Improved motion artifact compensation through metal artifact reduction.

Abstract

The present invention relates to a method for geometric calibration of DVT imaging in the dental field, the method comprising the following steps: (S1) reconstructing a first volume (1) from a sinogram using an initial projection geometry; (S2) detecting metal regions in the sinogram; (S3) correcting the metal regions in the sinogram; (S4) reconstructing a second volume (2) from the corrected sinogram from step (S3) using the initial projection geometry or a modified projection geometry; and (S5) comparing a simulated sinogram of the reconstructed second volume (2) geometrically calibrated by modifying the projection geometry with said sinogram or the corrected sinogram. and evaluating the sinogram by a similarity index between the sinogram and the second volume (2), wherein the simulated sinogram is calculated from a second volume (2) reconstructed using a modified projection geometry, wherein at least partial regions of the following data, namely (a) the sinogram from step (S1), (b) the corrected sinogram from step (S3), (c) the simulated sinogram, and (d) an intermediate result for calculating a similarity index derived from at least one of the sinograms (a) to (c), are evaluated differently from the remaining regions of the data during the calculation of the similarity index, the partial regions including the metal regions from step (S2).

Description

Technical Field

[0001] The present invention relates to a method for digital volume tomography (DVT) in the dental field. Specifically, the present invention relates to the reduction of motion artifacts in DVT imaging.

Background Art

[0002] In DVT imaging, the X-ray source and the X-ray detector, which are imaging components, face each other and rotate around the patient. A sequence of X-ray projection images that form a sinogram is generated. By knowing the projection geometry, a volume is reconstructed from the sinogram. The projection geometry describes the geometric characteristics of the DVT device and its trajectory during irradiation. It can be represented by a projection matrix.

[0003] In DVT imaging, radiopaque structures such as metal cause image artifacts in the reconstructed volume. Image artifacts occur when the sensitivity of the X-ray detector is not sufficient to physically image X-ray attenuation accurately enough. This causes problems especially behind highly absorbent structures such as metal where noise is dominant. This results in inconsistent values in the reconstruction procedure, often manifested as fringe artifacts around the metal region, inaccurate absorption values in the metal region, and inaccurately imaged metal contours in the volume.

[0004] If the patient moves during DVT acquisition, or if the device calibration is outdated, incorrect projection geometry may be applied during the reconstruction procedure. This results in motion artifacts in the reconstructed volume by taking in inconsistent data. Motion artifact compensation (MAC) procedures, or geometric calibration of images, can estimate projection geometry from a given patient image. Metal artifacts hinder the convergence and / or accuracy of MAC, as MAC performs projection geometry estimation based on a similarity index between a simulated sinogram of the reconstructed volume and a measured sinogram. The simulated sinogram is generated by projecting the artifact-affected volume. Metal artifacts result in strong data inconsistencies. [Overview of the project]

[0005] The object of the present invention is to provide a method for geometric calibration of digital volume tomography (DVT) imaging in the dental field that can provide improved motion artifact compensation (MAC) through metal artifact reduction (MAR).

[0006] This objective is achieved by the method described in claim 1. The subject matter of the dependent claims relates to further developments and preferred embodiments.

[0007] The method according to the present invention is used for geometric calibration of DVT imaging in the dental field. The method consists of the following steps: (S1) reconstructing a first volume from a sinogram using an initial projection geometry; (S2) detecting metal regions in the sinogram; (S3) correcting the metal regions in the sinogram; (S4) reconstructing a second volume from the corrected sinogram from step (S3) using the initial projection geometry or a modified projection geometry; and (S5) geometrically calibrating by modifying the projection geometry and using a similarity index between the simulated sinogram of the reconstructed second volume and the sinogram or corrected sinogram. The process comprises a step of evaluating, wherein the simulated sinogram is calculated from a second volume reconstructed using a modified projection geometry, wherein at least a partial region of the following data, namely (a) the sinogram from step (S1), (b) the corrected sinogram from step (S3), (c) the simulated sinogram, and (d) an intermediate result for calculating a similarity index derived from at least one of the sinograms (a) to (c), is evaluated differently from the rest of the data during the calculation of the similarity index, wherein the partial region includes the metal region from step (S2).

[0008] A key feature of the present invention is the use of metal artifact correction for reconstructing the volume used in MAC, and the improvement of MAC by special handling of metal regions during the calculation of the similarity index, preferably by filtering / weighting those metal regions. Another important and beneficial effect of the present invention is the improved accuracy and faster convergence of MAC.

