Method for image reconstruction of an object, in particular image reconstruction based on computed tomography, and device, system and computer program product for the method

By reconstructing N-dimensional image data from multiple angles and different input parameters in computed tomography, the problem of noise sensitivity and computational burden of image reconstruction process in the prior art is solved, and the efficiency of image reconstruction is improved and speed.

CN111670460BActive Publication Date: 2024-11-26MITOS GMBH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN201880088369.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2018-01-31
Publication Date
2024-11-26
Estimated Expiration
2038-01-31

AI Technical Summary

Technical Problem

Existing computed tomography techniques are noise-sensitive and have a high computational burden during image reconstruction, especially the iterative reconstruction method requires a large number of parameters, resulting in inefficiency.

Method used

By obtaining multiple (N-1) dimensional images at different angle positions and reconstructing N dimensional image data using an image reconstruction algorithm, preferably by stacking the represented pseudo-objects and different input parameter values ​​to reduce the computational burden and improve efficiency.

Benefits of technology

It achieves the reduction of computational burden and the improvement of image reconstruction while maintaining reliability and effectiveness, and is suitable for medical and non-destructive testing purposes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN111670460B_ABST
    Figure CN111670460B_ABST
Patent Text Reader

Abstract

The invention relates to a method for image reconstruction of an object, in particular based on computed tomography, as well as a device, a system and a computer program product for the method. A plurality of two-dimensional images of an object are obtained at different angular positions relative to a straight line between a light source and a two-dimensional image detector, and three-dimensional image data are reconstructed based on an image reconstruction algorithm, which uses the data of the plurality of two-dimensional images of the object and at least one input parameter as input. In particular, the three-dimensional image data are reconstructed based on a three-dimensional pseudo object represented by a stack of a plurality of slices, wherein each of the plurality of slices of the pseudo object is associated with the same base slice of the object. Furthermore, reconstructing the three-dimensional image data comprises using different values ​​of the at least one input parameter as input to the image reconstruction algorithm for each slice of the pseudo object.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a method for image reconstruction of an object, in particular image reconstruction based on computed tomography, as well as a device, a system and a computer program product for the method.

[0002] Multiple (N-1)-dimensional images of the object are obtained at different angular positions relative to a straight line between a light source and a two-dimensional image detector, and N-dimensional image data is reconstructed based on an image reconstruction algorithm, which uses the data of the multiple (N-1)-dimensional images of the object and at least one input parameter as input, where N=3, N=4 or greater. Background Art

[0003] X-ray computed tomography (CT) is an imaging modality for non-destructive three-dimensional investigations that is very common in medical applications and is increasingly used in non-destructive testing in industry and science.

[0004] The most common nondestructive inspection alternatives are X-ray radiography, light microscopy, ultrasound imaging, and magnetic resonance imaging (MRI).

[0005] Compared to X-ray CT, the latter three methods do not carry the risk of radiation damage. For X-ray radiography, this risk is greatly reduced. In addition, light microscopy, ultrasound and radiography are much cheaper. By far, the greatest advantage of X-ray CT is its most precise spatial information. In particular, these three cheaper methods only allow very limited three-dimensional studies. Compared to MRI, X-ray CT is faster, cheaper and more precise.

[0006] X-ray computed tomography allows the creation of a three-dimensional map of the coefficients that describe how strongly the X-rays interact with the material. This map allows different materials or mass densities to be distinguished. The working principle of CT consists of two basic steps. The first step is data acquisition, in which the desired quantity can only be measured indirectly. The second step is reconstruction on the computer, in which the three-dimensional map of the coefficients is read from the measurement. The grayscale image of a two-dimensional plane represented by volume data is the most common way for people to evaluate this map.

[0007] The basic setup for a tomographic scan consists of an x-ray source placed on one side of the specimen, a 2D detector placed on the other side, and a mechanism to rotate the specimen in the middle perpendicular to the x-ray beam, or vice versa, to rotate both devices around the specimen.

[0008] Detailed information on x-ray sources, the interaction of x-rays with matter, and the different acquisition techniques described can be found in the literature, both scientific and patent.

[0009] As mentioned above, the idea behind computed tomography is to create a three-dimensional map (three-dimensional image data) of the local attenuation coefficients from a set of transmission views through a sample (object). More detailed information on analytical tomographic reconstruction can be found in the literature (e.g., A. K. Ak and M. Slaney, "Principles of computerized tomographic imaging", Classics in applied mathematics, IEEE Press, 1988, and H. Turbell's doctoral thesis "Cone-Beam Reconstruction Using Filtered Back Projection", University of Sweden, 2001).

[0010] Although it is theoretically possible to analytically reconstruct 3D image data, this analytical reconstruction method is very sensitive to noise and requires a large number of 2D images of the object to be measured. Another less sensitive method is called iterative reconstruction, which provides results of greatly improved quality. However, iterative reconstruction methods usually have a high computational burden and additional parameters.

[0011] For patent literature related to iterative reconstruction, refer to, for example, WO 2013 / 088294 A1 and US Pat. No. 9,449,404 B2.

[0012] In view of the above, the object of the present invention is to provide a concept for image reconstruction of an object, in particular a concept for image reconstruction based on computed tomography, and in particular a concept for reconstructing three-dimensional reconstruction image data. In particular, the object of the present invention is to provide a concept for image reconstruction of an object, which concept can be practically applied for medical purposes and for non-destructive testing of objects, and in particular for analytical methods as well as for iterative methods, and which concept is as reliable and effective as known reconstruction algorithms, but which allows to find parameters with less computational burden and / or in a faster way. Summary of the invention

[0013] To achieve the above-mentioned object, according to the invention, a method for image reconstruction of an object, in particular based on computed tomography, a device, a system and a computer program product according to the independent claims are proposed. The dependent claims relate to preferred exemplary embodiments.

[0014] According to an exemplary aspect of an exemplary embodiment, a method for image reconstruction of an object is proposed, in particular a method for image reconstruction based on computed tomography, comprising obtaining multiple (N-1)-dimensional images of the object at different angular positions relative to a straight line between a light source and a two-dimensional image detector or an (N-1)-dimensional image detector, and reconstructing N-dimensional image data based on an image reconstruction algorithm, wherein the image reconstruction algorithm uses the data of the multiple (N-1)-dimensional images of the object and at least one input parameter as input, where N is N=3 or N=4 or greater.

[0015] When N=3, a method for image reconstruction of an object, in particular a method for image reconstruction based on computed tomography, is proposed, comprising obtaining multiple two-dimensional images of the object at different angular positions relative to a straight line between a light source and a two-dimensional image detector, and reconstructing three-dimensional image data based on an image reconstruction algorithm, wherein the image reconstruction algorithm uses the data of the multiple two-dimensional images of the object and at least one input parameter as input.

[0016] On the other hand, in the case of N=4, a method for image reconstruction of an object, in particular a method for image reconstruction based on computed tomography, is proposed, comprising obtaining a plurality of three-dimensional images of the object at different angular positions relative to a straight line between a light source and a two-dimensional image detector or a three-dimensional image detector, and reconstructing four-dimensional image data based on an image reconstruction algorithm, the image reconstruction algorithm using the data of the plurality of three-dimensional images of the object and at least one input parameter as input. In this case, for example in the case of an energy-resolving detector and / or a time-resolving detector, the three-dimensional image of the object can be resolved in one or more dimensions, for example energy and / or time.

