Method for generating a volume model of an object under investigation, control unit, X-ray device, computer program and electronically readable data carrier

Abstract

The invention relates to a method for generating a volume model (11) of an object of investigation (5) from a data set (9), wherein the data set (9) comprises first 2D projection images (7) of an object of investigation (5), created in a first X-ray spectrum and second 2D projection images (8) of the object of investigation (5), created in a second X-ray spectrum.The method comprises supplementing the 2D projection images (7, 8) with fill projections (15), and reconstructing a preliminary volume model (19) of the object under investigation (5) from the supplemented 2D projection images (7, 8); the method comprises determining a spectral X-ray absorption property of at least one volume element (18) and generating virtual 2D projection images (20) from the preliminary volume model (19) of the object under investigation (5) depending on the spectral X-ray absorption property of the at least one volume element (18); and reconstructing the volume model (11) of the object under investigation (5) from the virtual 2D projection images (20) and / or the supplemented 2D projection images (7, 8) and the virtual 2D projection images (20).

Description

[0001] The invention relates to a method for generating a volume model of an examination object. The invention also relates to a control device, an X-ray device, a computer program, and an electronically readable data storage medium.

[0002] Cone beam computed tomography (CBCT) and limited angle tomography are imaging techniques used to create a reconstruction image of an object under examination.

[0003] In both techniques, X-rays are emitted from an X-ray source onto the object under examination, with a main beam region emanating from the X-ray source. In cone-beam computed tomography, this main beam region is cone-shaped. Opposite the X-ray source is a surface detector, which detects the emitted X-rays as a two-dimensional projection image of the object under examination. The object under examination is positioned between the X-ray source and the surface detector. To enable the reconstruction of a reconstruction image of the object under examination, a data set comprising several of the 2D projection images of the object under examination is acquired.

[0004] To acquire the 2D projection images of the object under examination for the dataset, the X-ray source and the area detector are moved along a usually circular trajectory around the object under examination. The respective 2D projection images of the object under examination are generated at specific positions along the trajectory. This results in the dataset comprising multiple 2D projection images that depict the object under examination in different orientations. The 2D projection images of the dataset are processed according to a predetermined reconstruction method to generate a reconstruction image of the object under examination.

[0005] For optimal reconstruction of the examination subject in cone-beam computed tomography (CBCT), it is necessary that the X-ray beams penetrate the examination subject in each position. In other words, the entire examination subject must be located within the cone-shaped main beam area during the acquisition of the respective projection image.

[0006] A special type of cone-beam computed tomography (CBCT) is dual-energy cone-beam computed tomography (DE-CBCT). This technique involves acquiring 2D projection images of the object under examination using two different X-ray spectra.

[0007] The 3D reconstruction and spectral analysis of 2D projection images from dual-energy cone-beam computed tomography is significantly hampered by a number of effects. These include scattered radiation, beam hardening, metal artifacts, and cupping artifacts.

[0008] A particular challenge, however, arises when truncation occurs in the 2D projection images for dual-energy cone-beam computed tomography. Truncation occurs when parts of the patient or the subject under examination are not captured from the full angular range required for 3D reconstruction, resulting in insufficient information about these parts of the patient for reconstruction.

[0009] For conventional cone-beam computed tomography with only a single acquisition spectrum, there are numerous known correction approaches to deal with truncation and the associated reconstruction artifacts. However, spectral effects are not taken into account in these models, which leads to reconstruction artifacts in the models from dual-energy cone-beam computed tomography.

[0010] The fundamental problem with limited-angle tomography is that the 2D projection images are only recorded over a relatively limited angular range. This can also lead to reconstruction artifacts.

[0011] The following methods are known from the state of the art in this field, but they are not directed at the problem mentioned above.

[0012] The publication by Clark, Darin P., et al. (Clark, Darin P., et al. “Deep learning based spectral extrapolation for dual-source, dual-energy x-ray computed tomography.” Medical physics 47.9 (2020): 4150-4163.) describes a deep learning-based approach for dual-energy fan beam CT. This publication concerns a dual-energy CT scanner that has a larger image reconstruction volume for a first X-ray spectrum than for a second X-ray spectrum. In this case, a full 3D reconstruction is available for the first spectrum, while truncation occurs for the second spectrum. A 3D reconstruction of the second X-ray spectrum is predicted from the 3D reconstruction of the first X-ray spectrum using a neural network.

[0013] The publication by Huang, Yixing, et al. (Huang, Yixing, et al. “Data extrapolation from learned prior images for truncation correction in computed tomography.” IEEE Transactions on Medical Imaging 40.11 (2021): 3042-3053.) describes a method for correcting truncation, although this correction is not directed to the case of using two X-ray spectra.

[0014] Furthermore, the paper by Liu, Chi-Kuang and Huang, Hsuan-Ming, “Unsupervised deep learning based image outpainting for dual-source, dual-energy computed tomography.” In: Radiation Physics and Chemistry, Vol. 188, 2021, p. 109635, discloses a method based on unsupervised deep learning, whereby computed tomography images acquired using two sources (dual-source) and two energy ranges (dual-energy) can be extended outside of an original imaging range.

[0015] It is an object of the invention to provide a solution for compensating truncation for tomography methods that acquire projection images with at least two X-ray spectra. In particular, a solution is to be provided for dual-energy cone-beam computed tomography and limited-angle tomography.

[0016] This object is achieved by the respective subject matter of the independent claims. Advantageous further developments and preferred embodiments are the subject matter of the dependent claims.

[0017] One approach of the invention is to create a truncation model of the object under investigation with spectral properties and / or an assumed material composition from 2D projection images of a measurement for dual-energy cone beam computed tomography or limited-angle tomography.

[0018] A first aspect of the invention relates to a method for generating a volume model of an object under investigation from a data set.

[0019] The dataset includes first 2D projection images of an examination object, captured in a first X-ray spectrum of the examination object. The dataset also includes second 2D projection images of the examination object, captured in a second X-ray spectrum of the examination object. In other words, the dataset includes the 2D projection images that separately image the examination object in the two X-ray spectra. The X-ray spectra can include different energy spectra.