[0009] The present invention will be described in more detail below with reference to exemplary embodiments and the drawings. [Brief explanation of the drawing]

[0010] [Figure 1]A diagram illustrating an overview of a method for geometric calibration of DVT imaging in the dental field according to one embodiment. [Figure 2] Flowchart of a further embodiment. [Figure 3] Flowchart of a further embodiment. [Figure 4] Flowchart of a further embodiment. [Figure 5] Flowchart of a further embodiment. [Figure 6] A diagram showing a computer-assisted DVT system in which the method according to the present invention can be performed. [Modes for carrying out the invention]

[0011] The reference numerals shown in the drawings indicate the elements listed below, which are referred to in the following description of exemplary embodiments. 1. DVT system 2. X-ray devices 3.X-ray source 4. X-ray detector 5. Control Unit 6.Head fixation device 7 bytes 8. Computers 9. Display definition I1, I2: Projected images

[0012]

number

[0013] : Average value of Image I N: Number of pixels in the projected image I(u,v): Pixel values ​​of projected image I at position (u,v) s(I1,I2): Similarity index of projected images MSE(I1,I2): Mean Squared Error MAD(I1,I2): Mean absolute difference NCC1(I1,I2): Normalized Cross-Correlation NCC2(I1, I2): Pixel-based Normalized Cross-Correlation GC1(I1, I2): Gradient Correlation 1 GC2(I1, I2): Gradient Correlation 2 GI1(I1, I2): Gradient Information GI2(I1, I2): Gradient Information Using Linear Scaling GO(I1, I2): Gradient Direction LoG: Combination of Gaussian and Laplacian MI(A, B): Mutual Information The method according to the present invention is used for geometric calibration of DVT imaging in the dental field. As shown in FIG. 2, the method includes the following steps: (S1) reconstructing a first volume (1) from a sinogram using an initial projection geometry; (S2) detecting a metal region in the sinogram; (S3) correcting the metal region in the sinogram; (S4) reconstructing a second volume (2) from the corrected sinogram from step (3) using the initial projection geometry or a modified projection geometry; (S5) geometrically calibrating by changing the projection geometry and evaluating by a similarity index between the simulated sinogram of the reconstructed second volume (2) and the sinogram or the corrected sinogram, wherein the simulated sinogram is calculated from the second volume (2) reconstructed using the modified projection geometry, and wherein at least a partial region of at least one of the following data: (a) the sinogram from step (S1), (b) the corrected sinogram from step (S3), (c) the simulated sinogram, and (d) intermediate results for calculating a similarity index derived from at least one of the sinograms (a)-(c) is evaluated to be different from the remaining region of the data during the calculation of the similarity index, and the partial region includes the metal region from step (S2).

[0014] Metal or artificial radiopaque structures form distinct contours in a sinogram. In a simulated sinogram, these contours are typically obscured or distorted due to inaccurate projection geometry in the reconstruction procedure, missing physical information in the metal regions, and metal artifacts in the reconstructed volume. While metal artifacts in the reconstructed volume can be reduced by using Metal Artifact Reduction (MAR) in steps (S2) to (S4), they are usually not fully or physically correctly restored. These inconsistent metal regions hinder the convergence of geometric calibration or motion artifact correction (MAC). By evaluating metal regions differently during the calculation of the similarity index, the convergence of geometric calibration can be improved, often accelerated, and its accuracy increased. The partial region in step (S5) is preferably larger than the metal region detected from step (S2) so that artifacts at the edges of the metal region can also be detected. Correction of metal regions in the sinogram in step (S3) can also be performed on a larger region to compensate for the inaccuracies in the detection of metal regions in step (S2).

[0015] Figure 3 shows a preferred embodiment of the method, which includes a step (S6) after step (S5) in which steps (S4) to (S5) are repeated. This is for the iterative improvement of MAC. The iteration terminates when an termination criterion or convergence criterion is reached. Possible termination criteria are: a) whether the change in projection geometry is less than a threshold; b) whether the change in final volume is less than a threshold; c) whether the number of iteration steps is greater than a threshold; or d) whether the computation time is greater than a threshold.

