Method for correcting metal artifacts in x-ray imaging, and x-ray device

1. A method for correcting artifacts caused by at least one metal object in a target region to be imaged during a reconstruction of a three-dimensional image dataset from a plurality of two-dimensional projection images of the target region recorded from mutually different projection directions with an x-ray device, the method which comprises:

determining an orientation and a position of a three-dimensional model of the at least one metal object in a coordinate system of the x-ray device by way of a 2D-3D registration with at least one two-dimensional projection image;

inserting into the three-dimensional image dataset an image of the metal object based on the known orientation and position of the model;

for each radiation path of the projection images running through the metal object, based on attenuation information through the metal object assigned to the model, calculating a proportion of attenuation through the metal object for the projection image data assigned to the respective radiation path and using a result of the calculation to determine modified projection images from the projection image data and to thereby correct artifacts caused by the metal object in the target region; and

determining the three-dimensional image dataset from the modified projection images.

2. The method according to claim 1 , which comprises carrying out at least one of the following steps: determining the model from previously known geometrical data and determining the attenuation information from previously known material data of the metal object.

3. The method according to claim 2 , which comprises, where the metal object is an implant in a patient, determining data selected from the group consisting of the geometrical data and the material data from a manufacturer of the metal object.

4. The method according to claim 3 , wherein the metal object is selected from the group consisting of a screw, a plate, and a metal object forming at least one part of an artificial joint, and the geometrical data and material data is acquired from an information system.

5. The method according to claim 2 , which comprises, where the metal object is an implant in a patient, determining data selected from the group consisting of the geometrical data and the material data via a database.

6. The method according to claim 1 , which comprises using a model selected from the group of models consisting of a model based on finite elements and a model based on a defined, regular grid.

7. The method according to claim 6 , wherein spatial elements respectively defined by the finite elements and the grid are each assigned attenuation information.

8. The method according to claim 6 , wherein a size of spatial elements respectively defined by the finite elements and the grid is selected with reference to a voxel size in the three-dimensional image dataset.

9. The method according to claim 1 , which comprises carrying out at least one of the following steps: determining the model from known geometrical data and determining the attenuation information from known material data of the metal object.

10. The method according to claim 1 , which comprises, in determining and evaluating radiation paths during a correction, taking into consideration already-recorded image information, which was recorded from a projection direction differing from the projection direction of a currently observed projection image.

11. The method according to claim 10 , wherein the projection direction of the already-recorded image information differs from the projection direction of the currently observed projection image by a 90° rotation.

12. The method according to claim 10 , which comprises using image information selected from the group consisting of a further projection image contributing to the reconstruction and an additional image additionally recorded beforehand as the image information.

13. The method according to claim 12 , which comprises using the image information of a projection image recorded with a relatively lower radiation dose than the projection images used for reconstruction.

14. The method according to claim 1 , wherein, as starting point for the 2D-3D-registration with a further projection image, using an orientation and a position of the model in the projection image observed beforehand determined in the projection image observed immediately beforehand, adjacent as regards the projection direction.

15. A method for correcting artifacts caused by at least one metal object in a target region to be imaged during a reconstruction of a three-dimensional image dataset from a plurality of two-dimensional projection images of the target region recorded from mutually different projection directions with an x-ray device, the method which comprises:

determining an orientation and a position of a three-dimensional model of the at least one metal object in a coordinate system of the x-ray device by way of a 2D-3D registration with at least one two-dimensional projection image;

for each radiation path of the projection images running through the metal object, based on attenuation information through the metal object assigned to the model, calculating a proportion of attenuation through the metal object for the projection image data assigned to the respective radiation path and using a result of the calculation to determine modified projection images from the projection image data and to thereby correct artifacts caused by the metal object in the target region;

determining beam hardening information produced by the at least one metal object from the model known in orientation and position and using the hardening information as part of a beam hardening correction, and thereby making a distinction along a radiation path between structures lying in front of the at least one metal object and structures lying behind the at least one metal object; and

determining the three-dimensional image dataset from the modified projection images.

16. A method for correcting artifacts caused by at least one metal object in a target region to be imaged during a reconstruction of a three-dimensional image dataset from a plurality of two-dimensional projection images of the target region recorded from mutually different projection directions with an x-ray device, the method which comprises:

determining an orientation and a position of a three-dimensional model of the at least one metal object in a coordinate system of the x-ray device by way of a 2D-3D registration based on a single two-dimensional projection image;

for each radiation path of the projection images running through the metal object, based on attenuation information through the metal object assigned to the model, calculating a proportion of attenuation through the metal object for the projection image data assigned to the respective radiation path and using a result of the calculation to determine modified projection images from the projection image data and to thereby correct artifacts caused by the metal object in the target region; and

determining the three-dimensional image dataset from the modified projection images.

17. The method according to claim 16 , which comprises determining the modified projection images for each projection image directly after recording.

18. A method for correcting artifacts caused by at least one metal object in a target region to be imaged during a reconstruction of a three-dimensional image dataset from a plurality of two-dimensional projection images of the target region recorded from mutually different projection directions with an x-ray device, the method which comprises:

determining an orientation and a position of a three-dimensional model of the at least one metal object in a coordinate system of the x-ray device by way of a 2D-3D registration with a least one two-dimensional projection image;

for each radiation path of the projection images running through the metal object, based on attenuation information through the metal object assigned to the model, calculating a proportion of attenuation through the metal object for the projection image data assigned to the respective radiation path and using a result of the calculation to determine modified projection images from the projection image data and to thereby correct artifacts caused by the metal object in the target region;

effecting a fine correction by determining virtual projection images from the three-dimensional image dataset with an inserted image of the metal object through forward projection and comparing with the recorded projection images, wherein the image data of the projection images is corrected as a function of the comparison, optionally with an iterative process; and

determining the three-dimensional image dataset from the modified projection images.

19. The method according to claim 18 , which comprises, for a multiple reconstruction of three-dimensional image datasets, especially as part of an iterative process, determining correction information, especially related to the beam modeling and taking into account the orientation and position and the attenuation information, and including the correction information for correction of the image data.

20. An x-ray device, comprising:

an imaging arrangement including an x-ray emitter and an x-ray detector for recording projection images of a target region from a plurality of different projection directions; and

a control device configured for correcting artifacts caused by at least one metal object in a target region to be imaged during a reconstruction of a three-dimensional image dataset from a plurality of two-dimensional projection images of the target region recorded from mutually different projection directions with said imaging arrangement, by carrying out the following process steps:

determining an orientation and a position of a three-dimensional model of the at least one metal object in a coordinate system of the x-ray device by way of a 2D-3D registration with at least one two-dimensional projection image;

for each radiation path of the projection images running through the metal object, based on attenuation information through the metal object assigned to the model, calculating a proportion of attenuation through the metal object for the projection image data assigned to the respective radiation path and using a result of the calculation to determine modified projection images from the projection image data; and

determining the three-dimensional image dataset from the modified projection images.