Bias field estimation for intensity inhomogeneity correction in MR images
AssigneeSiemens Corporate Research, Inc.
InventorsShang-Hong Lai, Ming Fang
Life of this patent
Estimating the term…
Abstract
An intensity inhomogeneity correction system includes a log operator which performs a log operation on the intensity values of the original MR image. A regular grid generator partitions the image domain and then a compute and select establishes reliable orientation constraints at grid locations. A bias field surface reconstructor reconstructs a bias field surface from selected orientation constraints with a thin-plate spline by using a preconditioned conjugate gradient. An intensity inhomogeneities remover subtracts an estimated bias function from the original log image and an exponential operation on the bias-field corrected log image provides a corrected image.
Claims20
A variational shape-from-orientation apparatus for correcting intensity inhomogeneities in MR images comprising:
- a log operator for receiving an original MR image;
- a grid operator connected to said log operator;
- a compute and select connected to said grid generator;
- a bias field surface reconstructor connected to said compute and select and adapt to reconstruct a bias field surface without a prior knowledge of any tissue types depicted in the original MR image and without segmentation of the original MR image with respect to the tissue types;
- an intensity inhomogeneities remover connected to said bias field surface reconstructor and said log operator;
- and
- an exponential operator connected to said intensity inhomogeneities remover for providing a corrected image.
A variational shape-from-orientation method for correcting intensity inhomogeneities in MR images comprising the steps of:
- taking a log operation on intensity values of an original MR image;
- partitioning an image domain using a grid;
- computing and selecting reliable orientation constraints at grid locations;
- reconstructing a bias field surface from selected orientation constraints with a thin-plate spline by using a preconditioned conjugate gradient;
- removing intensity inhomogeneities by subtracting an estimated bias field function from an original log image;
- and
- taking an exponential operation on a bias-field corrected log image, wherein said reconstructing step reconstructs the bias field surface without a prior knowledge of any tissue types depicted in the original MR image and without segmentation of the original MR image with respect to the tissue types.
A variational shape-from-orientation apparatus for correcting intensity inhomogeneities in MR image comprising:
- a log operator for receiving an original MR image;
- a grid operator connected to said log operator;
- a compute and select connected to said grid generator;
- a bias field surface reconstructor connected to said compute and select and adapted to reconstruct a bias field surface without a prior knowledge of any tissue types depicted in the original MR image and without segmentation of the original MR image with respect to the tissue types;
- an intensity inhomogeneities remover connected to said bias field surface reconstructor and said log operator;
- an exponential operator connected to said intensity inhomogeneities remover for providing a corrected image;
- wherein said bias field surface reconstructor comprises:
- a thin-plate spline surface constructor for providing a solution of a linear system which is obtained by an efficient preconditioned conjugate gradient algorithm.
Dependencies are inferred from the claim text.