[0020] The 2D projection images each depict a projection volume of the examination object, which is arranged in a field of view of the respective 2D projection images. In other words, each of the 2D projection images comprises a respective object projection, which depicts the projection volume of the examination object. The respective projection volume is a partial volume of a total volume of the examination object. The projection volume is the volume that is arranged in the respective field of view of the X-ray device when the respective 2D projection image is created. The projection volume of the 2D projection image is identical to the total volume of the examination object if the examination object is completely located in the field of view of the X-ray device when the respective 2D projection image is taken.If the object under examination is not completely within the field of view, the 2D projection image in question will only image that part of the object's total volume that is within the field of view. In this case, the projection volume does not include the entire total volume. The 2D projection image in question therefore lacks defective regions that would image the part of the total volume outside the projection volume. Because the entire volume of the object under examination is not captured in the 2D projection image in question, there is a lack of information for reconstructing the non-captured region. This region is referred to as the truncation volume. In contrast, the part of the total volume that was within the respective field of view for each acquisition of the 2D projection image is referred to as the primary capture region.

[0021] The method comprises steps which are carried out by a control device.

[0022] A first step involves the control device receiving the data set from an X-ray device. The X-ray device can be a device for performing a CT examination, which has created the 2D projection images of the examination subject.

[0023] In a subsequent step, the 2D projection images are supplemented. The supplementation comprises extending at least one truncated 2D projection image of the first 2D projection images and / or a truncated 2D projection image of the second 2D projection images with a fill projection. The fill projection depicts a fill volume of the examination object located outside the field of view of the at least one truncated 2D projection image.

[0024] The fill volume is a part of the examination object that is located outside the respective field of view of the respective 2D projection image. In other words, it is a part of the total volume of the examination object that is located outside the projection volume depicted by the 2D projection image. In order to be able to take into account the parts of the total volume not covered by the projection volume in the respective 2D projection images, the individual 2D projection images are supplemented with the fill projections. In other words, individual 2D projection images are supplemented and / or extrapolated to compensate for missing areas that arise because the examination object is not completely positioned in the field of view. Therefore, not all of the 2D projection images need to be supplemented, but only those in which part of the examination object is truncated.

[0025] In one step of the method, a preliminary volume model of the examination object is reconstructed from the supplemented 2D projection images. In other words, the supplemented 2D projection images, which include the at least one extended truncated 2D projection image, are used to reconstruct the preliminary volume model of the examination object. The reconstruction can be performed using a conventional reconstruction method according to the state of the art.

[0026] In a further step, a spectral X-ray absorption property of at least one volume element is determined, wherein the at least one volume element is located within the fill volume of the at least one truncated 2D projection image. In other words, the at least one volume element is located within a truncation volume of the preliminary volume model. The spectral X-ray absorption property is determined depending on the supplemented 2D projection images.

[0027] In a subsequent step, virtual 2D projection images of the examination object are generated for the two X-ray spectra from the preliminary volume model of the examination object. The virtual 2D projection images are generated depending on the spectral X-ray absorption property of the at least one volume element. In other words, the control device performs a virtual examination of the examination object. This generates the virtual 2D projection images of the preliminary volume model. When generating the virtual 2D projection images, the spectral X-ray absorption property of the at least one volume element is taken into account. In other words, for example, an absorption of rays from the X-ray spectra that depends on the X-ray absorption property of the at least one volume element can be simulated.

[0028] In a subsequent step, the volume model of the examination object was reconstructed from the virtual 2D projection images or from the supplemented 2D projection images and the virtual 2D projection images. A known, state-of-the-art reconstruction method can be applied. The reconstruction can be based solely on the virtual 2D projection images. Alternatively, the reconstruction can be based on the supplemented 2D projection images and the virtual 2D projection images. For example, virtual 2D projection images can be used for areas of the examination object that were not irradiated by the acquired 2D projection images.

[0029] The invention offers the advantage that truncation in dual-energy cone beam computed tomography can be compensated.

[0030] The invention also includes further developments which result in further advantages.

[0031] A further development of the invention provides that the volume model of the examination object comprises a single-energy volume model of the examination object for the first X-ray spectrum and a single-energy volume model of the examination object for the second X-ray spectrum. In other words, the volume model comprises two separate single-energy volume models of the respective X-ray spectra. It is provided that the single-energy volume model of the examination object for the first X-ray spectrum is reconstructed from the virtual 2D projection images generated in the first X-ray spectrum or from the supplemented first 2D projection images. In other words, the reconstruction of the single-energy volume model of the examination object for the first X-ray spectrum takes place from the virtual 2D projection images generated for the first X-ray spectrum.Alternatively, the reconstruction is performed from the first projection images and the virtual 2D projection images generated for the first X-ray spectrum. Accordingly, the single-energy volume model of the examination object for the second X-ray spectrum is reconstructed from the virtual 2D projection images generated in the second X-ray spectrum or from the supplemented second 2D projection images. In other words, the reconstruction of the single-energy volume model of the examination object for the second X-ray spectrum is performed from the virtual 2D projection images generated for the second X-ray spectrum. Alternatively, the reconstruction is performed from the second projection images and the virtual 2D projection images generated for the second X-ray spectrum.

[0032] A further development of the invention provides that the volume model of the examination object comprises a dual-energy volume model of the examination object for both X-ray spectra. In other words, a dual-energy volume model is generated that characterizes the examination object in both X-ray spectra. It is provided that the dual-energy volume model of the examination object for both X-ray spectra is reconstructed from the virtual 2D projection images generated in both X-ray spectra or from the supplemented first 2D projection images and the supplemented second 2D projection images.

[0033] A further development of the invention provides that the supplementation of the 2D projection images comprises a deep learning method. In other words, the deep learning method is applied to supplement the 2D projection images with the fill projections. For example, it can be provided that an artificial neural network underlying the deep learning method is trained on 2D projection images and / or volume models of objects of a category of the examination object. This makes it possible to reconstruct the truncated regions of the examination object based on known objects. One possible method is disclosed in the publication by Huang, Yixing, et al. (Huang, Yixing, et al. "Data extrapolation from learned prior images for truncation correction in computed tomography." IEEE Transactions on Medical Imaging 40.11 (2021): 3042-3053.).For example, the truncated 2D projection image can be fed to the artificial neural network along with additional information about the object under investigation as input data. Additional 2D projection images can also be provided as input data. The neural network can reproduce the truncated 2D projection image, supplemented by the filler projection, as output data.

