Partial deformation maps for reconstructing motion-affected treatment dose
A method comprises identifying a treatment planning image of a target subject, the treatment planning image comprising information associated with an arrangement of structures within the target subject. The method further comprises generating, based on the information, a set of reference data associated with the target subject, the reference data indicating a plurality of positions of the target subject. The method further comprises generating target-subject-specific models based on the reference data and modifying one or more hyper-parameters of the target-subject-specific mode to generate second target-subject-specific models corresponding to a second position of the plurality of positions. The method further comprises controlling a radiation treatment delivery device based on the second target-subject-specific model to deliver a radiation treatment to the target subject.
1. A method, comprising:
identifying a treatment planning image of a target subject, the treatment planning image comprising information associated with an arrangement of structures within the target subject;
generating, based on the information, a set of reference data associated with the target subject, the reference data indicating a plurality of positions of the target subject;
generating a first target-subject-specific model based on the reference data;
applying, by a processing device, the first target-subject-specific model to the treatment planning image to generate a transformed treatment planning image corresponding to a first position of the plurality of positions;
comparing the transformed treatment planning image to a reference image;
based on the comparing, modifying one or more hyper-parameters of the first target-subject-specific model to generate a second target-subject-specific model corresponding to a second position of the plurality of positions; and
controlling a radiation treatment delivery device based on the second target-subject-specific model to deliver a radiation treatment to the target subject.
2. The method of claim 1 , wherein the treatment planning image is one of three-dimensional (3-D) or four-dimensional (4-D) anatomical images and wherein the transformed treatment planning image and the reference image are two-dimensional (2-D) images.
3. The method of claim 1 , further comprising constructing the first target-subject-specific model from a 3-D, 4-D, or other derived image of the target subject.
4. The method of claim 1 , further comprising constructing the first target subject-specific model from at least one deformable registration between the target subject and another subject.
5. The method of claim 1 , further comprising constructing the first target-subject-specific model from at least one deformable registration between the target subject at a first time and the target subject at a second time.
6. The method of claim 1 , further comprising constructing the first target subject-specific model from at least one deformable registration between the target subject and a synthesized atlas subject.
7. The method of claim 1 , wherein the subject-specific models comprise a central atlas of the target subject and a set of basis transformations and appearance changes, and wherein the method further comprises linearly combining the central atlas, the set of basis transformations, and the appearance changes to generate an estimate of one or more conformations in which the target subject may be found.
8. The method of claim 1 , wherein the subject-specific models are developed using dimensionality reduction on a set of at least one of: real or synthetic data.
9. The method of claim 1 , wherein the information and one or more parameters of the target-subject-specific models are used to construct a fully 4-D spatiotemporal correspondence mapping between an anatomy of the target subject at a reference time and the anatomy of the target subject at time of image acquisition.
10. The method of claim 1 , wherein the information and one or more parameters of the target-subject-specific models are used to construct a fully 4-D spatiotemporal appearance model of the target subject at a time of image acquisition.
11. The method of claim 1 , further comprising receiving an auxiliary anatomical or pseudo-anatomical signal simultaneously with receiving the information.
12. The method of claim 11 , wherein the auxiliary signal is correlated with the first target-subject-specific model and the information to increase a temporal resolution of the second target-subject-specific model.
13. The method of claim 11 , wherein the auxiliary signal is obtained optically based on a set of triangulated light-emitting diode (LED) markers.
14. A radiation treatment delivery system comprising:
a memory to store information; and
a processing device, operatively coupled to the memory, the processing device to:
identify a treatment planning image of a target subject, the treatment planning image comprising the information associated with an arrangement of structures within the target subject;
generate, based on the information, a set of reference data associated with the target subject, the reference data indicating a plurality of positions of the target subject;
generate a first target-subject-specific model based on the reference data;
apply the first target-subject-specific model to the treatment planning image to generate a transformed treatment planning image corresponding to a first position of the plurality of positions;
compare the transformed treatment planning image to a reference image;
based on the comparing, modify one or more hyper-parameters of the first target-subject-specific model to generate a second target-subject-specific model corresponding to a second position of the plurality of positions; and
control a radiation treatment delivery device based on the second target-subject-specific model to deliver a radiation treatment to the target subject.
15. The radiation treatment delivery system of claim 14 , wherein the treatment planning image is one of three-dimensional (3-D) or four-dimensional (4-D) anatomical images and wherein the transformed treatment planning image and the reference image are two-dimensional (2-D) images.
16. The radiation treatment delivery system of claim 14 , the processing device further to construct the first target-subject-specific model from a 3-D, 4-D, or other derived image of the target subject.
17. The radiation treatment delivery system of claim 14 , the processing device further to construct the first target subject-specific model from at least one deformable registration between the target subject and another subject.
18. The radiation treatment delivery system of claim 14 , the processing device further to construct the first target-subject-specific model from at least one deformable registration between the target subject at a first time and the target subject at a second time.
19. The radiation treatment delivery system of claim 14 , the processing device further to construct the first target subject-specific model from at least one deformable registration between the target subject and a synthesized atlas subject.
20. The radiation treatment delivery system of claim 14 , wherein the subject-specific models comprise a central atlas of the target subject and a set of basis transformations and appearance changes, and wherein the processing device is further to linearly combine the central atlas, the set of basis transformations, and the appearance changes to generate an estimate of one or more conformations in which the target subject may be found.