[0017] Preferably, in some exemplary embodiments, the N-dimensional image data is reconstructed based on an N-dimensional pseudo object represented by a stack of a plurality of slices. Preferably, each slice of the plurality of slices of the pseudo object is associated with the same underlying slice of the object.

[0018] Preferably, in some exemplary embodiments, reconstructing the N-dimensional image data comprises using a different value of at least one input parameter as input to an image reconstruction algorithm for each slice of the pseudo object.

[0019] Preferably, in some exemplary embodiments, a plurality of (N-1)-dimensional images are obtained from different viewpoints with respect to a rotation axis arranged along the Z direction.

[0020] Preferably, in some exemplary embodiments, the elementary slices of the object are arranged substantially along a two-dimensional XY plane passing through the object perpendicular to the rotation axis.

[0021] Preferably, in some exemplary embodiments, the N-dimensional pseudo object is represented by a stack of a plurality of slices stacked along the Z direction.

[0022] Preferably, in some exemplary embodiments, different planar portions of the N-dimensional image data perpendicular to the stacking direction of the pseudo object represent image reconstructions of the same basic slice of the object at different values ​​of the at least one input parameter of the image reconstruction algorithm.

[0023] Preferably, in some exemplary embodiments, the method further comprises: extracting slice image data from each of the multiple (N-1)-dimensional images obtained, preferably, each slice image data represents one or more adjacent pixel rows or columns of the corresponding obtained (N-1)-dimensional image.

[0024] Preferably, in some exemplary embodiments, the method further includes: preferably, generating a corresponding modified (N-1)-dimensional image for each of the multiple (N-1)-dimensional images obtained by stacking the corresponding extracted slice image data of the corresponding (N-1)-dimensional image multiple times, particularly stacking multiple times in the pixel row direction or the pixel column direction.

[0025] Preferably, in some exemplary embodiments, the N-dimensional image data is reconstructed based on an image reconstruction algorithm which uses the modified data of the plurality of (N-1)-dimensional images as input.

[0026] Preferably, in some exemplary embodiments, the method further comprises selecting a specific value of at least one input parameter based on the reconstructed N-dimensional image data.

[0027] Preferably, in some exemplary embodiments, the method further comprises reconstructing the final N-dimensional image data based on an image reconstruction algorithm, which uses the obtained multiple (N-1)-dimensional images of the object and / or a specific value of the selected at least one input parameter as input.

[0028] Preferably, in some exemplary embodiments, selecting a specific value of at least one input parameter includes: preferably, based on applying a predetermined metric calculation algorithm to the (N-1)-dimensional image data of the corresponding plane portion associated with the corresponding slice of the pseudo object of the N-dimensional image data, calculating a corresponding metric parameter for each slice of the pseudo object, identifying the corresponding slice of the pseudo object whose corresponding metric parameter has its extreme value, and / or identifying the corresponding input value of the at least one input parameter as the specific value of the at least one input parameter selected, and the corresponding input value of the at least one input parameter is used to reconstruct the corresponding plane portion of the N-dimensional image data associated with the identified corresponding slice of the pseudo object.

[0029] Preferably, in some exemplary embodiments, the at least one input parameter represents: a weight parameter, which is used to define the weights of data terms and regularization terms according to an iterative image reconstruction algorithm; a penalty function parameter, which is used in a penalty function for the regularization term according to an iterative method; a center offset parameter, which represents the offset of the actual rotation axis of the measurement system relative to the target rotation axis of the measurement system; a correction parameter, which represents a beam hardening correction; a filtering parameter, which is used in image filtering applied to an (N-1)-dimensional image; and / or a weight factor, which weights different image source energies in dual-energy CT image reconstruction.

[0030] According to a further exemplary aspect of an exemplary embodiment, a device for image reconstruction of an object, in particular a device for image reconstruction based on computed tomography, and in particular a device according to a method of at least one of the above-mentioned aspects is proposed.

[0031] Preferably, the apparatus comprises an interface configured to receive data indicative of a plurality of (N-1)-dimensional images of the object obtained at different angular positions relative to a line between the light source and the two-dimensional image detector.

[0032] Preferably, in some exemplary embodiments, the device comprises an image reconstruction unit configured to reconstruct three-dimensional image data based on an image reconstruction algorithm using data of multiple (N-1)-dimensional images of the object and at least one input parameter as input.

[0033] Preferably, in some exemplary embodiments, the image reconstruction unit reconstructs the N-dimensional image data based on an N-dimensional pseudo object represented by a stack of a plurality of slices, each slice of the plurality of slices of the pseudo object being associated with a same base slice of the object.

[0034] Preferably, in some exemplary embodiments, the image reconstruction unit is configured to reconstruct N-dimensional image data, and reconstructing the N-dimensional image data comprises using a different value of at least one input parameter as an input to an image reconstruction algorithm for each slice of the pseudo object.

[0035] Preferably, in some exemplary embodiments, recursive optimization of the range of the parameter used in the parameter dimension may be provided, or the range of the parameter may be limited to improve accuracy, or the range of the parameter may be expanded or changed to extend the search range to obtain the optimal value.

[0036] According to another exemplary aspect of an exemplary embodiment, a system for image reconstruction of an object, in particular a system for image reconstruction based on computed tomography, is proposed, comprising: a light source (preferably an X-ray light source), a two-dimensional image detector, and / or a device according to at least one of the above aspects.

[0037] According to a further exemplary aspect of the exemplary embodiments, a computer program product is proposed, comprising a computer program comprising computer program instructions adapted to cause a computer, a processor or a computed tomography controller to perform the steps of the method of at least one of the above aspects. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A schematic diagram of a measurement system is shown exemplarily;

[0039] Figure 2 It is shown by way of example that Figure 1 In the measurement system shown, multiple two-dimensional images of an object are obtained from different viewing angles at different angular positions;

[0040] Figure 3 The concept of iterative reconstruction of three-dimensional image data of an object based on computed tomography is exemplarily shown;

[0041] Figure 4 The process of iterative reconstruction of three-dimensional image data based on computer tomography is exemplarily shown;

[0042] Figure 5 The concept of using a pseudo volume generated based on a two-dimensional image of an object according to an exemplary embodiment is exemplarily illustrated;

[0043] Figure 6 The concept of image reconstruction of three-dimensional image data of an object based on computed tomography according to an exemplary embodiment is exemplarily illustrated;

[0044] Figure 7 The process of image reconstruction of three-dimensional image data of an object based on computer tomography according to an exemplary embodiment is exemplarily shown;

[0045] Figure 8 A process of image reconstruction of three-dimensional image data of an object based on computer tomography according to another exemplary embodiment is exemplarily shown; and

[0046] Fig. 9 A schematic diagram of a measuring system according to an exemplary embodiment is shown exemplarily, which comprises a device for image reconstruction of three-dimensional image data of an object based on computed tomography. DETAILED DESCRIPTION

[0047] The preferred aspects and embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. The same or similar features in different drawings and embodiments are represented by similar reference numerals. It should be understood that the following detailed description of various preferred aspects and preferred exemplary embodiments is not meant to limit the scope of the present invention.

[0048] Figure 1 The schematic diagram of the measuring system is shown exemplarily. The measuring system exemplarily comprises a light source S (especially an X-ray source) and a two-dimensional image detector D arranged opposite to the light source S.

[0049] Exemplarily, an object P to be measured is arranged between the light source S and the detector D and on a straight line L extending between the light source S and the detector D.

