Method and systems for automatic gantry tilt estimation for head CT scans from camera images

1 . A method for estimating an optimal gantry orientation for a CT scan using a CT scanning protocol without using a scout scan, the method comprising:

acquiring, by at least one camera, one or more images of a patient;

estimating, using an image to image network, 3D internal landmarks of the patient from the one or more images, wherein the image to image network is trained to convert the one or more images into a segmented image with identified internal landmarks;

defining, based on the estimated 3D internal landmarks and the CT scanning protocol, one or more protocol dependent 3D orientation boxes, each orientation box comprising a geometric volume defined in a coordinate system of the CT scanner and representing at least one of a scan region or an avoidance region for the CT scanning protocol;

determining the optimal gantry orientation by evaluating an orientation of a CT scan plane relative to the one or more protocol dependent 3D orientation boxes to satisfy protocol-specific scan constraints without acquisition of a scout scan; and

performing the CT scan with the CT scanning protocol, by the CT scanner, wherein a gantry of the CT scanner is positioned using the optimal gantry orientation.

2 . The method of claim 1 , further comprising:

determining an optimal scan range based on the protocol dependent 3D orientation boxes, wherein the CT scan is performed using the optimal scan range.

3 . The method of claim 1 , wherein the at least one camera comprises an RGBD camera.

4 . The method of claim 1 , wherein the at least one camera is located inside a bore of the gantry.

5 . The method of claim 1 , wherein the optimal gantry orientation comprises tilting the gantry in order to keep the patient's eye lenses outside a range of a beam during the CT scan.

6 . The method of claim 1 , wherein the optimal gantry orientation comprises moving the gantry horizontally in order to keep the patient's eye lenses outside a range of a beam during the CT scan.

7 . The method of claim 1 , wherein the steps of acquiring, estimating, defining, and determining are repeated continuously while the patient is in the gantry.

8 . A method for estimating gantry orientation of a gantry for a CT scan without using a scout scan, the method comprising:

acquiring, by at least one camera, one or more images of a patient;

defining a CT scanning protocol for the CT scan by a CT scanner;

defining using a machine learning system, the one or more images, and the CT scanning protocol, one or more protocol dependent 3D orientation boxes, each protocol dependent 3D orientation box comprising a geometric volume defined in a coordinate system of the CT scanner and representing at least one of a scan region or an avoidance region for the CT scanning protocol;

determining an optimal gantry orientation directly from the one or more protocol dependent 3D orientation boxes; and

performing the CT scan, wherein the gantry is positioned using the optimal gantry orientation.

9 . The method of claim 8 , wherein the at least one camera comprises an RGBD camera.

10 . The method of claim 8 , wherein the at least one camera is located inside a bore of the gantry.

11 . The method of claim 8 , wherein the gantry orientation comprises tilting or adjusting a position of the gantry in order to keep the patient's eye lenses outside a range of a beam during the CT scan.

12 . The method of claim 8 , wherein the steps of acquiring and determining are repeated continuously while the patient is in the gantry.

13 . The method of claim 8 , further comprising:

determining, using the machine learning system, one or more scanning parameters from the one or more images; wherein the CT scan is performed with the one or more scanning parameters.

14 . The method of claim 8 , wherein the one or more images cover at least a head region of the patient.

15 . A system for estimating gantry orientation for CT scans without using a topogram, the system comprising:

at least one camera configured to acquire one or more images of a patient;

a CT scanning system configured to perform a scan of the patient, the CT scanning system comprising at least a gantry and a gantry control system; and

a control unit configured to determine an optimal gantry orientation based on the one or more images using at least an image to image network trained to convert the one or more images into a segmented image with identified three-dimensional internal landmarks that are used to define one or more protocol dependent 3D orientation boxes, each orientation box comprising a geometric volume defined in a coordinate system of the CT scanner and representing at least one of a scan region or an avoidance region for the CT scanning protocol, determine an optimal gantry orientation by evaluating an orientation of a CT scan plane relative to the one or more protocol dependent 3D orientation boxes to satisfy protocol-specific scan constraints without acquisition of a scout scan,

and adjust the gantry using the gantry control system.

16 . The system of claim 15 , wherein the at least one camera comprises an RGBD camera.

17 . The system of claim 15 , wherein the at least one camera is located inside a bore of the gantry.

18 . The system of claim 15 , wherein the optimal gantry orientation comprises tilting or adjusting a position of the gantry in order to keep the patient's eye lenses outside a range of a beam during the scan.