[0016] In another preferred embodiment, after step (S5), the method comprises step (S7) which repeats steps (S2) to (S5) or steps (S2) to (S6). This helps to iteratively improve the MAR by using intermediate results of the MAC. This also means iterative improvement of the MAC. The iterations terminate when a second termination criterion or convergence criterion is reached. Possible second termination criteria are a) whether the change in projection geometry is less than a threshold, b) whether the change in final volume is less than a threshold, c) whether the number of iteration steps is greater than a threshold, d) whether the computation time is greater than a threshold, or e) whether the amount of metal in the volume or sinogram is greater than a threshold. Figure 4 shows this embodiment combined with the previous preferred embodiment.

[0017] In a further preferred embodiment, the method comprises, after step (5), a step (S8) of reconstructing the third volume (3) while taking into account the estimated projection geometry and applying metal artifact correction. The third volume takes into account the results of the MAR and MAC correction methods. Figure 5 shows this embodiment combined with other preferred embodiments.

[0018] In another preferred embodiment, the detection of metal regions in the sinogram in step (S2) comprises (a) detecting metal regions in the first volume (1) and (b) projecting the detected metal regions into the sinogram. Detecting metal regions in a volume typically has the advantage of being more robust or easier to implement than direct detection of metal regions in a sinogram. By applying projected geometry or estimated projected geometry, the detected metal regions can be converted into a sinogram. The metal regions detected in the volume can be enlarged to compensate for inaccurate projected geometry and / or inaccurate detection of metal regions.

[0019] In another preferred embodiment, the detection of metal regions in the sinogram in step (2) comprises (a') generating a simulated sinogram of the first volume (1) and (b') detecting metal regions in the simulated sinogram from step (S2)(a) and transferring the detected metal regions to the sinogram. Detecting metal regions in the simulated sinogram is easier than detecting metal regions in the sinogram because a simplified representation of the superimposed structures is possible when generating the simulated sinogram. The metal regions detected in the simulated sinogram can be transferred to the sinogram because the projected geometry estimated with respect to the sinogram is used in generating the simulated sinogram.

[0020] In a further preferred embodiment, the method further comprises the step (a) determining the metal regions in the second volume (2) using the detected metal regions in the sinogram after step (S4). The determination of the metal regions in the second volume is performed by backprojecting or reconstructing the detected metal regions in the sinogram using the projected geometry or estimated projected geometry. This step is required if detection of metal regions in a simulated sinogram has been performed.

[0021] In a more preferred embodiment, the method further comprises (b) correcting the metal region in the second volume (2) after step (S4), wherein step (b) comprises one or more of the following steps: (b1) filling the metal region in the second volume (2) with values ​​from the reconstructed first volume (1) or artificial values ​​representing the metal region; and (b2) mixing the values ​​from step (S4)(b1) with the values ​​from the second volume (2) while weighting them. This has the advantage that the metal region in the simulated sinogram in step (S5) is better suited for comparison with the sinogram and calculation of similarity indices. Figure 5 shows this embodiment combined with other preferred embodiments.

[0022] In a further preferred embodiment, the correction of metal regions in the sinogram in step (S3) comprises one or more of the following steps: (a) filling the metal regions in the corrected sinogram with new pixel values, which are either calculated from adjacent pixels or correspond to artificial pixel values; and (b) mixing the new pixel values ​​from step (S3)(a) with the pixel values ​​of the sinogram in a weighted manner. Metal regions in a sinogram cannot be measured physically accurately due to strong absorption. By replacing them with plausible pixel values, the occurrence of metal artifacts during reconstruction can be avoided. The new plausible pixel values ​​can be artificial values ​​or can be calculated from adjacent pixels in the sinogram. Mixing the new pixel values ​​with the original pixel values ​​in a weighted manner preserves some of the uncertain physical information.

[0023] In a further preferred embodiment, the different evaluation of a partial region in at least one of the data during the calculation of the similarity index in step (S5) preferably comprises pixel-by-pixel weightings within a value range of 0 to 1. This has the advantage that the metal regions and their surroundings are included more weakly in the calculation of the similarity index and therefore do not hinder, or hinder, the convergence of the MAC.

[0024] In a further preferred embodiment, the different evaluation of a partial region in at least one of the data during the calculation of the similarity index in step (S5) comprises local filtering. This has the advantage that the metal regions and their surroundings are included more weakly in the calculation of the similarity index and therefore do not hinder, or hinder, the convergence of the MAC.

[0025] In a further preferred embodiment, the projection geometry of the DVT imaging on the patient's head, or the estimated projection geometry, is described by geometric parameters. The projection geometry is preferably described by internal and external parameters, wherein the internal parameters include the relative positions between the X-ray source and the X-ray detector and their resolutions, and the external parameters include transformations consisting of rotation and translation of selected subregions of the projection image and volume.