[0050] Exemplarily, the light source S is configured to emit light, in particular X-ray light, in a cone-shaped beam geometry (cone beam geometry) towards the detector D. That is, the light beam CB emitted from the light source S exemplarily has a cone shape ("cone beam"), and exemplarily, an area of ​​the light beam CB (e.g. a quasi-rectangular area) hits the planar detector surface of the detector D.

[0051] However, the present invention is not limited to the use of a light source S that emits light toward the detector D in a cone-beam beam geometry. In other exemplary embodiments, light sources with other beam geometries may also be used, for example, a light source that emits light (especially X-ray light) toward the detector D in a parallel beam or fan beam shape (parallel beam geometry or fan beam geometry).

[0052] exist Figure 1 In the figure, the detector D is exemplarily shown as having a planar detection surface, however, the present invention is not limited to the use of a detector having a planar detection surface, and in an exemplary embodiment of the present invention, a detector having a curved detection surface may also be used.

[0053] This principle construction of a measuring system based on computed tomography is known in the prior art and is used in many applications of computed tomography, including medical applications and non-destructive testing of materials and physical objects, in particular non-biological objects.

[0054] For computer tomography-based measurements, multiple two-dimensional images of the object to be measured are usually acquired from different viewpoints (viewing directions) at different angular positions of the object. This can be achieved by rotating the object between a stationary light source and detector assembly, or by keeping the object stationary and rotating the light source and detector assembly around a rotation axis extending through the object.

[0055] It should be noted that the present invention is not limited to any one of the above. In an exemplary embodiment of the present invention, a two-dimensional image of the object can be obtained from different viewpoints (observation directions) at different angular positions of the object by rotating the object between a stationary light source and a detector assembly, and / or by moving the light source and the detector assembly along any other trajectory around P.

[0056] Figure 2An example is shown in Figure 1 In the measurement system shown, a plurality of two-dimensional images of an object P are obtained from different viewpoints (observation directions) at different angular positions, which exemplarily shows an example in which the object P remains stationary and the light source S and detector D assembly rotate around a rotation axis R extending through the object S. However, as described above, the present invention is not limited to this configuration, and additionally or alternatively, a plurality of two-dimensional images may also be obtained by rotating or moving the object P.

[0057] Figure 2 Three different rotational angular positions of the light source S and detector D assembly relative to an object P arranged on a straight line between the light source S and the detector D are exemplarily shown. In each position, the light source S is activated to emit a light beam CB through the object P toward the detector D, and the detector D detects two-dimensional image data associated with the corresponding angular position.

[0058] Although Figure 2 Three different angular positions of the light source S and the detector D assembly relative to the object P are exemplarily shown, but the two-dimensional image of the object P can be captured / detected at more than three different angular positions of the light source S and the detector D assembly, preferably at the same angular distance. That is, for example, the two-dimensional image of the object P can be captured / detected at an angular distance of k / M*360°, where M is an integer of 2 or greater, and k={1, 2, ..., M}.

[0059] Figure 3 The well-known concept of iterative reconstruction of three-dimensional image data of an object based on computed tomography is shown by way of example.

[0060] As mentioned above, first, a plurality of two-dimensional images are captured / detected at more than three different relative angular positions of the light source S and detector D assembly relative to the object to be measured P. For example, six two-dimensional images can be obtained at an angular interval of k / 6*360° (where k={1, 2, 3, 4, 5, 6}), and these six two-dimensional images are exemplarily referred to as I1, I2, I3, I4, I5 and I6.

[0061] The two-dimensional images I1, I2, I3, I4, I5 and I6 are respectively provided as image data of m×n pixels, where m is the width of the corresponding two-dimensional image and n is the height of the corresponding two-dimensional image. Figure 3 In the example, 5×5 pixels of image data are provided (ie, 25 pixels per image, for example), and each pixel is represented by Figure 3 The cube representation in , wherein, in actual embodiments, in order to obtain reconstructed three-dimensional image data with good resolution, the width and height of the image may be much higher.

[0062] Then, the plurality of two-dimensional images I1, I2, I3, I4, I5 and I6 are input to a three-dimensional image data reconstruction algorithm, for example, provided for iterative reconstruction, together with initial values ​​of one or more input parameters of the three-dimensional image data reconstruction algorithm.

[0063] For example, specifically with respect to the iterative reconstruction algorithm, the one or more input parameters may include at least one of the following:

[0064] A weight parameter, which is used to define the weights of the data term and the regularization term according to the iterative image reconstruction algorithm (commonly referred to as lambda or λ); and / or

[0065] • Penalty function parameters, which are used in the penalty function according to the iterative image reconstruction algorithm for the regularization term, for example for correcting the grayscale (commonly called gamma or γ).

[0066] Furthermore, as an example, additionally or alternatively, the one or more input parameters may include at least one of the following:

[0067] a center offset parameter, which represents the offset of the actual rotation axis of the measurement system relative to the target rotation axis of the measurement system;

[0068] Correction parameters, which represent beam hardening corrections;

[0069] Artifact correction parameters, which are used to correct image artifacts;

[0070] Filter parameters used in image filtering applied to the two-dimensional image; and / or

[0071] • Weight factors, which weight different image source energies in Dual-Energy-CT image reconstruction.

[0072] Then, based on the input two-dimensional images, such as image data of I1, I2, I3, I4, I5 and I6, and further based on the input initial values ​​of one or more input parameters used in the three-dimensional image data reconstruction algorithm, a three-dimensional image data reconstruction algorithm is executed to output reconstructed three-dimensional image data V of a reconstructed volume (including the volume of the object to be measured / reconstructed).

[0073] The output reconstructed three-dimensional image data V includes a plurality of voxels. Figure 3 In the example, the output reconstructed three-dimensional image data V includes 5×5×5 voxels (ie, 125 voxels, for example), each of which is composed of Figure 3 , wherein, in practical embodiments, the number of voxels may be much higher in order to obtain reconstructed three-dimensional image data with good resolution.

[0074] like Figure 3As exemplarily shown in , the three-dimensional voxel space of the reconstructed three-dimensional image data corresponds to three spatial dimensions X, Y and Z. That is, the reconstructed three-dimensional image data represents a reconstructed volume in the canonical space of the three spatial dimensions X, Y and Z.

[0075] Then, the characteristics of the obtained output reconstructed three-dimensional image data V (such as noise, blurriness, sharpness) can be analyzed, and if the output reconstructed three-dimensional image does not meet the requirements or appears to need to be optimized, one or more values ​​of one or more input parameters are adjusted, and then based on the input two-dimensional image, such as I1, I2, I3, I4, I5 and I6, and further based on the appropriate one or more input values ​​of one or more input parameters used in the three-dimensional image data reconstruction algorithm, the three-dimensional image reconstruction algorithm is executed again to output another reconstructed three-dimensional image data V of the reconstructed volume (including the volume of the object to be measured / reconstructed).

[0076] This process can be repeated iteratively until the obtained output reconstructed three-dimensional image data V meets the requirements.

[0077] Figure 4 The reconstruction process of three-dimensional image data based on computed tomography is shown by way of example.

[0078] In step S401 , a plurality of two-dimensional images of the object P to be measured are obtained at different angular positions of the light source S and the detector D assembly relative to the object P to be measured.

[0079] In step S402, the obtained multiple two-dimensional images of the object P to be measured are input as input image data into a three-dimensional image data reconstruction algorithm.