[0034] A further development of the invention provides that supplementing the 2D projection images comprises providing the fill projections from a predefined standard model of the examination object. In other words, a standard model is provided for the examination object, which is used to supplement the fill projections. For example, it can be provided that the standard model is a water cylinder model. When generating the fill projections, it can be assumed that the examination object is arranged in a cylindrical volume of water.

[0035] A further development of the invention provides that the method includes a step of reconstructing the preliminary volume model of the examination object from the non-supplemented 2D projection images. In other words, before generating the preliminary volume model from the supplemented 2D projection images, the preliminary volume model is generated from the non-supplemented 2D projection images. This provides the advantage that the supplementation of the 2D projection images can be performed by incorporating a non-truncated volume of the preliminary volume model.

[0036] According to a first variant of the method, the method comprises identifying a material at least in a partial region of one of the 2D projection images based on a pattern of the partial region. In other words, an identification method is provided for the at least one partial region of the 2D projection image in order to identify the material of the partial region. It can be provided that patterns assigned to different materials are stored in a database of the control device. A bone material can, for example, be distinguished from soft tissue by a characteristic structure of the pattern. The respective spectral absorption properties can be assigned to the material in the database. The spectral X-ray absorption properties of the at least one volume element are determined based on the material.

[0037] According to a second variant of the method, the method includes identifying a partial object volume of at least one partial object of the object under investigation in the preliminary volume model. In other words, the preliminary volume model is examined for the presence of the partial object volume.

[0038] In a subsequent step, the partial object volume is added to the preliminary solid model as a fill volume. In other words, a portion of the partial object volume may be located outside the primary acquisition area, making a complete reconstruction of the partial object volume from the unsupplemented 2D projection images impossible. By identifying the partial object, a partial object volume of the partial object can be retrieved from the database, and the partial object volume can be added to the solid model.

[0039] A further development of the invention provides that the method comprises identifying image features of at least one partial object of the examination object in the 2D projection images. In other words, it is provided that the image features assigned to the partial object of the examination object are identified in the 2D projection images. The image feature can be, for example, an outline, an edge, or a point of the partial object. The image feature can be identified in the 2D projection images by the control device using image recognition methods. It can be provided, for example, that the control device is provided with a database that can describe the partial object and / or the image features of the partial object.

[0040] In a subsequent step, a position of the partial object volume of the at least one partial object in the volume model of the examination object is determined depending on the positions of the respective image features of the at least one partial object in the 2D projection images. In other words, the image features of the at least one partial object are identified in the respective 2D projection images of the examination object, and their respective positions in the projection image are determined. The position of the partial object in the volume model is determined from the positions of the image features assigned to the respective partial object.

[0041] A further development of the invention provides that the at least one volume element is arranged within the truncated partial object volume, and determining the spectral X-ray absorption property of the at least one volume element comprises retrieving the spectral X-ray absorption property of the partial object volume from a database. In other words, the spectral X-ray absorption properties can be stored for the partial object. The spectral X-ray absorption property of at least the region of the partial object volume comprising the at least one volume element can be retrieved from the database. The spectral X-ray absorption property can be assigned to the volume element.

[0042] According to a third variant of the method, determining the spectral X-ray absorption property of the at least one volume element comprises the following steps: Identifying a surface element assigned to the at least one volume element in one of the first 2D projection images. In other words, the surface element that, for example, images the volume element in the first 2D projection image is identified. Identifying the surface element assigned to the at least one volume element in one of the second 2D projection images. In other words, the surface element is identified in the second 2D projection image. This makes it possible to provide projection values, also called signal intensity values, which are recorded in the surface element for determining the X-ray absorption property of the volume element.In a subsequent step, the spectral X-ray absorption property of the volume element is determined from the projection values ​​of the surface element in the 2D projection images of the two X-ray spectra.

[0043] A further development of the invention provides that the method comprises the following steps to be carried out by the X-ray device.

[0044] Acquiring the first 2D projection images of the examination subject in the first X-ray spectrum. Acquiring the second 2D projection images of the examination subject in the second X-ray spectrum. One step comprises providing the data set comprising the first 2D projection images and the second 2D projection images to the control device.

[0045] A further development of the invention provides that the 2D projection images are generated using a dual-energy cone beam computed tomography method. The 2D projection images can be provided to the control device by an external X-ray device. The 2D projection images can also be acquired by the X-ray device and provided to the control device as part of the method.

[0046] A further development of the invention provides that the 2D projection images are generated using a limited-angle tomography method. The 2D projection images can be provided to the control device by an external X-ray device. The 2D projection images can also be acquired by the X-ray device and provided to the control device as part of the method.

[0047] In other words, the method is applied to 2D projection images from a limited-angle tomography procedure. For example, additional information for the limited-angle tomography can be generated through multiple spectral calculations of projection images with the various X-ray spectra. This spectral information can be used as input data for model-based or iterative limited-angle tomography.

[0048] For use cases or application situations that may arise during the method and which are not explicitly described here, it may be provided that, in accordance with the method, an error message and / or a request to enter user feedback is issued and / or a default setting and / or a predetermined initial state is set.

[0049] A second aspect of the invention relates to a control device configured to generate a volume model of an examination object from a data set.

[0050] The data set comprises first 2D projection images of an examination subject, created in a first X-ray spectrum, and second 2D projection images of the examination subject, created in a second X-ray spectrum. The 2D projection images each depict a projection volume of the examination subject, which is arranged in a field of view of the respective 2D projection image. The control device is configured to receive the data set from an X-ray device. The control device is configured to supplement the 2D projection images.

[0051] The control device is configured to supplement the 2D projection images by extending at least one truncated 2D projection image of the first 2D projection images and / or the second 2D projection images with a fill projection by the control device. The fill projection depicts a fill volume of the examination object located outside the field of view of the at least one truncated 2D projection image.

