Apparatus and method for deep-learning-based scatter estimation and correction

1 . A method for performing X-ray scatter correction for computed tomography (CT) images, the method comprising:

obtaining a set of projection data acquired from a CT scan of an object;

generating, based on the set of projection data, one or more sets of preliminary scattering data, wherein the one or more sets of preliminary scattering data are different from the set of projection data; and

inputting the obtained set of projection data and the generated one or more sets of preliminary scattering data into a trained machine-learning model, wherein, based on the set of projection data and on the one or more sets of preliminary scattering data, the trained machine-learning model outputs scatter-distribution data.

2 . The method of claim 1 , wherein generating the one or more sets of preliminary scattering data includes:

generating a reconstructed image of the object based on the set of projection data, and

wherein generating the one or more sets of preliminary scattering data is based on the reconstructed image of the object.

3 . The method of claim 1 ,

wherein the set of projection data includes subsets of projection data that were acquired at a plurality of detection angles,

wherein at least some of the plurality of detection angles are different from other detection angles of the plurality of detection angles,

wherein each subset of the subsets of projection data was acquired at a respective detection angle of the plurality of detection angles,

wherein the respective detection angle of each subset of the subsets of projection data is different from the respective detection angle of at least one other subset of the subsets of projection data, and

wherein the one or more sets of preliminary scattering data include a respective set of preliminary scattering data for each detection angle.

4 . The method of claim 3 , wherein the trained machine-learning model outputs respective scatter-distribution data for each subset of the subsets of projection data.

5 . The method of claim 4 , further comprising:

correcting each subset of the subsets of projection data based on the respective scatter-distribution data of the subset.

6 . The method of claim 1 , wherein the one or more sets of preliminary scattering data include one or more sets of first-scatter distributions.

7 . The method of claim 1 , wherein the one or more sets of preliminary scattering data include one or more structure-mask projections.

8 . An X-ray imaging apparatus comprising:

an X-ray tube;

an X-ray detector; and

one or more memories and processing circuitry configured to

obtain a set of projection data acquired from a computed tomography (CT) scan of an object;

generate, based on the set of projection data, one or more sets of preliminary scattering data, wherein the one or more sets of preliminary scattering data are different from the set of projection data; and

input the obtained set of projection data and the generated one or more sets of preliminary scattering data into a trained neural network, wherein, based on the set of projection data and on the one or more sets of preliminary scattering data, the trained neural network outputs scatter-distribution data.

9 . The X-ray imaging apparatus of claim 8 , wherein, to generate the one or more sets of preliminary scattering data, the one or more memories and processing circuitry are further configured to

generate a reconstructed image of the object based on the set of projection data, and

generate the one or more sets of preliminary scattering data based on the reconstructed image of the object.

10 . The X-ray imaging apparatus of claim 8 ,

wherein the set of projection data includes subsets of projection data that were acquired at a plurality of detection angles,

wherein at least some of the plurality of detection angles are different from other detection angles of the plurality of detection angles,

wherein each subset of the subsets of projection data was acquired at a respective detection angle of the plurality of detection angles,

wherein the respective detection angle of each subset of the subsets of projection data is different from the respective detection angle of at least one other subset of the subsets of projection data, and

wherein the one or more sets of preliminary scattering data include a respective set of preliminary scattering data for each detection angle.

11 . The X-ray imaging apparatus of claim 10 , wherein the trained neural network outputs respective scatter-distribution data for each subset of the subsets of projection data.

12 . The X-ray imaging apparatus of claim 11 , wherein the one or more memories and processing circuitry are further configured to

correct each subset of the subsets of projection data based on the respective scatter-distribution data of the subset.

13 . The X-ray imaging apparatus of claim 10 , wherein the one or more sets of preliminary scattering data include one or more sets of first-scatter distributions.

14 . The X-ray imaging apparatus of claim 10 , wherein the one or more sets of preliminary scattering data include one or more structure-mask projections.

15 . A medical image processing apparatus comprising:

one or more memories and processing circuitry configured to

obtain a set of projection data acquired from a computed tomography (CT) scan of an object;

generate, based on the set of projection data, one or more sets of preliminary scattering data, wherein the one or more sets of preliminary scattering data are different from the set of projection data; and

input the obtained set of projection data and the generated one or more sets of preliminary scattering data into a trained neural network, wherein, based on the set of projection data and on the one or more sets of preliminary scattering data, the trained neural network outputs scatter-distribution data.

16 . The medical image processing apparatus of claim 15 , wherein the trained neural network operates in a projection space, and wherein the projection data in the set of projection data are defined in the projection space.

17 . The medical image processing apparatus of claim 15 , wherein the one or more sets of preliminary scattering data include one or more sets of first-scatter distributions or one or more structure-mask projections.

18 . The medical image processing apparatus of claim 15 , wherein the one or more sets of preliminary scattering data are defined in a projection space, and wherein the projection data in the set of projection data are defined in the projection space.

19 . The medical image processing apparatus of claim 15 , wherein, to generate the one or more sets of preliminary scattering data, the one or more memories and processing circuitry are further configured to

generate a reconstructed image of the object based on the set of projection data, and

generate the one or more sets of preliminary scattering data based on the reconstructed image of the object.

20 . The medical image processing apparatus of claim 19 ,

wherein the set of projection data includes subsets of projection data that were acquired at a plurality of detection angles,

wherein at least some of the plurality of detection angles are different from other detection angles of the plurality of detection angles,

wherein each subset of the subsets of projection data was acquired at a respective detection angle of the plurality of detection angles,

wherein the respective detection angle of each subset of the subsets of projection data is different from the respective detection angle of at least one other subset of the subsets of projection data, and

wherein the one or more sets of preliminary scattering data include a respective set of preliminary scattering data for each detection angle.