[0080] In step S403, an initial value (or multiple initial values) of one or more input parameters may be selected (eg, based on manual input by an operator or based on one or more pre-stored default parameter values).

[0081] In step S404, an initial value (or multiple initial values) of the selected one or more input parameters is input into the three-dimensional image data reconstruction algorithm, and in step S405, the three-dimensional image data reconstruction algorithm is executed based on the input initial value (or multiple initial values) of the one or more input parameters and the input multiple two-dimensional images of the object P to be measured.

[0082] In step S406, reconstructed three-dimensional image data is output from the executed three-dimensional image data reconstruction algorithm, and in step S407, the image characteristics or features of the output reconstructed three-dimensional image data are analyzed, such as the presence of noise, the degree of blur, the sharpness, etc.

[0083] In step S408, it is determined whether the output reconstructed 3D image data satisfies one or more analysis criteria, and if step S408 returns "YES", then in step S410, the reconstructed 3D image data is determined to represent the final reconstructed 3D image data of the object P, and the process ends.

[0084] However, if step S408 returns "NO", then in step S409, other values ​​of one or more parameters are selected for repeating the process. The process then returns to step S404, and steps S404 to S408 are repeated until step S408 returns "YES".

[0085] Specifically, in the next iteration, one or more other values ​​of the one or more parameters selected in step S409 are input into the three-dimensional image data reconstruction algorithm in step S404, and step S405 executes the three-dimensional image data reconstruction algorithm based on a newly input value (or multiple newly input values) of the one or more input parameters for the next iteration and the input multiple two-dimensional images of the object P to be measured.

[0086] For more information on iterative reconstruction, see WO 2013 / 088294 A1 and US 9,449,404 B2.

[0087] Figure 5 The concept of using a pseudo volume generated based on a two-dimensional image of an object according to an exemplary embodiment is exemplarily illustrated.

[0088] In contrast to the above-mentioned conventional method based on iterative reconstruction, an exemplary embodiment of the present invention does not initially use the complete two-dimensional information of multiple two-dimensional images of object P, but only uses that information of multiple two-dimensional images related to a certain slice (roughly a two-dimensional plane slice) (hereinafter referred to as a base slice BS) passing through object P.

[0089] For example, in the case of a light source emitting a beam with a cone-beam geometry, such as Figure 5 As shown, the central slice of the object P is selected as the base slice BS.

[0090] Specifically, in the case of a cone-beam geometry, preferably, a central slice of the object P is represented by a planar slice through the object P, the planar slice being represented by a plane including the center line of the cone-beam geometry (exemplarily, with Figure 5 The plane is defined by a plane (coinciding with the straight line L between the light source S and the detector D in the measuring system), which plane is perpendicular to the rotation axis R of the measuring system. That is, in the case of a cone-beam geometry, the center line (cone axis) of the cone-beam geometry of the light beam emitted from the light source S is located within a plane slice through the object P, and the plane slice is arranged perpendicular to the rotation axis R of the measuring system.

[0091] Preferably, the pixel rows of the pixel sensor of detector D are arranged so that a plane slice representing a central slice passing through object P intersects the detector surface of detector D parallel to the pixel rows of the pixel sensor, i.e., preferably, the pixel column direction of detector D is arranged parallel to the rotation axis R of the measuring system.

[0092] In the case of a cone-beam geometry, the height of the base slice BS through the object P is preferably such that a horizontally oriented projection DS of the base slice BS onto the detector surface of the detector D is approximately one pixel height.

[0093] Preferably, in the case of a cone-beam geometry, the projection DS of the base slice BS onto the detector surface of the detector D corresponds to one pixel row.

[0094] On the other hand, preferably if a light source S emitting light according to a parallel beam geometry is used, the height of the base slice BS through the object P is such that a horizontally oriented projection DS of the base slice BS on the detector surface of the detector D is approximately one or more pixel heights.

[0095] Preferably, in the case of a parallel beam geometry, the projection DS of the base slice BS onto the detector surface of the detector D corresponds to one or more adjacent pixel rows.

[0096] Furthermore, in the case of a parallel beam geometry, preferably, a central slice of the object P is represented by a plane slice through the object P, the plane slice being defined by a plane including the beam lines of the parallel beam geometry and being perpendicular to the rotation axis R of the measuring system. That is, in the case of a parallel beam geometry, the beam lines of the parallel beam geometry of the light beam emitted from the light source S are located within the plane slice through the object P, and the plane slice is arranged perpendicular to the rotation axis R of the measuring system.

[0097] In summary, in the exemplary embodiment, preferably independently of the beam geometry, the planar slices of the base slices BS through the object P are arranged perpendicular to the rotation axis R of the measuring system.

[0098] from Figure 5 As can be seen from the middle part of FIG. , a base slice BS of the object P is determined, which represents a slice volume of a part of the object P along a plane slice, which is arranged perpendicular to the rotation axis R of the measuring system.

[0099] According to an exemplary embodiment of the present invention, based on the determined base slice BS, a pseudo volume PV is generated by stacking the same slice volume of the base slice BS along a stacking direction, which is represented by the direction of the rotation axis R of the measuring system (e.g., Figure 5in the Z direction).

[0100] Thus, when generating a pseudo-volume PV (which still has the general spatial directions X and Y), the third spatial dimension is reduced, since each slice volume of the slice stack of the pseudo-volume PV represents the same slice volume of the base slice BS of the object P, and movement in the stacking direction no longer represents movement in the Z direction (compared to Figure 3 on the contrary).

[0101] As a basic idea and main aspect of the exemplary embodiment, the dimension of the stacking direction in the pseudo-volume PV is used to apply different values ​​of the input parameters of the image reconstruction algorithm.

[0102] That is, the spatial dimension corresponding to the Z direction, instead, the third (exemplarily, vertical) directional dimension of the pseudo volume PV is used as an input parameter (in Figure 5 In the example, it is called a dimension in the parameter space of the input parameter a), which is Figure 5 , where the three-dimensional coordinate space of the pseudo volume PV includes an X-axis and a Y-axis as two-dimensional representations of the spatial dimensions X and Y, and a third coordinate axis is labeled with input parameter "a" of the one-dimensional parameter space of the input parameter a.

[0103] Likewise, the input parameter "a" may exemplarily represent at least one of the following:

[0104] A weight parameter, which is used to define the weights of the data term and the regularization term according to the iterative image reconstruction algorithm (commonly referred to as lambda or λ); and / or

[0105] • Penalty function parameters, which are used in the penalty function according to the iterative image reconstruction algorithm for the regularization term, for example for correcting the grayscale (commonly called gamma or γ).

[0106] Furthermore, additionally or alternatively, the input parameter "a" may exemplarily represent at least one of the following:

[0107] a center offset parameter, which represents the offset of the actual rotation axis of the measurement system relative to the target rotation axis of the measurement system;

[0108] Correction parameters, which represent beam hardening corrections;

[0109] Artifact correction parameters, which are used to correct image artifacts;

[0110] Filter parameters used in image filtering applied to the two-dimensional image; and / or

[0111] • Weight factors, which weight different image source energies in Dual-Energy-CT image reconstruction.

[0112] Thus, a parameter space of a single input parameter can be used as the domain for the third dimension of the pseudo-volume PV, but because the space of the pseudo-volume PV is still three-dimensional (2 spatial dimensions, e.g., X and Y, and one parameter space dimension for the input parameter a), three-dimensional image reconstruction algorithms optimized for creating three-dimensional image data can still be reliably and efficiently used.