[0052] The control device is configured to determine a spectral X-ray absorption property of at least one volume element arranged within the fill volume of the at least one truncated 2D projection image as a function of the supplemented 2D projection images. The control device is configured to generate virtual 2D projection images of the X-ray spectra of the examination object from the preliminary volume model of the examination object as a function of the spectral X-ray absorption property of the at least one volume element. The control device is configured to reconstruct the volume model of the examination object from the virtual 2D projection images or from the supplemented 2D projection images and the virtual 2D projection images.

[0053] According to a first variant of the control device, it is provided that the control device is designed to: - to identify a material of at least a partial area of ​​one of the 2D projection images based on a pattern of the partial area; - to determine the X-ray absorption property of the at least one volume element arranged within the filling volume of the at least one truncated 2D projection image as a function of the material of the partial region.

[0054] According to a second variant of the control device, it is provided that the control device: - a surface element assigned to the volume element is identified in one of the first 2D projection images, - the surface element assigned to the volume element is identified in one of the second 2D projection images and - the spectral X-ray absorption property of the volume element is determined from projection values ​​of the surface element in the 2D projection images of the two X-ray spectra.

[0055] According to a third variant of the control device, it is provided that the control device is designed to: - to identify a sub-object volume of at least one sub-object of the object under investigation in the preliminary volume model; - to add a truncated sub-object volume as a filling volume in the preliminary volume model.

[0056] The control device can in particular contain one or more computers, one or more microcontrollers and / or one or more integrated circuits, for example one or more application-specific integrated circuits (ASICs), one or more field-programmable gate arrays (FPGAs), and / or one or more single-chip systems (SoCs). The computing unit can also contain one or more processors, for example one or more microprocessors, one or more central processing units (CPUs), one or more graphics processing units (GPUs), and / or one or more signal processors, in particular one or more digital signal processors (DSPs). The computing unit can also contain a physical or virtual network of computers or other of the aforementioned units.

[0057] A third aspect of the invention relates to an X-ray device. The X-ray device has at least one control device according to the second aspect of the invention. The X-ray device is configured to acquire first 2D projection images of an examination object, created in a first X-ray spectrum, and second 2D projection images of the examination object, created in a second X-ray spectrum, wherein each of the 2D projection images depicts a respective object projection of a respective projection volume of the examination object. The respective projection volume is a partial volume of a total volume of the examination object, which is arranged in a respective field of view of the respective 2D projection image. The X-ray device is configured to provide a data set of the 2D projection images of the examination object to the control device for generating a volume model of the examination object.

[0058] A fourth aspect of the invention relates to a computer program that can be loaded directly into a memory of a control device, comprising program means for executing the steps of the above-mentioned method when the program is executed in the control device. The method described herein can therefore also be in the form of a computer program product that implements the method on a control device when it is executed on the control device.

[0059] A fifth aspect of the invention relates to an electronically readable data carrier with electronically readable control information stored thereon, which comprises at least one described computer program and is designed such that it carries out a described method when the data carrier is used in a control device.

[0060] Regardless of the grammatical gender of a particular term, persons with male, female or other gender identity are included.

[0061] Further features of the invention emerge from the claims, the figures and the description of the figures. The features and combinations of features mentioned above in the description as well as the features and combinations of features mentioned below in the description of the figures and / or shown in the figures can be encompassed by the invention not only in the respectively specified combination, but also in other combinations. In particular, the invention can also encompass embodiments and combinations of features that do not have all the features of an originally formulated claim. Furthermore, the invention can encompass embodiments and combinations of features that go beyond the combinations of features set out in the backreferences to the claims or deviate from them.

[0062] The invention is explained in more detail below using specific embodiments and associated schematic drawings. In the figures, identical or functionally equivalent elements may be provided with the same reference numerals. The description of identical or functionally equivalent elements may not necessarily be repeated for different figures. The figures show: Fig. 1 a schematic representation of the acquisition of 2D projection images of an examination object by an X-ray device; Fig. 2 shows a schematic representation of 2D projection images of an object under investigation; Fig. 3 shows a schematic representation of a reconstructed volume model; and Fig. 4 shows a schematic representation of a process flow.

[0063] Fig. 1 shows a schematic representation of the acquisition of 2D projection images of an examination object by an X-ray device.

[0064] The X-ray device 1 can have a detection device 2, which can include an X-ray source 3 and a detection screen 4. The X-ray source 3 can be configured to emit X-rays along a cone volume in the direction of the detection screen 4. The X-ray device 1 can be configured to perform a dual-energy cone beam computed tomography method and / or a limited-angle tomography method. For this purpose, the X-ray device 1 can emit the X-rays in two different X-ray spectra.

[0065] An examination object 5 can be arranged between the X-ray source 3 and the detection screen 4. The X-rays are absorbed by the examination object 5, so that the detection screen 4 captures a two-dimensional projection image 7, 8 of the examination object 5. The absorption of the X-rays in the examination object 5 can depend on the X-ray spectrum of the X-rays. The X-rays of the first X-ray spectrum can be captured in first 2D projection images 7, and the X-rays of the second X-ray spectrum can be captured in second 2D projection images 8.

[0066] Reconstructing a volume model 11 of the examination object 5 requires capturing a plurality of 2D projection images 7, 8 of the examination object 5 from different directions. For this purpose, the X-ray device 1 can be configured to move the X-ray source 3 and the detection screen 4 along a, for example, circular trajectory 6 around the examination object 5. The respective 2D projection images 7, 8 of the examination object 5 can be captured in predetermined directions in the respective X-ray spectrum and added to a data set 9. A control device 10 of the X-ray device 1 can be configured to reconstruct the volume model 11 of the examination object 5 from the first 2D projection images 7 and the second 2D projection images 8.

[0067] For a complete reconstruction of the volume model 11 of the examination object 5, it is necessary that a total volume 14 of the examination object 5 is completely captured during each measurement. To do this, the total volume 14 must be arranged within the cone for each recording of the respective 2D projection images 7, 8, so that the examination object 5 is completely irradiated and the projection image 7, 8 encompasses the examination object 5 over its entire respective dimension. However, it may be the case that, at least in some positions of the detection device 2, defective areas 28 of the examination object 5 lie outside the cone and are therefore not captured by the respective projection image 7, 8. In this case, a so-called truncation occurs. This can lead to errors in the volume model 11 of the examination object 5.

