Back to results
US 12,718,448 B2Granted
Apparatus and method for beam-hardening correction in computed tomography
AssigneeCanon Kabushiki Kaisha
InventorsYujie Lu, Liang Cai, Zhou Yu, Jian Zhou
Life of this patent
Estimating the term…
Abstract
According to some embodiments, a method comprises obtaining a group of reconstructed-image data; converting the group of reconstructed-image data to a derivative of Radon space, thereby generating Radon-space data, wherein the Radon-space data have a radial sampling pattern in the derivative of Radon space; and generating resampled data by inputting the Radon-space data into a first trained machine-learning model for resampling Radon-space data, wherein the resampled data have a cone-beam-projection-geometry-shaped sampling pattern in the derivative of Radon space.
Claims17
A method for generating computed-tomography (CT) data, the method comprising:
- obtaining a group of reconstructed-image data;
- converting the group of reconstructed-image data to a derivative of Radon space, thereby generating Radon-space data, wherein the Radon-space data have a radial sampling pattern in the derivative of Radon space;
- generating resampled data by inputting the Radon-space data into a first trained machine-learning model for resampling Radon-space data, wherein the resampled data have a cone-beam-projection-geometry-shaped sampling pattern in the derivative of Radon space;
- generating first pathlength data based on the resampled data;
- generating second pathlength data by applying a second trained machine-learning model to the first pathlength data, wherein the second trained machine-learning model has been trained based on third pathlength data, wherein the third pathlength data were obtained by applying a pathlength-estimation process to one or more sets of reconstructed-image data, wherein the pathlength-estimation process includes ray tracing;
- and
- generating reconstructed-image data based on the second pathlength data and on projection data.
4 more elements
A medical image-processing method comprising:
- obtaining a set of projection data that was generated by scanning an object;
- generating a set of reconstructed-image data, wherein generating the set of reconstructed-image data includes performing a reconstruction process on the set of projection data;
- obtaining first pathlength data by applying a first pathlength-estimation process to a group of reconstructed-image data from the set of reconstructed-image data, wherein the first pathlength-estimation process includes a Fourier-transform process;
- generating second pathlength data by applying a first trained machine-learning model to the first pathlength data, wherein the first trained machine-learning model has been trained based on third pathlength data, wherein the third pathlength data were obtained by applying a second pathlength-estimation process to one or more sets of reconstructed-image data, and wherein the second pathlength-estimation process includes ray tracing;
- obtaining corrected projection data by applying a beam-hardening-correction process to the set of projection data based on the second pathlength data;
- and
- generating corrected reconstructed-image data based on the corrected projection data.
4 more elements
An image-processing apparatus comprising:
- one or more computer-readable storage media storing instructions;
- and
- one or more processors, wherein the one or more processors and the one or more computer-readable storage media are configured to control the apparatus to perform operations that include:
- obtaining a set of projection data that was generated by scanning an object;
- generating a set of reconstructed-image data, wherein generating the set of reconstructed-image data includes performing a reconstruction process on the set of projection data;
- obtaining first pathlength data by applying a first pathlength-estimation process to a group of reconstructed-image data from the set of reconstructed-image data, wherein the first pathlength-estimation process includes a Fourier-transform process;
- generating second pathlength data by applying a first trained machine-learning model to the first pathlength data, wherein the first trained machine-learning model has been trained based on third pathlength data, wherein the third pathlength data were obtained by applying a second pathlength-estimation process to one or more sets of reconstructed-image data, and wherein the second pathlength-estimation process includes ray tracing;
- obtaining corrected projection data by applying a beam-hardening-correction process to the set of projection data based on the second pathlength data;
- and
- generating corrected reconstructed-image data based on the corrected projection data.
7 more elements
A medical-image-processing method, the method comprising:
- obtaining a set of projection data, wherein the set of projection data was generated by scanning an object;
- generating a set of reconstructed-image data, wherein generating the set of reconstructed-image data includes performing a reconstruction process on the set of projection data;
- obtaining first pathlength data corresponding to a specific material by applying a first pathlength-estimation process on a first group of reconstructed-image data from the set of reconstructed-image data, wherein the first pathlength-estimation process includes a Fourier transform;
- obtaining second pathlength data corresponding to the specific material by applying a first trained machine-learning model to the first pathlength data, wherein the first trained machine-learning model has been trained based on third pathlength data, wherein the third pathlength data were obtained by applying a second pathlength-estimation process to one or more sets of reconstructed-image data, wherein the second pathlength-estimation process includes ray tracing;
- generating corrected projection data, wherein generating the corrected projection data includes applying a beam-hardening-correction process to the set of projection data based on the second pathlength data;
- and
- generating corrected reconstructed-image data based on the corrected projection data.
4 more elements
Dependencies are inferred from the claim text.