Non-invasive method for using 2D angiographic images for radiosurgical target definition
A non-invasive method and system for using 2D angiographic images for radiosurgical target definition uses non-invasive calibration devices and methods to calibrate an angiographic imaging system and a six-parameter registration algorithm to register angiographic images with 3D scan data for radiation treatment planning.
1. A method, comprising:
acquiring a plurality of two-dimensional (2D) angiographic images with two or more orientations of an angiographic imaging system, each orientation having an unknown imaging geometry, wherein each of the plurality of 2D angiographic images includes a projection of a plurality of non-invasive fiducial markers having a known three-dimensional (3D) configuration;
registering, by a processing device, selected 2D angiographic images to a 3D scan volume produced by a 3D imaging system, wherein the imaging geometry of each orientation of the angiographic imaging system is unknown with respect to the 3D imaging system prior to acquiring the plurality of 2D angiographic images, wherein registering selected 2D angiographic images to the 3D scan volume comprises:
generating one or more sets of digitally reconstructed radiographs (DRRs) from the 3D scan volume using the imaging geometry of the two or more orientations of the angiographic imaging system;
comparing selected DRRs to the selected 2D angiographic images; and
finding a transformation between the 3D scan volume and the 3D object space that maximizes a similarity measure between the selected DRRs and the selected 2D angiographic images;
determining the imaging geometry of each of the two or more orientations of the angiographic imaging system from the projection of the plurality of non-invasive fiducial markers in each of the 2D angiographic images and the known 3D configuration of the plurality of non-invasive fiducial markers;
identifying contours of a target vasculature in one or more of the plurality of 2D angiographic images in each of the two or more orientations;
back-projecting the contours of the target vasculature, through the imaging geometry of the two or more orientations, to a 3D object space; and
rendering a volume of the target vasculature in the 3D object space.
2. The method of claim 1 , wherein the plurality of non-invasive fiducial markers is arrayed on a non-invasive calibration device of known configuration, wherein the calibration device has a fixed location relative to a patient's head.
3. The method of claim 1 , wherein the plurality of non-invasive fiducial markers is temporarily attached to a patient's head, the method further comprising determining the 3D configuration of the plurality of non-invasive fiducial markers by,
acquiring a calibrated 3D image of the patient, and
measuring the configuration of the plurality of non-invasive fiducial markers in the calibrated 3D image.
4. The method of claim 1 , wherein the plurality of non-invasive fiducial markers comprise tracking objects temporarily attached to a patient's head, and wherein the configuration of the tracking objects is determined by a 3D tracking system.
5. The method of claim 1 , wherein the target vasculature comprises a nidus of an arteriovenous malformation (AVM), wherein the plurality of 2D angiographic images comprises one or more time-series of angiographic images recording progress of a contrast agent from injection through infusion of the nidus, and wherein the selected 2D angiographic images include images without contrast agent and images with contrast agent.
6. The method of claim 5 , wherein the selected 2D angiographic images comprise the images without contrast agent.
7. The method of claim 5 , wherein the selected 2D angiographic images comprise the images with contrast agent.
8. The method of claim 1 , wherein the similarity measure comprises an image intensity similarity measure, and wherein finding the transformation between the 3D scan volume and the 3D object space comprises applying a six-parameter registration algorithm to maximize the similarity measure.
9. The method of claim 1 , further comprising:
fusing the 3D object space with the 3D scan volume using the transformation between the 3D scan volume and the 3D object space;
generating contours of the target vasculature from the volume of the target vasculature in the 3D object space; and
developing a radiation treatment plan.
10. A system, comprising:
a storage device; and
a processing device operatively coupled with the storage device to:
receive data comprising a plurality of two-dimensional (2D) angiographic images of an anatomical region in two or more different orientations of an angiographic imaging system, each orientation having an unknown imaging geometry, wherein each of the plurality of 2D angiographic images includes a projection of a plurality of non-invasive fiducial markers having a known three-dimensional (3D) configuration; and
register data of selected 2D angiographic images with data comprising a calibrated 3D scan volume of the anatomical region produced by the 3D imaging system, wherein the imaging geometry of each orientation of the angiographic imaging system is unknown with respect to the 3D imaging system prior to the receive of data, wherein to register data of the selected 2D angiographic images to the 3D scan volume, the processing device is to:
generate digitally reconstructed radiographs (DRRs) from the 3D scan volume using the imaging geometry of the two or more orientations of the angiographic imaging system;
compare selected DRRs to the selected 2D angiographic images; and
find a transformation between the 3D scan volume and the 3D object space that maximizes a similarity measure between the selected DRRs and the selected 2D angiographic images;
wherein the processing device is further to:
determine the imaging geometry of each of the two or more orientations of the angiographic imaging system from the projection of the plurality of non-invasive fiducial markers in each of the 2D angiographic images and the known 3D configuration of the plurality of non-invasive fiducial markers;
identify contours of a target vasculature in one or more of the plurality of 2D angiographic images in each of the two or more orientations;
back-project the contours of the target vasculature, through the imaging geometry of the two or more orientations, to a 3D object space;
render a volume of the target vasculature in the 3D object space;
identify contours of a target vasculature in the 3D scan volume;
project the contours of the target vasculature, through the imaging geometry of one or more orientations of the angiographic imaging system; and
display the projections of the target vasculature in one or more 2D angiographic images.