[0068] An area that is sufficiently irradiated is referred to as the primary reconstruction volume 27. The other area is referred to as the truncation volume 25. The truncation volume 25 can comprise the volumes that were outside the field of view in at least one recording of one of the projection recordings 7, 8. For reconstruction, it may be necessary to supplement the 2D projection recordings 7, 8 with fill projections 15. The control device 10 can be configured to supplement the 2D projection recordings 7, 8 with the fill projections 15 and / or to supplement a preliminary volume model 19 created from the unsupplemented 2D projection recordings 7, 8 with fill volumes 16. For the supplementation of the fill projections 15, a standard model of the examination object 5 can be assumed, which can depict a water cylinder, for example. The truncation volume 25 can be supplemented by a water cylinder as filling volume 16.

[0069] Partial objects 23, which may have respective partial object volumes 24, can be arranged in the examination object 5. The partial object volumes 24 can be located partly within the primary reconstruction volume 27 and partly outside the primary reconstruction volume 27. To generate the fill volume 16, for example, image features 21 of the partial objects 23 in the 2D projection images 7, 8 can be identified, and their positions in the 2D projection images 7, 8 can be determined. The control device 10 can use this to determine a position of the partial object volume 24 of the partial object 23 in the volume model 11. The partial object volume 24 of the partial object 23 can be stored, for example, in a database of the control device 10. As a result, the control device 10 can identify the partial object 23 in the 2D projection images 7, 8 of the examination object 5 based on the image features 21.A portion of the partial object volume 24 in the truncation volume 25, which is not depicted in one of the 2D projection images 7, 8, can be added as a fill projection 15 in the other projection image 7, 8. The missing portion can be determined or estimated from the partial object volume 24 stored in the database.

[0070] The partial object volumes 24 can also be identified in a preliminary volume model 19 of the examination object 5, which can be determined from the unsupplemented 2D projection images 7, 8. An X-ray absorption property can be determined for at least one volume element 18 of the preliminary volume model 11. The X-ray absorption property can be determined, for example, based on a material 22, which can be identified based on a structure in a partial area of ​​one of the 2D projection images 7, 8. The X-ray absorption property of the material 22 can be stored in a database of the control device 10.

[0071] After the preliminary volume model 11 has been generated, the control device 10 can create virtual 2D projection images 20 from the preliminary volume model 18.

[0072] Fig. 2 shows a schematic representation of 2D projection images 7, 8 of an object under investigation 5.

[0073] The 2D projection images 7, 8 can have an object projection 12 that depicts the projection volume 13 of the examination object 5 of the respective measurement. The projection volume 13 can describe the portion of the total volume 14 of the examination object 5 that is located within the cone and thus the field of view 17 of the respective measurement. The missing area 28 outside the object projection 12 can show the portion of the 2D projection images 7, 8 that has been supplemented.

[0074] The 2D projection images 7, 8 of the examination object 5 can be created from two positions for the respective X-ray spectra. Due to the X-ray absorption properties of materials, intensity values ​​of the 2D projection images 7, 8 can differ from one another depending on the X-ray spectrum. This can, for example, be used to identify a material. The outlines of the partial objects 23 can represent the image features 21 of the partial objects 23, which can be identified by the control device 10. The position of the partial object 23 in the volume model 11 can be determined from the position of the image features 21.

[0075] Fig. 3 shows a schematic representation of a reconstructed volume model 11.

[0076] It can be seen that the truncation volume 25 outside the primary reconstruction volume 27 is supplemented by the filling volume 16. A volume element 18 is shown as an example, the X-ray absorption property of which has been determined.

[0077] Fig. 4 shows a schematic representation of a process flow.

[0078] In a step S1, a DE-CBCT scan of an examination object 5 can be performed using an X-ray device 1. During the DE-CBCT scan, 2D projection images 7, 8 can be acquired with a first and a second X-ray spectrum. The DE-CBCT scan can be acquired using known dual energy methods such as dual source, fast kV switching, fast filter switching, or dual layer & photon counting detectors. The two X-ray spectra can differ from one another in at least one wavelength range. Truncation may have occurred, resulting in 2D projection images 7, 8 in which a defective region 28 of the examination object 5, for example a lateral region, is cut off. The cut-off defective region 28 may have been located outside the field of view 17 when the projection image 7, 8 was acquired.The 2D projection images 7, 8 of the examination object 5 of the X-ray spectra can be added to a data set 9, which can be provided to a control device 10 for generating a volume model 11.

[0079] Optionally, in a step S2, the control device 10 can perform a first image reconstruction for reconstructing a preliminary volume model 19 using known steps according to the prior art for reconstructing the primary reconstruction volume 27. The non-truncated region can describe the total volume 14 of the examination object 5 in the primary reconstruction volume 27. The reconstruction can optionally be performed separately for the 2D projection images 7 of the first X-ray spectrum and the 2D projection images 8 of the second X-ray spectrum. Optionally, the truncation volume 25 shown in the 2D projection images 7, 8 of the two X-ray spectra can also be estimated using known methods for truncation correction.

[0080] In an optional step S3, if possible with the selected dual energy imaging technique, dual energy 2D projection images 7, 8 can be calculated from corresponding 2D projection images 7, 8 of the respective X-ray spectra using weighted subtraction or other image processing methods. If necessary, the corresponding 2D projection images 7, 8 can be combined in a motion-compensated manner.

[0081] In the dual-energy 2D projection images 7, 8, sub-objects 23, such as bones or a contrast agent, can be highlighted or, alternatively, suppressed. For example, using a method described in the publication Maaß, Nicole, et al. Maaß, Nicole, et al. "Empirical multiple energy calibration EMEC for material-selective CT." 2011 IEEE Nuclear Science Symposium Conference Record. IEEE, 2011., material-specific projection values ​​and / or X-ray absorption properties can be determined from the corresponding 2D projection images 7, 8 of the different X-ray spectra.

