X-ray system and method for standing subject
A system includes: a movable X-ray tube scanner; a range sensor movable with the X-ray tube scanner; an X-ray detector positioned to detect X-rays from the X-ray tube passing through a standing subject between the X-ray tube and the X-ray detector; and a processor configured for automatically controlling the X-ray tube scanner to transmit X-rays to a region of interest of the patient while the subject is standing between the X-ray tube and the X-ray detector.
1. A method comprising:
receiving range data of a subject from a range sensor, where the range sensor is movable with an X-ray tube, and the subject is standing between the X-ray tube and an X-ray detector;
computing a surface model of the subject from the 3D depth image using a parametric model, wherein at least one parameter of the parametric model is estimated by machine learning;
selecting the region of interest of the subject based on the surface model; and
automatically controlling the X-ray tube to transmit X-ray radiation to a region of interest of the subject while the subject is standing between the X-ray tube and the X-ray detector.
2. The method of claim 1 , wherein the range sensor includes a three-dimensional (3D) camera and the range data includes a 3D depth image.
3. A method comprising:
receiving range data of a subject from a range sensor, where the range sensor is movable with an X-ray tube, and the subject is standing between the X-ray tube and an X-ray detector;
computing a surface model of the subject from the 3D depth image using a parametric model, wherein at least one parameter of the parametric model is estimated by machine learning;
selecting the region of interest of the subject based on the surface model; and
automatically controlling the X-ray tube to transmit X-ray radiation to a region of interest of the subject while the subject is standing between the X-ray tube and the X-ray detector;
wherein computing the surface model from the 3D depth image of the subject comprises:
transforming a 3D point cloud representation of the 3D depth image to align the 3D depth image with a predetermined field of view of the X-ray detector using an estimated transformation between a coordinate system of the 3D camera and a coordinate system of the X-ray detector;
projecting the 3D point cloud representation to generate a re-projected image comprising a color and depth image pair; and
estimating the surface model using the re-projected image.
4. A method comprising:
receiving range data of a subject from a range sensor, where the range sensor is movable with an X-ray tube, and the subject is standing between the X-ray tube and an X-ray detector;
computing a surface model of the subject from the 3D depth image using a parametric model, wherein at least one parameter of the parametric model is estimated by machine learning;
selecting the region of interest of the subject based on the surface model; and
automatically controlling the X-ray tube to transmit X-ray radiation to a region of interest of the subject while the subject is standing between the X-ray tube and the X-ray detector;
wherein computing the surface model from the 3D depth image of the subject comprises:
detecting a body pose in the 3D depth image using one or more machine learning-based pose classifiers; and
detecting anatomical landmarks of the subject in the 3D data based on the detected patient pose.
5. The method of claim 2 , further comprising computing a collimation of the X-ray radiation based on the parametric model.
6. The method of claim 2 , further comprising selecting a dose of the X-ray radiation based on the parametric model.
7. The method of claim 2 , further comprising computing a transformation between a coordinate system of the 3D camera and a coordinate system of an X-ray detector.
8. The method of claim 2 , wherein the 3D camera is mounted to the X-ray tube.
9. The method of claim 1 , wherein the X-ray detector is arranged parallel to a coronal plane of the subject during the step of automatically controlling the X-ray tube.
10. A system comprising:
a movable X-ray tube scanner;
a range sensor movable with the X-ray tube scanner;
an X-ray detector positioned to detect X-rays from the X-ray tube passing through a standing subject between the X-ray tube and the X-ray detector;
a processor configured for automatically controlling the X-ray tube scanner to transmit X-rays to a region of interest of the patient while the subject is standing between the X-ray tube and the X-ray detector; and
a non-transitory, machine-readable storage medium storing a parametric deformable model, wherein the processor is also configured for estimating a surface model of the subject based on the parametric deformable model and a 3D depth image of the subject captured by the range sensor while the subject is standing between the X-ray tube and the X-ray detector, wherein at least one parameter of the parametric deformable model is estimated by machine learning.
11. The system of claim 10 , wherein the range sensor is a 3D camera.
12. The system of claim 11 , wherein the 3D camera is a 3D depth camera.
13. The system of claim 10 , wherein the X-ray detector is mounted on a stand above a floor and is oriented to receive X-rays traveling in a horizontal direction.
14. The system of claim 13 , wherein the X-ray detector is mechanically movable in a vertical direction relative to the stand.
15. The system of claim 14 , wherein the processor is configured to control the X-ray tube scanner and the X-ray detector to move so that the X-ray tube scanner, the region of interest, and the X-ray detector are aligned along a line.
16. The system of claim 10 , wherein the X-ray tube scanner is mechanically mounted for translating in three orthogonal dimensions and rotating in at least two orthogonal dimensions.