[0026] In a further preferred embodiment, in addition to image data, for example, initial projection geometry in the form of data from device calibration may be provided to the geometry calibration procedure as input data (step S5).

[0027] The method according to the present invention is computer-implementable and can be executed on a computerized DVT system (1). Figure 6 shows an example of one embodiment of the DVT system (1). In this regard, the present invention also includes a computer program having computer-readable code. The computer program may be provided on a data storage device. The computerized DVT system (1) comprises an X-ray device (2) for performing imaging of a patient, thereby generating a sinogram. The X-ray device (2) has an X-ray source (3) and an X-ray detector (4), which rotate around the patient's head during irradiation. The trajectories of the X-ray source (3) and X-ray detector (4) during irradiation may trace a circular path. Alternatively, the trajectories may be assumed to be of a deviant shape. If several actuators are controlled simultaneously, a device trajectory around the patient's head that deviates from a purely circular path can be achieved. The patient's head is positioned in the X-ray device using a bite (7) and a head restraint (6). The computerized DVT system (1) comprises a control unit (5), preferably a computer (8) or computing unit connectable to the X-ray device (2), and preferably a display (9) for visualizing the dataset, among other things. The computer (8) may be connected to the X-ray device (2) via a local area network (not shown) or alternatively via the internet. The computer (8) may be part of a cloud. Alternatively, the computer (8) may be integrated into the X-ray apparatus (2). Calculations may alternatively be performed in the cloud. The computer (8) runs computer programs and provides datasets, including visualizations on the display (9). The display (9) may be spatially separated from the X-ray device (2). Preferably, the computer (8) may also control the X-ray device (2). Alternatively, a separate computer may be used for control and reconfiguration.

[0028] According to the present invention, the dataset generated by the above embodiments can be presented to a physician for visualization, particularly for diagnostic purposes, preferably by a display (9) or printout. similarity index The similarity index s(I1,I2) is described in detail below. The similarity index is a scalar quantity that describes the similarity between two projected images I1 and I2. A negative difference index may also be used instead of a similarity index. Similarity indices can be divided into two classes: pixel-based similarity indices and histogram-based similarity indices. Examples of pixel-based similarity metrics: a) Mean squared error

[0029]

number

[0030] b) Mean absolute difference

[0031]

number

[0032] c) Normalized cross-correlation

[0033]

number

[0034] d) Pixel-based normalized cross-correlation

[0035]

number

[0036] e) Gradient correlation 1

[0037]

number

[0038] f) Gradient correlation 2

[0039]

number

[0040] g) Gradient information

[0041]

number

[0042]

number

[0043] and

[0044]

number

[0045] h) Gradient information using linear scaling

[0046]

number

[0047]

number

[0048] and

[0049]

number

[0050] And α is the scaling factor.

[0051] Here, g(u,v) and θ(u,v) are examples of intermediate results for calculating similarity indices derived from at least one of the sinograms a) to c). Here, a) shows the sinogram from step (S1), b) shows the corrected sinogram from step (S3), and c) shows the simulated sinogram.

[0052] i) Direction of gradient

[0053]

number

[0054] Here,

[0055]

number

[0056] And C1, t1, and t2 are scalar constants.

[0057] j) Combinations of Gaussian and Laplacian

[0058]

number

[0059] Here,

[0060]

number

[0061] It has a Gaussian kernel width σ. Examples of histogram-based similarity metrics: k) Mutual information

[0062]

number

[0063] Here, p(a) is the probability of the value a occurring in input image l1, p(b) is the probability of the value b occurring in input image l2, and p(a,b) is the conditional probability of the values ​​a and b occurring in input images l1 and l2, respectively.

[0064] For example, different evaluations of sub-regions in step (S5) of calculating the similarity index can be implemented as follows:

[0065] a') MSE(wI1,wI2) Here, w is the weight for each pixel, where w ≤ 1 in the partial region and w = 1 in the remaining region.

[0066] a'')

[0067]

number

[0068] Here, w1 and w2 are the weights for each pixel, and for each pixel, w1 + w2 = 1, f 11 ,f 12 ,f 21 , and f 22 This is local filtering.

[0069] g')

[0070]

number

[0071] Here, w is the weight for each pixel, where w ≤ 1 in the partial region and = 1 in the remaining region.