[0082] In a step S4, sufficiently contrasted partial objects 23 and / or image features 21 can be segmented and / or detected in the 2D projection images 7, 8 or in dual-energy 2D projection images derived from the respective 2D projection images 7, 8. The partial objects 23 can be, for example, bones, the skin surface, accumulations of contrast agent such as contrasted vessels, implants, external objects, or organs. Image features 21, such as landmarks, of the exemplary partial objects 23 can also be detected in the 2D projection images 7, 8. An entire partial object volume 24 of a partial object 23 or at least a portion of the partial object volume 24 can be detected.

[0083] Optionally, a classification and / or assignment of the sub-objects 23 can be performed. For example, a type of bone can be identified as the left elbow, possibly using a known orientation of the patient on the couch (head-first supine). The position of the bone can be estimated using corresponding image features 21 captured in several of the 2D projection images 7, 8. Similarly, the outline of an arm, the body surface, implants, or other organs / objects of the patient can be detected in the truncation volume 25.

[0084] Detection can be supported by the spectral differences in the 2D projection images 7, 8. For example, the segmentation of the sub-objects 23 can be performed using convolutional neural networks, with the corresponding 2D projection images 7, 8 of the different X-ray spectra being used as input values ​​for the convolutional neural networks. This makes it easier for the convolutional neural network to distinguish between sub-objects 23, especially overlapping sub-objects 23.

[0085] For example, an anatomical sub-object 23, such as a bone or another sub-object 23, can be highlighted by a 2D dual-energy offset of the corresponding 2D projection images 7, 8, which corresponds to a projection image 7, 8 of both different X-ray spectra. This sub-object 23 can, in particular, lie outside the primary reconstruction volume 27 due to lateral detector truncation. The dual-energy offset can enable detection and triangulation, or even limited-angle tomography, of sub-objects 23 in the truncation volume 25, as well as material determination or material estimation of a material of these sub-objects 23. This can improve the truncation model for the patient.

[0086] In a step S5, the 2D projection images 7, 8 can first be extrapolated using deep learning approaches as described in the publication by Huang, Yixing, et al., and / or using a standard model of the examination object 5, for example, a simple water cylinder model or a patient model. Subsequently, the supplemented image areas, also called fill projections 15, of the 2D projection images 7, 8 can be compared with the acquired and / or supplemented image areas in the other 2D projection images 7, 8.

[0087] In a step S6, the preliminary volume model 19 can be supplemented by a partial object volume 24 of a detected partial object 23. If, for example, the bone or a bone section or another detected partial object 23 is located at least partially in the truncation volume 25, the position of this bone and / or its outlines in the volume model 11 can be adjusted and / or this partial object volume 24 can be newly added to the volume model 19 if it is not yet present in at least one area of ​​the volume model 19.

[0088] In a step S7, the reconstruction of the primary reconstruction volume 27 and, via the fill projections 15 supplemented in the 2D projection images 7, 8, the preliminary volume model 19 of the examination object 5 can be present in an area relevant for imaging in the regions outside the primary reconstruction volume 27. In the truncated regions, spectral X-ray absorption properties of the at least one volume element 18 of the preliminary volume model 19 can be determined based on the previously performed classifications of the sub-objects 23.

[0089] In a step S8, virtual 2D projection images 20 of the preliminary volume model 19 can be generated. In the first and second X-ray spectra, the defect areas 28 of the respective virtual 2D projection images 20 can be supplemented compared to the 2D projection images 7, 8.

[0090] In a step S9, a separate reconstruction of the volume model 11 can be carried out for both X-ray spectra, taking into account the virtual 2D projection images 20. Alternatively, a joint reconstruction of the volume model 11 can be carried out for both X-ray spectra in order to reconstruct the volume model 11 as a dual energy reconstruction image. The reconstruction can be carried out using the virtual 2D projection images 20 alone. Alternatively, the reconstruction can be carried out using the supplemented 2D projection images 7, 8 and the virtual 2D projection images 20. In this case, for example, the supplemented 2D projection images 7, 8 can be supplemented at least partially with the virtual 2D projection images 20. This can be done, for example, at the edges of the field of view in order to expand the field of view using the virtual 2D projection images 20. This can avoid truncations at the edge of the examination object 5.

[0091] The volume model 11 extracted from the 2D projection images 7, 8 of the dual-energy CBCT procedure can also be used for further correction steps that take material properties into account. These correction steps can include, for example, scatter correction and / or metal artifact correction. These correction steps can also be used during an interventional procedure, for example, when determining a dose distribution on the patient.