11. The system of claim 10 , wherein the target vasculature comprises a nidus of an arteriovenous malformation (AVM), wherein the plurality of 2D angiographic images comprises one or more time-series of angiographic images recording progress of a contrast agent from injection through infusion of the nidus, and wherein the selected 2D angiographic images include images without contrast agent and images with contrast agent.
12. The system of claim 11 , wherein the selected 2D angiographic images comprise the images without contrast agent.
13. The system of claim 11 , wherein the selected 2D angiographic images comprise the images with contrast agent.
14. The system of claim 10 , wherein the similarity measure comprises an image intensity similarity measure, and wherein to find the transformation between the 3D scan volume and the 3D object space, the processing device is configured to apply a six-parameter registration algorithm to maximize the similarity measure.
15. The system of claim 10 , wherein the processing device is further configured to:
fuse the 3D object space with the 3D scan volume using the transformation between the 3D scan volume and the 3D object space; and
generate contours of the target vasculature from the volume of the target vasculature in the 3D object space.
16. A non-transitory machine-readable medium including data that, when read by a processing device, cause the processing device to perform operations comprising:
acquiring a plurality of two-dimensional (2D) angiographic images with two or more orientations of an angiographic imaging system, each orientation having an unknown imaging geometry, wherein each of the plurality of 2D angiographic images includes a projection of a plurality of non-invasive fiducial markers having a known three-dimensional (3D) configuration;
registering selected 2D angiographic images to a 3D scan volume produced by a 3D imaging system, wherein the imaging geometry of each orientation of the angiographic imaging system is unknown with respect to the 3D imaging system prior to acquiring the plurality of 2D angiographic images, wherein registering selected 2D angiographic images to the 3D scan volume comprises:
generating one or more sets of digitally reconstructed radiographs (DRRs) from the 3D scan volume using the imaging geometry of the two or more orientations of the angiographic imaging system;
comparing selected DRRs to the selected 2D angiographic images; and
finding a transformation between the 3D scan volume and the 3D object space that maximizes a similarity measure between the selected DRRs and the selected 2D angiographic images
determining the imaging geometry of each of the two or more orientations of the angiographic imaging system from the projection of the plurality of non-invasive fiducial markers in each of the 2D angiographic images and the known 3D configuration of the plurality of non-invasive fiducial markers;
identifying contours of a target vasculature in one or more of the plurality of 2D angiographic images in each of the two or more orientations;
back-projecting the contours of the target vasculature, through the imaging geometry of the two or more orientations, to a 3D object space; and
rendering a volume of the target vasculature in the 3D object space.
17. The non-transitory machine-readable medium of claim 16 , wherein the plurality of non-invasive fiducial markers is arrayed on a non-invasive calibration device of known configuration, and wherein the calibration device has a fixed location relative to a patient's head.
18. The non-transitory machine-readable medium article of manufacture of claim 16 , wherein the plurality of non-invasive fiducial markers is temporarily attached to a patient's head, wherein the machine-readable medium further includes data that cause the processing device to perform operations comprising:
determining the 3D configuration of the plurality of non-invasive fiducial markers by,
acquiring a calibrated 3D image of the patient, and
measuring the configuration of the plurality of non-invasive fiducial markers in the calibrated 3D image.
19. The non-transitory machine-readable medium of claim 16 , wherein the plurality of non-invasive fiducial markers comprise tracking objects temporarily attached to a patient's head, and wherein the configuration of the tracking objects is determined by a 3D tracking system.
20. The non-transitory machine-readable medium of claim 16 , wherein the target vasculature comprises a nidus of an arteriovenous malformation (AVM), wherein the plurality of 2D angiographic images comprises one or more time-series of angiographic images recording progress of a contrast agent from injection through infusion of the nidus, and wherein the selected 2D angiographic images include images without contrast agent and images with contrast agent.
21. The non-transitory machine-readable medium of claim 20 , wherein the selected 2D angiographic images comprise the images without contrast agent.
22. The non-transitory machine-readable medium of claim 20 , wherein the selected 2D angiographic images comprise the images with contrast agent.
23. The non-transitory machine-readable medium of claim 16 , wherein the similarity measure comprises an image intensity similarity measure, and wherein finding the transformation between the 3D scan volume and the 3D object space comprises applying a six-parameter registration algorithm to maximize the similarity measure.
24. The non-transitory machine-readable medium of claim 16 , wherein the machine-readable medium further includes data that cause the processing device to perform operations comprising:
fusing the 3D object space with the 3D scan volume using the transformation between the 3D scan volume and the 3D object space;
generating contours of the target vasculature from the volume of the target vasculature in the 3D object space; and
developing a radiation treatment plan.