of the detected portion of each of said plurality of objects between the image and another image in said sequence of fluoroscopic images comprises:
estimating motion of the detected portion of each of said plurality of objects between the image and a consecutive image in said sequence of fluoroscopic images.
6. The method of claim 4 , wherein said step of estimating motion of the detected portion of each of said plurality of objects between the image and another image in said sequence of fluoroscopic images comprises:
calculating motion vectors showing movement of the detected portion of each of said plurality of objects from the image to said another image.
7. The method of claim 6 , wherein said step of extracting a plurality of image layers from said sequence of fluoroscopic images based on the motion estimated for said plurality of objects comprises:
estimating motion fields over an image in said sequence of fluoroscopic images based on the motion vectors calculated for the detected portion of each of said plurality of objects for the image; and
extracting said plurality of image layers for the image in said sequence of fluoroscopic images using least squares estimation based on the motion fields estimated for the image.
8. The method of claim 7 , wherein said step of estimating motion fields over an image in said sequence of fluoroscopic images comprises:
computing thin-plate spline models to extrapolate the motion vectors calculated for the detected portion of each of said plurality of objects over the image.
9. The method of claim 1 , wherein said step of estimating motion of the detected portion of each of said plurality of objects in said sequence of fluoroscopic images comprises:
tracking the detected portion of each of said plurality of objects over each image in the sequence of fluoroscopic images using learning-based tracking.
10. The method of claim 1 , wherein said plurality of objects includes coronary vessels and a diaphragm.
11. The method of claim 10 , wherein one of said plurality of image layers shows said coronary vessels.
12. An apparatus for extracting motion-based layers from a sequence of fluoroscopic images, comprising:
means for detecting a portion of each of a plurality of objects in said sequence of fluoroscopic images;
means for estimating motion of the detected portion of each of said plurality objects in said sequence of fluoroscopic images; and
means for extracting a plurality of image layers from said sequence of fluoroscopic images based on the motion estimated for said plurality of objects.
13. The apparatus of claim 12 , wherein said means for detecting a portion of each of a plurality of objects in said sequence of fluoroscopic images comprises:
means for automatically detecting the portion of each of said plurality objects using learning-based detection.
14. The apparatus of claim 12 , wherein said means for detecting a portion of each of a plurality of objects in said sequence of fluoroscopic images comprises:
means for detecting the portion of said plurality of objects in each image in said sequence of fluoroscopic images.
15. The apparatus of claim 12 , wherein said means for estimating motion of the detected portion of each of said plurality of objects comprises:
means for estimating motion of the detected portion of each of said plurality of objects for a first image in said sequence of fluoroscopic images between said first image and a second image in said sequence of fluoroscopic images.
16. The apparatus of claim 15 , wherein said means for estimating motion of the detected portion of each of said plurality of objects for said first image in said sequence of fluoroscopic images comprises:
means for calculating motion vectors showing movement of the detected portion of each of said plurality of objects from said first image in said sequence of fluoroscopic images to said second image in said sequence of fluoroscopic images.
17. The apparatus of claim 16 , wherein said first and second images are consecutive images in said sequence of fluoroscopic images.
18. The apparatus of claim 16 , wherein said means for extracting a plurality of image layers from said sequence of fluoroscopic images based on the motion estimated for said plurality of objects comprises:
means for estimating motion fields over an image in said sequence of fluoroscopic images based on the motion vectors calculated for the detected portion of each of said plurality of objects for the image; and
means for extracting said plurality of image layers for the image in said sequence of fluoroscopic images using least squares estimation based on the motion fields estimated for the image.
19. The apparatus of claim 18 , wherein said means for estimating motion fields over an image in said sequence of fluoroscopic images comprises:
means for computing thin-plate spline models to extrapolate the motion vectors calculated for the detected portion of each of said plurality of objects over the image.
20. The apparatus of claim 12 , wherein said means for estimating motion of the detected portion of each of said plurality of objects in said sequence of fluoroscopic images comprises:
means for tracking the detected portion of each of said plurality of objects over each image in the sequence of fluoroscopic images using learning-based tracking.
21. A non-transitory computer readable medium encoded with computer executable instructions for extracting motion-based layers from a sequence of fluoroscopic images, the computer executable instructions defining steps comprising:
detecting a portion of each of a plurality of objects in said sequence of fluoroscopic images;
estimating motion of the detected portion of each of said plurality of objects in said sequence of fluoroscopic images; and
extracting a plurality of image layers from said sequence of fluoroscopic images based on the motion estimated for said plurality of objects.
22. The computer readable medium of claim 21 , wherein the computer executable instructions defining the step of detecting a portion of each of a plurality of objects in said sequence of fluoroscopic images comprise computer executable instructions defining the step of:
automatically detecting the portion of each of said plurality of objects using learning-based detection.
23. The computer readable medium of claim 21 , wherein the computer executable instructions defining the step of detecting a portion of each of a plurality of objects in said sequence of fluoroscopic images comprise computer executable instructions defining the step of:
detecting the portion of said plurality of objects in each image in said sequence of fluoroscopic images.
24. The computer readable medium of claim 21 , wherein the computer executable instructions defining the step of estimating motion of the detected portion of each of said plurality of objects comprise computer executable instructions defining the step of:
for an image in said sequence of fluoroscopic images, estimating motion of the detected portion of each of said plurality of objects between the image and another image in said sequence of fluoroscopic images.
25. The computer readable medium of claim 24 , wherein the computer executable instructions defining the step of estimating motion of the detected portion of each of said plurality of objects between the image and another image in said sequence of fluoroscopic images comprise computer executable instructions defining the step of:
calculating motion vectors showing movement of the detected portion of each of said plurality of objects from the image to said consecutive image.
26. The computer readable medium of claim 25 , wherein the computer executable instructions defining the step of extracting a plurality of image layers from said sequence of fluoroscopic images based on the motion estimated for said plurality of objects comprise computer executable instructions defining the steps of:
estimating motion fields over an image in said sequence of fluoroscopic images based on the motion vectors calculated for the detected portion of each of said plurality of objects for the image; and
extracting said plurality of image layers for the image in said sequence of fluoroscopic images using least squares estimation based on the motion fields estimated for the image.
27. The computer readable medium of claim 26 , wherein the computer executable instructions defining the step of estimating motion fields over an image in said sequence of fluoroscopic images comprise computer executable instructions defining the step of:
computing thin-plate spline models to extrapolate the motion vectors calculated for the detected portion of each of said plurality of objects over the image.
28. The computer readable medium of claim 21 , wherein the computer executable instructions defining the step of estimating motion of the detected portion of each of said plurality of objects in said sequence of fluoroscopic images comprise computer executable instructions defining the step of:
tracking the detected portion of each of said plurality of objects over each image in the sequence of fluoroscopic images using learning-based tracking.