Claims

[1] Method for generating a volume model (11) of an object under investigation (5) from a data set (9), wherein - the data set (9) comprises first 2D projection images (7) of an examination object (5), created in a first X-ray spectrum, and second 2D projection images (8) of the examination object (5), created in a second X-ray spectrum, - the 2D projection images (7, 8) each image a projection volume (13) of the examination object (5) which is arranged in a field of view (17) of the respective 2D projection images (7, 8), wherein the method comprises the following steps to be carried out by a control device (10): - receiving the data set (9) from an X-ray device (1), - supplementing the 2D projection images (7, 8), comprising extending at least one truncated 2D projection image of the first 2D projection images (7) and / or the second 2D projection images (8) by a fill projection (15), wherein the fill projection (15) images a fill volume (16) of the examination object (5) arranged outside the field of view (17) of the at least one truncated 2D projection image; - reconstruction of a preliminary volume model (19) of the object under investigation (5) from the supplemented 2D projection images (7, 8); - identifying a material (22) of at least a partial area of ​​one of the 2D projection images (7, 8) based on a pattern of the partial area; - determining a spectral X-ray absorption property of at least one volume element (18) of the preliminary volume model (11) as a function of the material (22) of the partial region, wherein the at least one volume element (18) is arranged within the filling volume (16) of the at least one truncated 2D projection image; - generating virtual 2D projection images (20) of the X-ray spectra of the examination object (5) from the preliminary volume model (19) of the examination object (5) as a function of the spectral X-ray absorption property of the at least one volume element (18); and - Reconstruction of the volume model (11) of the object under investigation (5) from the virtual 2D projection images (20) or from the supplemented 2D projection images (7, 8) and the virtual 2D projection images (20). [2] Method for generating a volume model (11) of an object under investigation (5) from a data set (9), wherein - the data set (9) comprises first 2D projection images (7) of an examination object (5), created in a first X-ray spectrum, and second 2D projection images (8) of the examination object (5), created in a second X-ray spectrum, - the 2D projection images (7, 8) each image a projection volume (13) of the examination object (5) which is arranged in a field of view (17) of the respective 2D projection images (7, 8), wherein the method comprises the following steps to be carried out by a control device (10): - receiving the data set (9) from an X-ray device (1), - supplementing the 2D projection images (7, 8), comprising extending at least one truncated 2D projection image of the first 2D projection images (7) and / or the second 2D projection images (8) by a fill projection (15), wherein the fill projection (15) images a fill volume (16) of the examination object (5) arranged outside the field of view (17) of the at least one truncated 2D projection image; - reconstruction of a preliminary volume model (19) of the object under investigation (5) from the supplemented 2D projection images (7, 8); - determining a spectral X-ray absorption property of at least one volume element (18) of the preliminary volume model (11), depending on the supplemented 2D projection images (7, 8), wherein the at least one volume element (18) is arranged within the filling volume (16) of the at least one truncated 2D projection image, wherein a surface element (26) associated with the volume element (18) is identified in one of the first 2D projection images (7), wherein the surface element (26) associated with the volume element (18) is identified in one of the second 2D projection images (8), wherein the spectral X-ray absorption property of the volume element (18) is determined from projection values ​​of the surface element (26) in the 2D projection images (7, 8) of the two X-ray spectra; - generating virtual 2D projection images (20) of the X-ray spectra of the examination object (5) from the preliminary volume model (19) of the examination object (5) as a function of the spectral X-ray absorption property of the at least one volume element (18); and - Reconstruction of the volume model (11) of the object under investigation (5) from the virtual 2D projection images (20) or from the supplemented 2D projection images (7, 8) and the virtual 2D projection images (20). [3] Method for generating a volume model (11) of an object under investigation (5) from a data set (9), wherein - the data set (9) comprises first 2D projection images (7) of an examination object (5), created in a first X-ray spectrum, and second 2D projection images (8) of the examination object (5), created in a second X-ray spectrum, - the 2D projection images (7, 8) each depict a projection volume (13) of the examination object (5) which is arranged in a field of view (17) of the respective 2D projection images (7, 8), the method comprising the following steps to be carried out by a control device (10): - receiving the data set (9) from an X-ray device (1), - supplementing the 2D projection images (7, 8), comprising extending at least one truncated 2D projection image of the first 2D projection images (7) and / or the second 2D projection images (8) by a fill projection (15), wherein the fill projection (15) images a fill volume (16) of the examination object (5) arranged outside the field of view (17) of the at least one truncated 2D projection image; - reconstruction of a preliminary volume model (19) of the object under investigation (5) from the supplemented 2D projection images (7, 8); - identification of a partial object volume (24) of at least one partial object (23) of the object under investigation (5) in the preliminary volume model (11); - adding a truncated partial object volume (24) as a filling volume (16) in the preliminary volume model (19); - determining a spectral X-ray absorption property of at least one volume element (18) of the preliminary volume model (11), depending on the supplemented 2D projection images (7, 8), wherein the at least one volume element (18) is arranged within the filling volume (16) of the at least one truncated 2D projection image; - generating virtual 2D projection images (20) of the X-ray spectra of the examination object (5) from the preliminary volume model (19) of the examination object (5) as a function of the spectral X-ray absorption property of the at least one volume element (18); and - Reconstruction of the volume model (11) of the object under investigation (5) from the virtual 2D projection images (20) or from the supplemented 2D projection images (7, 8) and the virtual 2D projection images (20). [4] Method according to claim 3, wherein the method comprises the following steps to be performed by the control device (10): - identification of image features (21) of the at least one partial object (23) of the examination object (5) in the 2D projection images (7, 8); - Determining a position of the partial object volume (24) of the at least one partial object (23) in the preliminary volume model (19) of the examination object (5) as a function of positions of the respective image features (21) of the at least one partial object (23) in the 2D projection images (7, 8). [5] The method according to claim 3 or 4, wherein the at least one volume element (18) is arranged within the truncated partial object volume (24); and determining the spectral X-ray absorption property of the at least one volume element (18) comprises retrieving the spectral X-ray absorption property of the partial object (23) from a database. [6] Method according to one of the preceding claims, wherein the volume model (11) of the examination object (5) comprises a single-energy volume model of the examination object (5) for the first X-ray spectrum and a single-energy volume model of the examination object (5) for the second X-ray spectrum, wherein - the single-energy volume model of the examination object (5) for the first X-ray spectrum is reconstructed from the virtual 2D projection images (20) generated in the first X-ray spectrum or from the supplemented first 2D projection images (7); and - the single-energy volume model of the examination object (5) for the second X-ray spectrum is reconstructed from the virtual 2D projection images (20) generated in the second X-ray spectrum or from the supplemented second 2D projection images (7). [7] Method according to one of the preceding claims, wherein the volume model (11) of the examination object (5) comprises a dual-energy volume model of the examination object (5) for both X-ray spectra, wherein the dual-energy volume model of the examination object (5) for both X-ray spectra is reconstructed from the virtual 2D projection