Dose deformation error calculation method and system
AssigneeKoninklijke Philips N.V.
InventorsKarl Antonin Bzdusek, Mark David Wiesmeyer
Life of this patent
Estimating the term…
Abstract
A method and system for determining a radiation dose transformation error are provided, wherein there is a deformation in one or more imaged structures as recorded by at least one fixed image data set and at least one moving image data set. Corresponding landmark points in the fixed image and in the moving image are automatically or semi-automatically identified.
Claims20
A method for determining a radiation dose transformation error, wherein there is a deformation in one or more imaged structures as recorded by at least one fixed image data set and at least one moving image data set, the method comprising:
- automatically or semi-automatically identifying corresponding landmark points in the fixed image and in the moving image;
- obtaining a fixed image radiation dose grid and a moving image radiation dose grid, wherein the moving image radiation dose grid comprises actual moving image point doses representing radiation dose levels at the landmark points in the moving image radiation dose grid;
- determining a transformation matrix between the fixed image and the moving image;
- applying the transformation matrix to the moving image data set to generate a warped radiation dose grid;
- determining warped point doses representing radiation dose levels in the warped radiation dose grid at the landmark points in the fixed image radiation dose grid;
- and
- determining difference between the actual point dose and the warped point dose at corresponding landmark points in order to generate dose errors.
A method for identifying corresponding landmark points between a fixed image data set and a moving image data set, the method comprising:
- identifying a first region of interest in one of the fixed image data set and the moving image data set;
- identifying a second region of interest in the other of the fixed image data set and the moving image data set, wherein the second region of interest corresponds to the first region of interest;
- generating a first surface mesh to approximate the first region of interest, and generating a second surface mesh to approximate the second region of interest, wherein the first and second surface meshes comprise an equal number of vertices;
- and
- identifying at least one vertex in the first region of interest and at least one vertex in the second region as corresponding landmark points.
A system for determining a radiation dose transformation error, wherein there is a deformation in one or more imaged structures as recorded by at least one fixed image data set and at least one moving image data set, the system including a computer readable medium comprising logic to:
- automatically or semi-automatically identify corresponding landmark points in the fixed image and in the moving image;
- obtain a fixed image radiation dose grid and a moving image radiation dose grid, wherein the moving image radiation dose grid comprises actual moving image point doses representing radiation dose levels at the landmark points in the moving image radiation dose grid;
- determine a transformation matrix between the fixed image and the moving image;
- apply the transformation matrix to the moving image data set to generate a warped radiation dose grid;
- determine warped point doses representing radiation dose levels in the warped radiation dose grid at the landmark points in the fixed image radiation dose grid;
- and
- determine a difference between the actual point dose and the warped point dose at corresponding landmark points in order to generate dose errors.
Dependencies are inferred from the claim text.