21. The radiation treatment delivery system of claim 14 , wherein the subject-specific models are developed using dimensionality reduction on a set of at least one of: real or synthetic data.
22. The radiation treatment delivery system of claim 14 , wherein the information and one or more parameters of the target-subject-specific models are used to construct a fully 4-D spatiotemporal correspondence mapping between an anatomy of the target subject at a reference time and the anatomy of the target subject at time of image acquisition.
23. The radiation treatment delivery system of claim 14 , wherein the information and one or more parameters of the target-subject-specific models are used to construct a fully 4-D spatiotemporal appearance model of the target subject at a time of image acquisition.
24. The radiation treatment delivery system of claim 14 , the processing device further to receive an auxiliary anatomical or pseudo-anatomical signal simultaneously with receiving the information.
25. The radiation treatment delivery system of claim 24 , wherein the auxiliary signal is correlated with the first target-subject-specific model and the information to increase a temporal resolution of the second target-subject-specific model.
26. The radiation treatment delivery system of claim 24 , wherein the auxiliary signal is obtained optically based on a set of triangulated light-emitting diode (LED) markers.
27. The radiation delivery system of claim 14 , further comprising:
a gantry coupled to the radiation source, wherein the gantry is configured to rotate the radiation source about the target subject.
28. The radiation delivery system of claim 27 , wherein the gantry comprises a C-arm gantry.
29. The radiation delivery system of claim 27 , wherein the gantry comprises a ring gantry.
30. The radiation delivery system of claim 14 , further comprising:
a robotic arm coupled to the radiation source, wherein the robotic arm is configured to position the radiation source at a plurality of positions along a circular or elliptical trajectory.
31. The radiation delivery system of claim 30 , wherein the robotic arm positions the radiation source at the plurality of positions about the target subject.
32. The radiation delivery system of claim 14 , wherein the radiation beam comprises a kilovolt (kV) treatment beam.
33. The radiation delivery system of claim 14 , wherein the MLC comprises a binary MLC.
34. A non-transitory, computer-readable storage medium comprising instructions, which when executed by a processing device, cause the processing device to:
identify a treatment planning image of a target subject, the treatment planning image comprising information associated with an arrangement of structures within the target subject;
generate, based on the information, a set of reference data associated with the target subject, the reference data indicating a plurality of positions of the target subject;
generate a first target-subject-specific model based on the reference data;
apply, by the processing device, the first target-subject-specific model to the treatment planning image to generate a transformed treatment planning image corresponding to a first position of the plurality of positions;
compare the transformed treatment planning image to a reference image;
based on the comparing, modify one or more hyper-parameters of the first target-subject-specific model to generate a second target-subject-specific model corresponding to a second position of the plurality of positions; and
control a radiation treatment delivery device based on the second target-subject-specific model to deliver a radiation treatment to the target subject.
35. The non-transitory, computer-readable storage medium of claim 34 , wherein the treatment planning image is one of three-dimensional (3-D) or four-dimensional (4-D) anatomical images and wherein the transformed treatment planning image and the reference image are two-dimensional (2-D) images.
36. The non-transitory, computer-readable storage medium of claim 34 , the processing device further to construct the first target-subject-specific model from a 3-D, 4-D, or other derived image of the target subject.
37. The non-transitory, computer-readable storage medium of claim 34 , the processing device further to construct the first target subject-specific model from at least one deformable registration between the target subject and another subject.
38. The non-transitory, computer-readable storage medium of claim 34 , the processing device further to construct the first target-subject-specific model from at least one deformable registration between the target subject at a first time and the target subject at a second time.
39. The non-transitory, computer-readable storage medium of claim 34 , the processing device further to construct the first target subject-specific model from at least one deformable registration between the target subject and a synthesized atlas subject.
40. The non-transitory, computer-readable storage medium of claim 34 , wherein the subject-specific models comprise a central atlas of the target subject and a set of basis transformations and appearance changes, and wherein the processing device is further to linearly combine the central atlas, the set of basis transformations, and the appearance changes to generate an estimate of one or more conformations in which the target subject may be found.
41. The non-transitory, computer-readable storage medium of claim 34 , wherein the subject-specific models are developed using dimensionality reduction on a set of at least one of: real or synthetic data.
42. The non-transitory, computer-readable storage medium of claim 34 , wherein the information and one or more parameters of the target-subject-specific models are used to construct a fully 4-D spatiotemporal correspondence mapping between an anatomy of the target subject at a reference time and the anatomy of the target subject at time of image acquisition.
43. The non-transitory, computer-readable storage medium of claim 34 , wherein the information and one or more parameters of the target-subject-specific models are used to construct a fully 4-D spatiotemporal appearance model of the target subject at a time of image acquisition.
44. The non-transitory, computer-readable storage medium of claim 34 , the processing device further to receive an auxiliary anatomical or pseudo-anatomical signal simultaneously with receiving the information.
45. The non-transitory, computer-readable storage medium of claim 44 , wherein the auxiliary signal is correlated with the first target-subject-specific model and the information to increase a temporal resolution of the second target-subject-specific model.
46. The non-transitory, computer-readable storage medium of claim 44 , wherein the auxiliary signal is obtained optically based on a set of triangulated light-emitting diode (LED) markers.