[0113] At the same time, when executing the 3D image reconstruction algorithm, the effects of multiple different values ​​of one or more input parameters of the 3D image reconstruction algorithm can be tested, wherein there is no need to combine, for example Figure 3 and Figure 4 The illustrated iterative execution of the 3D image reconstruction algorithm multiple times can advantageously and efficiently test the effects of multiple different values ​​of one or more input parameters of the 3D image reconstruction algorithm in one or at least only a limited number of executions of the 3D image reconstruction algorithm. As a result, in exemplary embodiments, the computational burden and computational time can be significantly reduced compared to the commonly known iterative reconstruction process, while utilizing a similar reliable and efficient basic 3D image reconstruction algorithm.

[0114] Figure 6 The concept of image reconstruction of three-dimensional image data of an object based on computed tomography according to an exemplary embodiment is exemplarily shown.

[0115] For example, although multiple two-dimensional images, such as images I1, I2, I3, I4, I5 and I6, are also obtained at multiple different relative angular positions of the light source S and detector D components relative to the object to be measured P, only limited data of the multiple two-dimensional images are used (for example, based on the projection DS of the detector surface of the detector D according to the basic slice BS through the object).

[0116] For example, in Figure 6 In the present invention, it is assumed that the projection DS of a base slice BS through the object onto the detector surface of the detector D results in the extraction of a row of pixels from each image, thereby obtaining image data portions IS1, IS2, IS3, IS4, IS5 and IS6 which (exemplarily) contain data of only one row of pixels (e.g. in the case of a cone beam geometry), and the execution of the three-dimensional image reconstruction algorithm is based on the extracted image data portions IS1, IS2, IS3, IS4, IS5 and IS6 instead of the complete image data of the images I1, I2, I3, I4, I5 and I6.

[0117] Herein, in some exemplary embodiments, the three-dimensional image reconstruction algorithm may be configured to receive complete image data of images I1, I2, I3, I4, I5 and I6, and may be further configured to extract information of each image data part IS1, IS2, IS3, I4, I5 from the image data of images I1, I2, I3, IS4, IS5 during execution of the three-dimensional image reconstruction algorithm, for example, based on the determined position of the base slice BS, based on a predetermined pixel row position, and / or based on a predetermined data extraction algorithm.

[0118] In other exemplary embodiments, the modified image may be input into a three-dimensional image reconstruction algorithm, wherein modified image data may be created based on the image data of images I1, I2, I3, I4, I5, and I6. For example, for each of images I1, I2, I3, I4, I5, and I6, a corresponding number of one or more pixel rows may be read, and new image data may be gathered by stacking the same read pixel rows in the pixel column direction.

[0119] For example, also in the case of 5×5 pixels in each image, the pixel value of image I_i (i={1, 2, 3, 4, 5, 6}) is p_ijk, where pixel row j (j={1, 2, 3, 4, 5}), pixel column k (k={1, 2, 3, 4, 5}), if the pixel row to be extracted is, for example, j=3, then a corresponding modified image J_i (i={1, 2, 3, 4, 5, 6}) can be created, where q_ijk=p_i3k, that is, the third pixel row of the corresponding unmodified image I_i is repeated in each pixel row.

[0120] Then, in some exemplary embodiments, the modified image data of the modified image may be used as input data for a three-dimensional image reconstruction algorithm.

[0121] Furthermore, instead of inputting only a single value of an input parameter of the 3D image reconstruction algorithm as in the above-described iterative reconstruction, multiple values ​​of the input parameter of the 3D image reconstruction algorithm may be input to the 3D image reconstruction algorithm.

[0122] For example, the value of the input parameter may be input as a vector (the dimension of the vector is equal to the number of slices of the pseudo volume PV), each vector entry corresponding to another value of the same input parameter of the three-dimensional image reconstruction algorithm.

[0123] Alternatively, for example, the values ​​of the input parameters can be input as a matrix (the matrix dimension is equal to the number of slices of the pseudo-volume PV multiplied by the number of slices of the pseudo-volume PV), which is a diagonal matrix and each diagonal matrix entry corresponds to another value of the same input parameter of the three-dimensional image reconstruction algorithm.

[0124] Furthermore, exemplarily, the inputs may be only maximum and minimum values, wherein the algorithm interpolates between the maximum and minimum values ​​of the inputs to determine interpolated different parameter values ​​of the same input parameter of the three-dimensional image reconstruction algorithm.

[0125] According to a basic idea of ​​an exemplary embodiment of the present invention, a 3D image reconstruction algorithm uses different values ​​of the same input parameters for respective reconstruction of each slice of a pseudo volume (the same basic slices respectively representing the same slice volume in 3D space).

[0126] Therefore, instead of having to iteratively repeat the execution of the 3D image reconstruction algorithm for different values ​​of one or more input parameters, the effect of multiple different values ​​of one or more input parameter values ​​of the 3D image reconstruction algorithm can be tested in one or at least only a limited number of executions of the 3D image reconstruction algorithm. As a result, in exemplary embodiments, the computational burden and computational time can be significantly reduced compared to commonly known iterative reconstruction processes, while utilizing a similar reliable and efficient underlying 3D image reconstruction algorithm.

[0127] Then, based on the analysis of the slices of the output reconstructed 3D image data, for example, by observing the image characteristics of the slices of the output reconstructed 3D image data, such as noise, blur and / or sharpness, the best fit optimization values ​​of the input parameters can be determined. This can be done manually by an operator, for example, by visually observing the slices of the output reconstructed 3D image data, or this can also be done automatically by analyzing different slices of the output reconstructed 3D image data based on a predetermined metric calculated for each slice of the output reconstructed 3D image data, and then selecting the slice of the output reconstructed 3D image data that optimizes the predetermined metric (for example, as an extreme value of the predetermined metric).

[0128] Through such a process, optimized values ​​of input parameters can be selected or identified, and then the originally acquired image can be used as image input, and the 3D image reconstruction algorithm can be repeated using the selected / identified input parameter values ​​to generate reconstructed 3D image data of the actual object.

[0129] Figure 7 A process of image reconstruction of three-dimensional image data of an object based on computed tomography according to an exemplary embodiment is exemplarily shown.

[0130] In step S701 (similar to the above step S401 ), a plurality of two-dimensional images of the object P to be measured are obtained at different angular positions of the light source S and detector D assembly relative to the object P to be measured.

[0131] In step S702, a base slice BS of the object P is determined or identified, e.g. as described above. The base slice BS defines a planar slice volume of the object P in order to determine a projection DS of the base slice BS on the detector surface of the detector D. In some exemplary embodiments, the area of ​​the projection DS of the base slice BS on the detector surface of the detector D (e.g., one or more pixel rows, as described above) can be predetermined depending on the characteristics and construction of the measuring system, particularly preferably based on the beam shape geometry of the light source S and / or based on the geometric arrangement of the light source S relative to the detector D.

[0132] In step S703, a data portion corresponding to the determined or identified base slice BS is extracted from the two-dimensional image obtained in step S701. For example, pixel values ​​may be extracted based on a projection DS of the base slice BS on a detector surface of the detector D.

[0133] In step S704, the extracted data portion is input into a three-dimensional image reconstruction algorithm (e.g., as a modified image or as an extracted data portion as discussed above), wherein the extracted data portion of the two-dimensional image relates to the same base slice BS through the object P.

[0134] In step S705 , a set of values ​​for one or more input parameters of a three-dimensional image reconstruction algorithm is determined, each value being associated with another slice of the pseudo-volume PV.