images (20) generated in both X-ray spectra, or from the supplemented first 2D projection images (7) and the supplemented second 2D projection images (7). [8] Method according to one of the preceding claims, wherein supplementing the 2D projection images (7, 8) comprises applying a deep learning method. [9] Method according to one of the preceding claims, wherein the supplementing of the 2D projection images (7, 8) comprises providing the fill projection (15) from a predetermined standard model of the examination object (5). [10] Method according to one of the preceding claims, wherein the method comprises the following step to be carried out by the control device (10): - Reconstruction of the preliminary volume model (19) of the object under investigation (5) from the 2D projection images (7, 8). [11] Method according to one of the preceding claims, wherein the method comprises the following steps to be carried out by the X-ray device (1): - capturing the first 2D projection images (7, 8) of the examination object (5) in the first X-ray spectrum; - capturing the second 2D projection images (7, 8) of the examination object (5) in the second X-ray spectrum; and - Providing the data set (9) comprising the first 2D projection images (7) and the second 2D projection images (8) to the control device (10). [12] Method according to claim 11, wherein the 2D projection images (7, 8) are acquired by the X-ray device (1) according to a dual-energy cone beam computed tomography method. [13] Method according to claim 11, wherein the 2D projection images (7, 8) are acquired by the X-ray device (1) according to a limited-angle tomography method. [14] Control device (10) configured to generate a volume model (11) of an examination object (5) from a data set (9), wherein - the data set (9) comprises first 2D projection images (7) of an examination object (5), created in a first X-ray spectrum and second 2D projection images (8) of the examination object (5), created in a second X-ray spectrum, - the 2D projection images (7, 8) each image a projection volume (13) of the examination object (5) which is arranged in a field of view (17) of the respective 2D projection image (7, 8);characterized by that the control device (10) is designed to - to receive the data set (9) from an X-ray device (1), - to supplement the 2D projection images (7, 8), wherein at least one truncated 2D projection image of the first 2D projection images (7) and / or the second 2D projection images (8) is extended by a fill projection (15), wherein the fill projection (15) images a fill volume (16) of the examination object (5) arranged outside the field of view (17) of the at least one truncated 2D projection image; - to reconstruct a preliminary volume model (19) of the object under investigation (5) from the supplemented 2D projection images (7, 8); - to identify a material (22) of at least one partial area of ​​one of the 2D projection images (7, 8) based on a pattern of the partial area; - to determine a spectral X-ray absorption property of at least one volume element (18) arranged within the filling volume (16) of the at least one truncated 2D projection image as a function of the material (22) of the partial region; - to generate virtual 2D projection images (20) of the X-ray spectra of the examination object (5) from the preliminary volume model (19) of the examination object (5) as a function of the spectral X-ray absorption property of the at least one volume element (18); and - to reconstruct the volume model (11) of the object under investigation (5) from the virtual 2D projection images (20) or the supplemented 2D projection images (7, 8) and the virtual 2D projection images (20). [15] Control device (10) configured to generate a volume model (11) of an examination object (5) from a data set (9), wherein - the data set (9) comprises first 2D projection images (7) of an examination object (5), created in a first X-ray spectrum and second 2D projection images (8) of the examination object (5), created in a second X-ray spectrum, - the 2D projection images (7, 8) each image a projection volume (13) of the examination object (5) which is arranged in a field of view (17) of the respective 2D projection image (7, 8); characterized by that the control device (10) is designed to - to receive the data set (9) from an X-ray device (1), - to supplement the 2D projection images (7, 8), wherein at least one truncated 2D projection image of the first 2D projection images (7) and / or the second 2D projection images (8) is extended by a fill projection (15), wherein the fill projection (15) images a fill volume (16) of the examination object (5) arranged outside the field of view (17) of the at least one truncated 2D projection image; - to reconstruct a preliminary volume model (19) of the object under investigation (5) from the supplemented 2D projection images (7, 8); - to determine a spectral X-ray absorption property of at least one volume element (18) arranged within the filling volume (16) of the at least one truncated 2D projection recording as a function of the supplemented 2D projection recordings (7, 8), wherein the control device: - a surface element (26) assigned to the volume element (18) is identified in one of the first 2D projection images (7), - the surface element (26) assigned to the volume element (18) is identified in one of the second 2D projection images (8) and - the spectral X-ray absorption property of the volume element (18) is determined from projection values ​​of the surface element (26) in the 2D projection images (7, 8) of the two X-ray spectra; - to generate virtual 2D projection images (20) of the X-ray spectra of the examination object (5) from the preliminary volume model (19) of the examination object (5) as a function of the spectral X-ray absorption property of the at least one volume element (18); and - to reconstruct the volume model (11) of the object under investigation (5) from the virtual 2D projection images (20) or the supplemented 2D projection images (7, 8) and the virtual 2D projection images (20). [16] Control device (10) configured to generate a volume model (11) of an examination object (5) from a data set (9), wherein - the data set (9) comprises first 2D projection images (7) of an examination object (5), created in a first X-ray spectrum and second 2D projection images (8) of the examination object (5), created in a second X-ray spectrum, - the 2D projection images (7, 8) each image a projection volume (13) of the examination object (5) which is arranged in a field of view (17) of the respective 2D projection image (7, 8); characterized by that the control device (10) is designed to - to receive the data set (9) from an X-ray device (1), - to supplement the 2D projection images (7, 8), wherein at least one truncated 2D projection image of the first 2D projection images (7) and / or the second 2D projection images (8) is extended by a fill projection (15), wherein the fill projection (15) images a fill volume (16) of the examination object (5) arranged outside the field of view (17) of the at least one truncated 2D projection image; - to reconstruct a preliminary volume model (19) of the object under investigation (5) from the supplemented 2D projection images (7, 8); - to identify a partial object volume (24) of at least one partial object (23) of the object under investigation (5) in the preliminary volume model (11); - to add a truncated partial object volume (24) as a filling volume (16) in the preliminary volume model (19); - to determine a spectral X-ray absorption property of at least one volume element (18) arranged within the filling volume (16) of the at least one truncated 2D projection image as a function of the supplemented 2D projection images (7, 8); - to generate virtual 2D projection images (20) of the X-ray spectra of the examination object (5) from the preliminary volume model (19) of the examination object (5) as a function of the spectral X-ray absorption property of the at least one volume element (18); and - to reconstruct the volume model (11) of the object under investigation (5) from the virtual 2D projection images (20) or the supplemented 2D projection images (7, 8) and the virtual 2D projection images (20). [17] X-ray device (1) comprising at least one control device (10) according to one of claims 14 to 16. [18] A computer program which can be loaded directly into a memory of a control device (10) according to any one of claims 14 to 16, comprising program means for carrying out the steps of the method according to any one of claims 1 to 13 when the program is executed in the control device (10). [19] Electronically readable data carrier with electronically readable control information stored thereon, which comprises at least one computer program according to claim 18 and is designed such that, when the data carrier is used in a control device (10) according to one of claims 14 to 16, it carries out a method according to one of claims 1 to 13.

Images