Systems, methods and devices for automated target volume generation
Systems and method for automatically generating structures, such as target volumes, in a treatment image using structure-guided deformation to propagate the structures from a planning image onto the subsequently acquired treatment image.
1 . A method for generating deformation vector fields (DVFs) that allow for automatic propagation of contours of structures from a first image that includes a first set of structures and a second set of structures to a second image that includes the second set of structures, comprising:
using image data of the first image and image data of the second image as input in a deformable registration algorithm; and
computing the deformable registration algorithm through a plurality of computation steps that optimize similarity measures between the first image and the second image,
wherein the plurality of computation steps includes:
comparing and spatially registering the first image with the second image to obtain a plurality of vectors that map voxels of each prescribed location in the first image to a location in the second image;
aggregating the plurality of vectors into a first deformation map;
constraining the first deformation map to force points in the second set of structures of the first image to match with points in the second set of structures of the second image; and
mapping of the image data of the first image to the image data of the second image to obtain the deformation vector fields (DVFs).
2 . The method for generating deformation vector fields (DVFs) of claim 1 , wherein the deformable registration algorithm is a structure-guided deformable registration algorithm.
3 . The method for generating deformation vector fields (DVFs) of claim 2 , wherein the constraining includes implementing a constraint in the structure-guided deformable registration algorithm to force intensity matching between the second set of structures of the first image and the second set of structures of the second image.
4 . The method for generating deformation vector fields (DVFs) of claim 3 , wherein the constraining further includes implementing a constraint in the structure-guided deformable registration algorithm to force a first set of points in the first set of structures to deform rigidly and a second set of points in the first set of structures to deform non-rigidly.
5 . The method for generating deformation vector fields (DVFs) of claim 4 , wherein the first set of points includes points that do not move independently of the second set of structures, and the second set of points include points that move independently of the second set of structures.
6 . The method for generating deformation vector fields (DVFs) of claim 1 , wherein the first set of structures include one or more target volumes and one or more anatomical structures of interest, and the second set of structures include one or more anatomical influencer structures.
7 . The method for generating deformation vector fields (DVFs) of claim 6 , wherein the anatomical influencer structures include anatomical structures that influence one of a shape, size, or location of the one or more target volumes.
8 . A non-transitory computer-readable storage medium upon which is embodied a sequence of programmed instructions for the generation of treatment images to be used in adaptive radiation therapy, which when executed by a computer processing system cause the computer processing system to:
generate deformation vector fields between a planning image and a treatment image to be used to propagate structures from the planning image to the treatment image; and
apply the generated deformation vector fields to propagate structures from the planning image to the treatment image,
wherein the deformation vector fields take into consideration the motion and deformation properties of the propagated structures, and
wherein the generation of the deformation vector fields includes:
using image data of the planning image and image data of the treatment image as input in a deformable registration algorithm; and
computing the deformable registration algorithm through a plurality of computation steps that optimize similarity measures between the planning image and the treatment image,
wherein the planning image includes a first set of structures and a second set of structures, and the treatment image includes the second set of structures, and the plurality of computation steps includes:
comparing and spatially registering the planning image with the treatment image to obtain a plurality of vectors that map voxels of each prescribed location in the planning image to a location in the treatment image;
aggregating the plurality of vectors into a first deformation map;
constraining the first deformation map to force points in the second set of structures of the planning image to match with points in the second set of structures of the treatment image; and
mapping of the image data of the planning image to the image data of the treatment image to obtain the deformation vector fields (DVFs).
9 . The non-transitory computer-readable storage medium of claim 8 , wherein the deformable registration algorithm is a structure-guided deformable registration algorithm.
10 . The non-transitory computer-readable storage medium of claim 9 , wherein the constraining includes implementing a constraint in the structure-guided deformable registration algorithm to force intensity matching between the second set of structures of the planning image and the second set of structures of the treatment image.
11 . The non-transitory computer-readable storage medium of claim 10 , wherein the constraining further includes implementing a constraint in the structure-guided deformable registration algorithm to force a first set of points in the first set of structures to deform rigidly and a second set of points in the first set of structures to deform non-rigidly.
12 . The non-transitory computer-readable storage medium of claim 11 , wherein the first set of points includes points that do not move independently of the second set of structures, and the second set of points include points that move independently of the second set of structures.
13 . The non-transitory computer-readable storage medium of claim 8 , wherein the first set of structures include one or more target volumes and one or more anatomical structures of interest, and the second set of structures include one or more anatomical influencer structures, the anatomical influencer structures including anatomical structures that influence one of a shape, size, or location of the one or more target volumes.