[0135] In step S706 , a three-dimensional image reconstruction algorithm is executed based on the extracted data portion of the input obtained two-dimensional image and the determined value set of one or more input parameters of the three-dimensional image reconstruction algorithm.

[0136] In step S707, reconstructed three-dimensional image data of the pseudo volume PV is output, wherein each slice of the pseudo volume is reconstructed by the same extracted data portion of the obtained two-dimensional image corresponding to the same base slice BS of the object P, but wherein the corresponding values ​​of the value set of one or more input parameters of the three-dimensional image reconstruction algorithm are different.

[0137] Therefore, as described above, the third dimension of the output three-dimensional image data of the pseudo-volume PV does not represent a spatial dimension in the measurement system space, but rather represents a 1-dimensional axis of the parameter space of the input parameters of the three-dimensional image reconstruction algorithm. It should be noted that the underlying three-dimensional image reconstruction algorithm may be a true cone-beam or parallel-beam three-dimensional image reconstruction algorithm, and / or may include the execution of multiple two-dimensional fan-beam reconstruction algorithms, respectively executed for each slice of the pseudo-volume PV, preferably in parallel, wherein each two-dimensional fan-beam reconstruction algorithm takes as input another value of the input parameter from the input / determined value set.

[0138] In step S708 , the output three-dimensional image data of the pseudo volume PV is analyzed to identify slices of the three-dimensional image data of the pseudo volume PV that meet the analysis criteria.

[0139] For example, based on the analysis of the slices of the output reconstructed 3D image data, the best fit optimization values ​​of the input parameters can be determined, for example, by observing the image characteristics of the slices of the output reconstructed 3D image data, such as noise, blur and / or sharpness. This can also be done automatically by analyzing different slices of the output reconstructed 3D image data based on a predetermined metric calculated for each slice of the output reconstructed 3D image data, and then selecting the slice of the output reconstructed 3D data that optimizes the predetermined metric (e.g., as an extreme value of the predetermined metric).

[0140] Furthermore, additionally or alternatively, this may be done manually by an operator, for example by visually observing slices of the output reconstructed three-dimensional image data and selecting visually satisfactory image slices.

[0141] In step S709 , respective corresponding values ​​of the input parameters associated with the identified slices of the three-dimensional image data of the pseudo-volume PV that meet the analysis criteria are determined.

[0142] Then, similar to steps S402, S404 and S405, the process includes step S710: inputting the (initially) obtained multiple two-dimensional images of object P as image data input into the three-dimensional image reconstruction algorithm, inputting the values ​​of the input parameters determined in step S709 into the three-dimensional image reconstruction algorithm, and step S712: executing the three-dimensional image reconstruction algorithm based on the (initially) obtained multiple two-dimensional images of object P and the input values ​​of the input parameters determined in step S709.

[0143] Then, as an output of the three-dimensional image reconstruction algorithm executed in step S712 , in step S713 , the reconstructed three-dimensional image data of the object P is output as final reconstructed three-dimensional data of the object P, and the process ends.

[0144] Figure 8 A process for image reconstruction of three-dimensional image data of an object based on computed tomography according to a further exemplary embodiment is exemplarily shown.

[0145] In step S801 (similar to the above step S401 or S701 ), a plurality of two-dimensional images of the object P to be measured are obtained at different angular positions of the light source S and the detector D assembly relative to the object P to be measured.

[0146] In step S802 (similar to step S702 described above), a base slice BS of the object P is determined or identified, e.g. as described above. The base slice BS defines a planar slice volume of the object P in order to determine a projection DS of the base slice BS on the detector surface of the detector D. In some exemplary embodiments, the area of ​​the projection DS of the base slice BS on the detector surface of the detector D (e.g., one or more pixel rows, as described above) can be predetermined depending on the characteristics and configuration of the measuring system, particularly preferably based on the beam shape geometry of the light source S and / or based on the geometric arrangement of the light source S relative to the detector D.

[0147] In step S803 (similar to step S402 above), the two-dimensional image obtained in step S801 is input as input image data into a three-dimensional image data reconstruction algorithm.

[0148] In step S804 (similar to step S705 above), a set of values ​​for one or more input parameters of a three-dimensional image reconstruction algorithm is determined, each value being associated with another slice of the pseudo-volume PV.

[0149] In step S805 (similar to step S706 above), a three-dimensional image reconstruction algorithm is performed based on the extracted data portion of the obtained two-dimensional image and the determined value set of one or more input parameters of the three-dimensional image reconstruction algorithm.

[0150] However, it is assumed herein that the 3D image reconstruction algorithm itself is configured to automatically extract data slices corresponding to the determined or identified base slice BS from the input acquired 2D image. For example, pixel values ​​can be extracted based on the projection DS of the base slice BS on the detector surface of the detector D.

[0151] In step S806 (similar to step S707 above), each slice of the pseudo volume is reconstructed by obtaining the same extracted data portion of the two-dimensional image corresponding to the same base slice BS of the object P, but with different corresponding values ​​of the value set of one or more input parameters of the three-dimensional image reconstruction algorithm.

[0152] Therefore, as described above, the third dimension of the output three-dimensional image data of the pseudo-volume PV does not represent a spatial dimension in the measurement system space, but rather represents a 1-dimensional axis of the parameter space of the input parameters of the three-dimensional image reconstruction algorithm. It should be noted that the underlying three-dimensional image reconstruction algorithm may be a true cone-beam or parallel-beam three-dimensional image reconstruction algorithm, and / or may include the execution of multiple two-dimensional fan-beam reconstruction algorithms, respectively, for each slice of the pseudo-volume PV, preferably in parallel, wherein each two-dimensional fan-beam reconstruction algorithm takes as input another value of the input parameter from the input / determined value set.

[0153] In step S807 (similar to step S708 above), the output three-dimensional image data of the pseudo-volume PV is analyzed to identify slices of the three-dimensional image data of the pseudo-volume PV that meet the analysis criteria.

[0154] For example, based on the analysis of the slices of the output reconstructed 3D image data, the best fit optimization values ​​of the input parameters can be determined, for example, by observing the image characteristics of the slices of the output reconstructed 3D image data, such as noise, blur and / or sharpness. This can also be done automatically by analyzing different slices of the output reconstructed 3D image data based on a predetermined metric calculated for each slice of the output reconstructed 3D image data, and then selecting the slice of the output reconstructed 3D data that optimizes the predetermined metric (e.g., as an extreme value of the predetermined metric).

[0155] Furthermore, additionally or alternatively, this may be done manually by an operator, for example by visually observing slices of the output reconstructed three-dimensional image data and selecting visually satisfactory image slices.

[0156] In step S808 (similar to step S709 above), respective corresponding values ​​of the input parameters associated with the identified slices of the three-dimensional image data of the pseudo-volume PV that meet the analysis criteria are determined.

[0157] Then, similar to the above-mentioned steps S711 and S712, the present process includes step S809: inputting the values ​​of the input parameters determined in step S808 into the three-dimensional image reconstruction algorithm, and step S810: executing the three-dimensional image reconstruction algorithm based on the (initially) obtained multiple two-dimensional images of the object P and the input values ​​of the input parameters determined in step S808.

[0158] Then, as an output of the three-dimensional image reconstruction algorithm executed in step S810 , in step S811 , reconstructed three-dimensional image data of the object P is output as final reconstructed three-dimensional data of the object P, and the process ends.

