Beam angle optimization for external beam radiation therapy using sectioning
Methods of beam angle optimization for intensity modulated radiotherapy (IMRT) treatment include determining beam's eye view (BEV) regions and a BEV region connectivity manifold by evaluating dose response of each region of interest for each vertex in a delivery coordinate space (DCS). The information contained in the BEV regions and the BEV region connectivity manifold is used to guide an optimizer to find optimal field geometries in the IMRT treatment. To improve the visibility of insufficiently exposed voxels of planning target volumes (PTVs), a post-processing step may be performed to enlarge certain BEV regions, which are considered for exposing during treatment trajectory optimization.
1. A method of beam angle optimization for an IMRT radiotherapy treatment, the method comprising:
providing a patient model including one or more planning target volumes (PTVs), the one or more PTVs including a first PTV;
defining a delivery coordinate space (DCS) having a plurality of vertices, each respective vertex defining a respective beam's eye view (BEV) plane;
for each respective BEV plane of a respective vertex, evaluating a respective BEV score of each respective pixel of the respective BEV plane for the first PTV based at least in part on a dose of the first PTV from a respective beamlet originating from the respective vertex and passing through the respective pixel of the respective BEV plane;
determining a threshold BEV score for the first PTV based on the BEV scores of the pixels of the BEV planes;
for each respective BEV plane of the respective vertex, determining one or more BEV regions by comparing the respective BEV score of each respective pixel to the threshold BEV score;
determining a BEV region connectivity manifold for the first PTV representing connections between BEV regions of adjacent vertices; and
selecting a plurality of optimal field geometries for the IMRT radiotherapy treatment based at least in part on the BEV region connectivity manifold for the first PTV.
2. The method of claim 1 , wherein pixels within the one or more BEV regions have BEV scores greater than or equal to the threshold BEV score.
3. The method of claim 1 , wherein each optimal field geometry of the plurality of optimal field geometries comprises at least one of a gantry angle, a couch angle, or a collimator angle.
4. The method of claim 1 , wherein selecting the plurality of optimal field geometries comprises performing an optimization on a graph defined by a plurality of nodes, each node associated with a respective vertex in the DCS.
5. The method of claim 4 , further comprising:
for each respective BEV plane of a respective vertex:
determining a region score for each respective pixel of the respective BEV plane based on the respective BEV score of the respective pixel; and
determining an integrated region score for each respective BEV region of the one or more BEV regions based on the region scores of the pixels of the respective BEV region;
wherein the optimization is performed using a max-distance function including a first term relating to the integrated region score for each respective BEV region.
6. The method of claim 5 , wherein the region score for each respective pixel has a normalized positive value if the respective pixel is within one of the one or more BEV regions, and a negative value if the respective pixel is not within any of the one or more BEV regions.
7. The method of claim 5 , wherein the max-distance function includes a second term relating to a set of PTV angular fluxes for a set of sampling points distributed within the first PTV, each PTV angular flux relating to novelty of directional vectors of incident beamlets through a closed surface centered at a respective sampling point of the set of sampling points.
8. The method of claim 5 , wherein the optimization is performed using a local gradient descent algorithm.
9. The method of claim 1 , wherein the patient model further includes a first organs at risk (OAR), the method further comprising:
for each respective BEV plane of a respective vertex:
for each respective pixel of the respective BEV plane:
evaluating a dose of the first OAR from a respective beamlet originating from the respective vertex and passing through the respective pixel of the respective BEV plane; and
wherein evaluating the respective BEV score of the respective pixel of the respective BEV plane comprises evaluating a weighted linear combination of the dose of the first PTV and the dose of the first OAR.
10. The method of claim 9 , wherein the first PTV is given a positive weight, and the first OAR is given a negative weight in the weighted linear combination.
11. A computer product comprising a non-transitory computer readable medium storing a plurality of instructions that when executed control a computer system to perform beam angle optimization for an IMRT radiotherapy treatment, the instructions comprising:
providing a patient model including one or more planning target volumes (PTVs), the one or more PTVs including a first PTV;
defining a delivery coordinate space (DCS) having a plurality of vertices, each respective vertex defining a respective beam's eye view (BEV) plane;
for each respective BEV plane of a respective vertex, evaluating a respective BEV score of each respective pixel of the respective BEV plane for the first PTV based at least in part on a dose of the first PTV from a respective beamlet originating from the respective vertex and passing through the respective pixel of the respective BEV plane;
determining a threshold BEV score for the first PTV based on the BEV scores of the pixels of the BEV planes;
for each respective BEV plane of the respective vertex, determining one or more BEV regions by comparing the respective BEV score of each respective pixel to the threshold BEV score;
determining a BEV region connectivity manifold for the first PTV representing connections between BEV regions of adjacent vertices; and
selecting a plurality of optimal field geometries for the IMRT radiotherapy treatment based at least in part on the BEV region connectivity manifold for the first PTV.
12. The computer product of claim 11 , wherein pixels within the one or more BEV regions have BEV scores greater than or equal to the threshold BEV score.
13. The computer product of claim 11 , wherein each optimal field geometry of the plurality of optimal field geometries comprises at least one of a gantry angle, a couch angle, or a collimator angle.
14. The computer product of claim 11 , wherein selecting the plurality of optimal field geometries comprises performing an optimization on a graph defined by a plurality of nodes, each node associated with a respective vertex in the DCS.
15. The computer product of claim 14 , wherein the instructions further comprise:
for each respective BEV plane of a respective vertex:
determining a region score for each respective pixel of the respective BEV plane based on the respective BEV score of the respective pixel; and
determining an integrated region score for each respective BEV region of the one or more BEV regions based on the region scores of the pixels of the respective BEV region;
wherein the optimization is performed using a max-distance function including a first term relating to the integrated region score for each respective BEV region.
16. The computer product of claim 15 , wherein the region score for each respective pixel has a normalized positive value if the respective pixel is within one of the one or more BEV regions, and a negative value if the respective pixel is not within any of the one or more BEV regions.
17. The computer product of claim 15 , wherein the max-distance function includes a second term relating to a set of PTV angular fluxes for a set of sampling points distributed within the first PTV, each PTV angular flux relating to novelty of directional vectors of incident beamlets through a closed surface centered at a respective sampling point of the set of sampling points.
18. The computer product of claim 15 , wherein the optimization is performed using a local gradient descent algorithm.
19. The computer product of claim 11 , wherein the patient model further includes a first organs at risk (OAR), and wherein the instructions further comprise:
for each respective BEV plane of a respective vertex:
for each respective pixel of the respective BEV plane:
evaluating a dose of the first OAR from a respective beamlet originating from the respective vertex and passing through the respective pixel of the respective BEV plane; and
wherein evaluating the respective BEV score of the respective pixel of the respective BEV plane comprises evaluating a weighted linear combination of the dose of the first PTV and the dose of the first OAR.
20. The computer product of claim 19 , wherein the first PTV is given a positive weight, and the first OAR is given a negative weight in the weighted linear combination.