# US 11,308,664 B2 · Method for reconstructing a three-dimensional image data set > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,308,664 B2 - Document ID: US11308664B2 - Status: Granted - Kind: B2 - Application number: 16/654,312 - Filed: 2019-10-16 - Published: 2020-04-23 - Granted: 2022-04-19 - Earliest claimed: 2018-10-17 - Assignees: Siemens Healthcare GmbH - Applicants: Siemens Healthcare GmbH - Inventors: Michael Manhart; Yixing Huang; Alexander Preuhs; Günter Lauritsch - Web page: https://patentgrep.com/patent/US11308664B2 ## CPC - G06T 12/30: Image post-processing, e.g. metal artefact correction - G06T 5/50: using two or more images, e.g. averaging or subtraction - G06T 12/20: Inverse problem, i.e. transformations from projection space into object space - G06T 2207/10116: X-ray image - G06T 2207/20081: Training; Learning - G06T 2207/20084: Artificial neural networks [ANN] - G06T 2207/20224: Image subtraction - G06T 2207/30004: Biomedical image processing - G06T 2211/421: Filtered back projection [FBP] - G06T 2211/424: Iterative - G06T 2211/436: Limited angle - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL ## Abstract Systems and methods are provided for reconstructing a three-dimensional result image data set from computed tomography from a plurality of two-dimensional images that create an image of an object undergoing examination from a particular imaging angle, The imaging angles of all the images lie within a restricted angular range. A three-dimensional artifact-reduced image data set is provided based on the two-dimensional images using an algorithm for reducing artifacts that are the result of a restriction of the angular range. The result image data set is reconstructed using a reconstruction algorithm that processes both the artifact-reduced image data set and the two-dimensional images as input data. ## Claims ### Claim 1 A method for reconstructing a three-dimensional result image data set from computed tomography from a plurality of two-dimensional images that create an image of an object undergoing examination from a particular imaging angle, the method comprising: providing a three-dimensional artifact-reduced image data set as a function of the plurality of two-dimensional images, wherein the plurality of two-dimensional images or a processing result that is dependent on the plurality of two-dimensional images are processed by an algorithm for reducing artifacts; and reconstructing the three-dimensional result image data set using a reconstruction algorithm that processes using as input data both the three-dimensional artifact-reduced image data set and the plurality of two-dimensional images. ### Claim 2 (depends on 1) The method of claim 1, wherein imaging angles of all the plurality of two-dimensional images lie within a restricted angular range, and in that the plurality of two-dimensional images or the processing result are processed by the algorithm for reducing artifacts resulting from a restriction on an angular range. ### Claim 3 (depends on 1) The method of claim 1, wherein the algorithm for reducing artifacts is an algorithm that has been or is trained by a machine learning method. ### Claim 4 (depends on 1) The method of claim 1, wherein the processing result of processing the plurality of two-dimensional images includes a reconstructed provisional three-dimensional image data set that is further processed by the algorithm for reducing artifacts to provide the artifact-reduced image data set. ### Claim 5 (depends on 1) The method of claim 1, wherein for reconstructing the result image data set, an iterative reconstruction method is used, wherein an assumed image data set is varied at each iteration step in dependence on images of the object undergoing examination. ### Claim 6 (depends on 5) The method of claim 5 wherein in a first iteration step of the iterative reconstruction method, the artifact-reduced image data set is used as the assumed image data set. ### Claim 7 (depends on 5) The method of claim 5, wherein a cost function is minimized during the iterative reconstruction method, wherein the minimization takes place with at least one boundary condition, the cost function depending on the assumed image data set and the artifact-reduced image data set. ### Claim 8 (depends on 7) The method of claim 7, wherein the cost function depends on difference data determined by subtracting data of the assumed image data set and the artifact-reduced image data set from one another. ### Claim 9 (depends on 1) The method of claim 1, wherein for reconstructing the result image data set, filtered back projection is used. ### Claim 10 (depends on 2) The method of claim 2, wherein for at least one imaging angle lying outside the restricted angular range, a synthetic two-dimensional image is generated by simulating an image of the artifact-reduced image data set, wherein the result image data set is reconstructed as a function of the synthetic image two-dimensional image. ### Claim 11 A method for reconstructing a three-dimensional result image data set, the method comprising: acquiring a plurality of two-dimensional images of an object undergoing examination from a particular imaging angle; generating, using an algorithm for reducing artifacts, a three-dimensional artifact-reduced image data set as a function of the plurality of two-dimensional images; and reconstructing the three-dimensional result image data set, using a reconstruction algorithm, from both the three-dimensional artifact-reduced image data set and the plurality of two-dimensional images. ### Claim 12 An X-ray system comprising: an image capturing device configured to capture two-dimensional images of an object undergoing examination from a plurality of imaging angles; and a processing device configured to generate a three-dimensional artifact-reduced image data set as a function of the two-dimensional images processed by an algorithm for reducing artifacts and reconstruct a three-dimensional result image data set using a reconstruction algorithm that processes both the artifact-reduced image data set or processing data that is generated from the artifact-reduced image data set and the two-dimensional images as input data. ### Claim 13 (depends on 12) The X-ray system of claim 12, wherein the plurality of imaging angles of all the two-dimensional images lie within a restricted angular range, and in that the two-dimensional images are processed by the algorithm for reducing artifacts resulting from a restriction on an angular range. ### Claim 14 (depends on 12) The X-ray system of claim 12, wherein the algorithm for reducing artifacts is an algorithm that has been or is trained by a machine learning method. ### Claim 15 (depends on 12) The X-ray system of claim 12, wherein the processing device is configured to reconstruct the three-dimensional result image data set using an iterative reconstruction method, wherein an assumed image data set is varied at each iteration step in dependence on images of the object undergoing examination. ### Claim 16 (depends on 15) The X-ray system of claim 15, wherein in a first iteration step, the artifact-reduced image data set is used as the assumed image data set. ### Claim 17 (depends on 15) The X-ray system of claim 15, wherein a cost function is minimized during the iterative reconstruction method, wherein the minimization takes place with at least one boundary condition, the cost function depending on the assumed image data set and the artifact-reduced image data set. ### Claim 18 (depends on 17) The X-ray system of claim 17, wherein the cost function depends on difference data determined by subtracting data of the assumed image data set and the artifact-reduced image data set from one another. ### Claim 19 (depends on 12) The X-ray system of claim 12, wherein the processing device is configured to reconstruct the three-dimensional result image data set using filtered back projection. ### Claim 20 (depends on 12) The X-ray system of claim 12, wherein for at least one imaging angle lying outside the restricted angular range, a synthetic two-dimensional image is generated by the processing device by simulating an image of the artifact-reduced image data set, wherein the three-dimensional result image data set is reconstructed by the processing device in dependence on the synthetic image two-dimensional image.