[0159] Fig. 9 A schematic diagram of a measuring system according to an exemplary embodiment is shown exemplarily, which comprises a device for image reconstruction of three-dimensional image data of an object based on computed tomography.

[0160] Similar to Figure 1 The measuring system comprises a light source S, in particular an X-ray source, and a two-dimensional image detector D arranged opposite to the light source S. Exemplarily, an object P to be measured is arranged between the light source S and the detector D and on a straight line L extending between the light source S and the detector D.

[0161] Exemplarily, the light source S is configured to emit light, in particular X-ray light, in a cone-shaped beam geometry (cone beam geometry) towards the detector D. That is, the light beam CB emitted from the light source S exemplarily has a conical shape ("cone beam"), and exemplarily a circular area of ​​the light beam CB hits the planar detector surface of the detector D.

[0162] However, the present invention is not limited to the use of a light source S that emits light in a cone-beam beam geometry toward the detector D. In other exemplary embodiments, light sources with other beam geometries may also be used, for example, a light source that emits light (particularly X-rays) in a parallel beam (parallel beam geometry) toward the detector D.

[0163] exist Fig. 9 In the figure, the detector D is exemplarily shown as having a planar detection surface, however, the present invention is not limited to the use of a detector having a planar detection surface, and in an exemplary embodiment of the present invention, a detector having a curved detection surface may also be used.

[0164] As mentioned above, for computer tomography-based measurements, multiple two-dimensional images of the object to be measured are usually acquired from different viewpoints (viewing directions) at different angular positions of the object. This can be achieved by rotating the object between a stationary light source and detector assembly, or by keeping the object stationary and rotating the light source and detector assembly around a rotation axis extending through the object.

[0165] It should be noted that the present invention is not limited to any of the above, and in exemplary embodiments of the present invention, a two-dimensional image of an object at different angular positions of the object can be obtained by rotating the object between a stationary light source and a detector assembly, and / or by rotating the light source and the detector assembly around a rotation axis extending through the object.

[0166] Furthermore, the measuring system comprises an apparatus 100 for image reconstruction of an object, in particular based on one, more or all of the above aspects. The apparatus 100, which can be implemented in a computer or a computer-implemented CT controller, is exemplarily connected to the detector D in communication, for example via a wired or wireless communication connection, via a detector communication interface 110, to receive image data from the detector D.

[0167] The apparatus 100 further includes a two-dimensional image data storage section 120 configured to store two-dimensional image data such as a plurality of acquired two-dimensional images received from the detector D through the detector communication interface 110 .

[0168] Furthermore, the apparatus 100 comprises an image reconstruction algorithm data store 130 configured to store computer program instructions for executing one or more predetermined image reconstruction algorithms, such as one or more three-dimensional image reconstruction algorithms discussed above.

[0169] Furthermore, the device 100 comprises an input parameter data storage 140 configured to store predetermined values ​​and / or value sets of one or more input parameters of one or more image reconstruction algorithms stored in the image reconstruction algorithm data storage 130 .

[0170] The device 100 also exemplarily includes a processing unit 150 (e.g., including one or more processors, central processing units, graphics processing units and / or programmable logic circuits), which is configured to execute one or more image reconstruction algorithms stored in the image reconstruction algorithm data storage unit 130 based on the image data stored in the two-dimensional image data storage unit 120 and / or the parameter data stored in the input parameter data storage unit 140.

[0171] The device 100 also exemplarily includes a three-dimensional image data storage unit 160 configured to store reconstructed three-dimensional image data output from the execution of one or more image reconstruction algorithms executed by the processing unit 150 .

[0172] Furthermore, the device 100 includes input / output interfaces including an input interface 171 configured to allow an operator to input instructions, data, values ​​and other user activities, and an output interface 172 configured to output information to the operator.

[0173] The input interface 171 may include a keyboard, a mouse, a touch screen, buttons, and sensors for receiving voice commands and / or visual commands such as gestures, etc. The output interface 172 may include a display, a speaker, a touch screen, etc.

[0174] The input / output interface may also include a logical user interface, such as a graphical user interface (GUI) and / or a command line interface (CLI).

[0175] ***

[0176] As will be appreciated by those skilled in the art, the present invention and the accompanying drawings as described above may be implemented as a method (e.g., a computer-implemented process or any other process), a control apparatus (including a device, a machine, a system, a computer program product and / or any other apparatus), or a combination of the foregoing.

[0177] Thus, embodiments of the present invention may take the form of a fully hardware embodiment, a fully software embodiment (including firmware, resident software, microcode, etc.), or a combination of software and hardware aspects (generally referred to herein as a "system"). In addition, embodiments of the present invention may take the form of a computer program product on a computer-readable medium having computer-executable program code embodied therein.

[0178] The above reference method flowchart and / or block diagram and device describe embodiments of the present invention.It should be understood that each frame of the flowchart and / or block diagram, and / or the combination of frames in the flowchart and / or block diagram can be implemented by computer executable program code.

[0179] Computer executable program code may be provided to a processor of a general purpose computer, a special purpose computer, or other programmable data processing device (e.g., a controller) to produce a specific machine, so that the program code executed by the processor of the computer or other programmable digital processing device creates a method for implementing a flowchart, one or more block diagram blocks, drawings, and / or written descriptions. These computer executable program codes may also be stored in a computer readable memory, which may instruct a computer or other programmable data processing device to work in a specific manner, so that the program code stored in the computer readable memory produces a product including an instruction method, which implements the function / action / output specified in the flowchart, one or more block diagram blocks, drawings, and / or written descriptions. The computer executable program code may also be loaded onto a computer or other programmable data processing device so that a series of operation steps to be executed on a computer or other programmable device produce a computer-implemented process, so that the program code executed on the computer or other programmable device provides steps for implementing the function / action / output specified in the flowchart, one or more block diagram blocks, drawings, and / or written descriptions. Alternatively, the steps or actions implemented by the computer program may be combined with the steps or actions implemented by an operator or manually to implement an embodiment of the present invention.

[0180] It should also be noted that logic flow can be described herein to illustrate various aspects of the present invention, and logic flow should not be interpreted as limiting the present invention to any specific logic flow or logic implementation. The described logic can be divided into different logic blocks (e.g., programs, modules, functions or subroutines) without changing the overall result or departing from the true scope of the present invention. Typically, logic elements can be added, modified, omitted, performed in different orders, or implemented using different logic structures (e.g., logic gates, loop primitives, conditional logic and other logic structures) without changing the overall result or departing from the true scope of the present invention.

[0181] Although certain exemplary embodiments have been described and shown in the accompanying drawings, it should be understood that these embodiments are merely exemplary and not limiting of the broad invention, and that embodiments of the present invention are not limited to the specific constructions and arrangements shown and described, because various other variations, combinations, omissions, modifications, and substitutions are possible in addition to those set forth in the above paragraphs. It will be understood by those skilled in the art that various adjustments, modifications, and / or combinations of the embodiments just described may be configured without departing from the scope and spirit of the present invention. Therefore, it should be understood that within the scope of the appended claims, the present invention may be practiced in a manner different from that specifically described herein. For example, unless otherwise expressly stated, the steps of the process described herein may be performed in an order different from that described herein, and one or more steps may be combined, separated, or performed simultaneously. Based on this disclosure, it will also be understood by those skilled in the art that the different embodiments of the present invention described herein may be combined to form other embodiments of the present invention.

Claims

1. A method for image reconstruction based on computed tomography for image reconstruction of an object, include: - obtaining (S701, S801) a plurality of (N-1)-dimensional images of the object, wherein N=3 or N=4, in which case the plurality of (N-1)-dimensional images of the object are obtained at different angular positions relative to a straight line between a light source and a two-dimensional image detector, or, in which case N=4, the plurality of (N-1)-dimensional images of the object are obtained at different angular positions relative to a straight line between a light source and a three-dimensional image detector; and - reconstructing N-dimensional image data based on an image reconstruction algorithm, the image reconstruction algorithm using as input the data of the plurality of (N-1)-dimensional images of the object and at least one input parameter; wherein the image reconstruction algorithm outputs the reconstructed image data as a reconstructed volume having at least three dimensions; wherein the image reconstruction algorithm is performed (S706, S805) based on the plurality of (N-1)-dimensional images of the object and the determined value set of the at least one input parameter to output (S707, S806) reconstructed N-dimensional image data of the N-dimensional pseudo object as a reconstructed volume represented by a stack of a plurality of slices, wherein each slice of the pseudo object is reconstructed at a different value of the determined set of values ​​of the at least one input parameter as input to the image reconstruction algorithm, and each slice of the plurality of slices of the pseudo object is associated with the same underlying slice of the object, wherein the underlying slice is a two-dimensional planar slice substantially passing through the object, and wherein the N-dimensional pseudo object has at least three dimensions, the three dimensions comprising a first dimension and a second dimension corresponding to two spatial dimensions X and Y of a planar slice volume of the object, and a third dimension of the pseudo object, the third dimension of the pseudo object corresponding to the one-dimensional parameter space of the at least one input parameter of the image reconstruction algorithm instead of corresponding to the third spatial dimension Z; and Based on executing the image reconstruction algorithm, reconstruct (S712, S713, S810, S811) the final N-Dimensional image data of the object as a reconstructed volume, the image reconstruction algorithm uses the multiple (N-1)-Dimensional images of the object obtained and a specific value of the at least one input parameter as input, and the specific value of the at least one input parameter is selected based on the N-Dimensional image data of the N-dimensional pseudo object, wherein the reconstructed volume of the final N-Dimensional image data of the object has at least three dimensions corresponding to three spatial dimensions X, Y and Z.

2. The method according to claim 1, It is characterized in that The plurality of (N-1)-dimensional images are obtained from different viewpoints with respect to a rotation axis arranged along the Z direction, and The basis slices of the object are generally arranged along a two-dimensional XY plane passing through the object, the two-dimensional XY plane being perpendicular to the rotation axis.

3. The method according to claim 2, It is characterized in that The N-dimensional pseudo object is represented by the stack of the plurality of slices stacked along the Z direction.

4. The method according to any one of claims 1 to 3, It is characterized in that - extracting (S703) slice image data from each of the plurality of (N-1)-dimensional images obtained, each slice image data representing one or more adjacent pixel rows of the corresponding (N-1)-dimensional image obtained; as well as - generating a corresponding modified (N-1)-dimensional image for each of the plurality of (N-1)-dimensional images obtained by stacking the corresponding extracted slice image data of the corresponding (N-1)-dimensional image a plurality of times, in particular a plurality of times in the pixel row direction or the pixel column direction; The N-dimensional image data is reconstructed based on the image reconstruction algorithm, and the image reconstruction algorithm uses the modified data of the plurality of (N-1)-dimensional images as input.

5. The method according to any one of claims 1 to 3, wherein Selecting a specific value of the at least one input parameter based on the N-dimensional image data of the N-dimensional pseudo object by the following steps: - calculating a respective metric parameter for each slice of the pseudo object based on applying a predetermined metric calculation algorithm to the (N-1)-dimensional image data of the respective planar portion of the N-dimensional image data associated with the respective slice of the pseudo object, and - identifying the corresponding slice of said pseudo-object for which the corresponding metric parameter has an extreme value, and - identifying a respective input value of the at least one input parameter as the selected specific value of the at least one input parameter, the respective input value of the at least one input parameter being used to reconstruct a respective planar portion of the N-dimensional image data associated with the identified respective slice of the pseudo object.

6. The method according to any one of claims 1 to 3, It is characterized in that The at least one input parameter represents: - weight parameters, which are used to define weights of data terms and regularization terms according to an iterative image reconstruction algorithm; - a penalty function parameter, which is used in a penalty function for a regularization term according to an iterative image reconstruction algorithm; a center offset parameter representing an offset of an actual rotation axis of the measuring system relative to a target rotation axis of the measuring system; - a correction parameter representing a beam hardening correction; - artifact correction parameters, which are used to correct image artifacts; - filtering parameters for use in image filtering applied to the (N-1)-dimensional image; and / or - Weighting factors that weight the different image source energies in dual energy CT image reconstruction.

7. A device for image reconstruction based on computed tomography for image reconstruction of an object, The device include: - an interface (110) configured to receive data indicating a plurality of (N-1)-dimensional images of the object, wherein N=3 or N=4, in which case the plurality of (N-1)-dimensional images of the object are obtained at different angular positions relative to a straight line between a light source and a two-dimensional image detector, or, in which case N=4, the plurality of (N-1)-dimensional images of the object are obtained at different angular positions relative to a straight line between a light source and a three-dimensional image detector; and - an image reconstruction unit (150) configured to reconstruct N-dimensional image data based on an image reconstruction algorithm, the image reconstruction algorithm using as input the data of the plurality of (N-1)-dimensional images of the object and at least one input parameter, wherein the image reconstruction algorithm outputs the reconstructed image data as a reconstructed volume having at least three dimensions; The image reconstruction unit (150) is configured to execute the image reconstruction algorithm based on the plurality of (N-1)-dimensional images of the object and the determined value set of the at least one input parameter to output N-dimensional image data of the N-dimensional pseudo object as a reconstructed volume represented by a stack of a plurality of slices; wherein each slice of the pseudo object is reconstructed at a different value of the determined set of values ​​of the at least one input parameter as input to the image reconstruction algorithm, and each slice of the plurality of slices of the pseudo object is associated with a same base slice of the object, wherein the base slice is a two-dimensional planar slice substantially through the object; and wherein the N-dimensional pseudo object has at least three dimensions, the three dimensions comprising a first dimension and a second dimension corresponding to two spatial dimensions X and Y of a planar slice volume of the object, and a third dimension of the pseudo object, the third dimension of the pseudo object corresponding to the one-dimensional parameter space of the at least one input parameter of the image reconstruction algorithm instead of corresponding to the third spatial dimension Z; and The image reconstruction unit (150) is configured to reconstruct final N-dimensional image data of the object as a reconstructed volume based on executing the image reconstruction algorithm, the image reconstruction algorithm using the multiple (N-1)-dimensional images of the object obtained and a specific value of the at least one input parameter as input, the specific value of the at least one input parameter being selected based on the N-dimensional image data of the N-dimensional pseudo object, wherein the reconstructed volume of the final N-dimensional image data of the object has at least three dimensions corresponding to three spatial dimensions X, Y and Z.

8. A system for image reconstruction of an object, in particular image reconstruction based on computed tomography, include: -light source - a two-dimensional image detector or a three-dimensional image detector, and - Apparatus according to claim 7.

9. A computer program product comprising a computer program comprising computer program instructions adapted to cause a computer, a processor or a computed tomography controller to perform the steps of the method according to any one of claims 1 to 6.