[0072] k') In mutual information computation, p(a) is calculated from (wI1) and p(b) is calculated from (wI2), where w is the pixel-wise weight [0,1]. The invention described in the original claims of this application is listed below. [C1] A method for geometric calibration of DVT imaging in the dental field, (S1) A step of reconstructing the first volume (1) from the sinogram using the initial projection geometry, (S2) A step of detecting the metal region in the sinogram, (S3) A step of correcting the metal region in the sinogram, (S4) A step of reconstructing the second volume (2) from the corrected sinogram from step (S3) using the initial projection geometry or the modified projection geometry, (S5) The step of geometrically calibrating by changing the projection geometry and evaluating the simulated sinogram of the reconstructed second volume (2) by a similarity index between the sinogram or the corrected sinogram. Equipped with, Herein, the simulated sinogram is calculated from the reconstructed second volume (2) using the modified projection geometry, and herein, the following data, namely, (a) Sinogram from step (S1), (b) Corrected sinogram from step (S3), (c) Simulated sinogram and (d) A method characterized in that at least a partial region of the intermediate result for calculating the similarity index derived from at least one of the sinograms (a) to (c) is evaluated differently from the rest of the data during the calculation of the similarity index, wherein the partial region includes the metal region from step (S2). [C2] The method according to C1, characterized in that, after step (S5), step (S6) is a step in which steps (S4) to (S5) are repeated. [C3] The method according to C1 or 2, characterized in that, after step (S5), there is a step (S7) which repeats steps (S2) to (S5) or steps (S2) to (S6). [C4] The method according to any one of C1 to 3, further comprising step (S8) of reconstructing the third volume (3) after step (S5) by taking into account the estimated projection geometry and applying metal artifact correction. [C5] The detection of the metal region in the sinogram in step (S2) is (a) Detection of the metal region in the first volume (1) and (b) Projection of the detected metal region onto the sinogram and The method according to any one of C1 to C4, characterized by comprising the above. [C6] The detection of the metal region in the sinogram in step (S2) is (a') To generate a simulated sinogram of the first volume (1), (b') The method according to any one of C1 to 4, characterized by comprising detecting the metal region in the simulated sinogram of step (S2)(a) and transferring the detected metal region to the sinogram. [C7] The method according to C6, characterized in that, after step (S4), (a) a step of determining the metal region in the second volume (2) using the detected metal region in the sinogram. [C8] The method according to C5 or 7, wherein, after step (S4), (b) a step of correcting the metal region in the second volume (2), the step (b) comprising one or more of the following steps: (b1) filling the metal region in the second volume (2) with a value from the reconstructed first volume (1) or an artificial value representing the metal region; and (b2) mixing the value from step (S4)(b1) with the value from the second volume (2) while weighting them. [C9] The method according to any one of C1 to 8, wherein the correction of the metal region in the sinogram in step (S3) comprises one or more of the following steps: (a) filling the metal region in the corrected sinogram with new pixel values, wherein the step is either calculated from adjacent pixels or corresponds to artificial pixel values; and (b) mixing the new pixel values ​​from step (S3)(a) with the pixel values ​​of the sinogram in a weighted manner. [C10] The method according to any one of C1 to 9, characterized in that the different evaluations of at least a portion of the data during the calculation of the similarity index in step (S5) comprises pixel-by-pixel weightings which are preferably in the range of values ​​from 0 to 1. [C11] The method according to any one of C1 to 10, wherein the different evaluation of at least a portion of the data during the calculation of the similarity index in step (S5) comprises local filtering. [C12] A computer program comprising computer-readable code that, when executed by a computer-assisted DVT system (1), causes the computer-assisted DVT system (1) to perform the method step of the method described in any one of C1 to C11. [C13] A computerized DVT system comprising an X-ray device (2) and a computing unit (8) configured to execute the computer program described in C12. [C14] A method for using the dataset for visualization provided by the method described in any one of items C1 through C11.

Claims

1. A method for geometric calibration of DVT imaging in the dental field, (S1) A step of reconstructing the first volume (1) from the sinogram using the initial projection geometry, (S2) A step of detecting the metal region in the sinogram, (S3) A step of correcting the metal region in the sinogram, (S4) A step of reconstructing the second volume (2) from the corrected sinogram from step (S3) using the initial projection geometry or the modified projection geometry, (S5) The step of geometrically calibrating by changing the projection geometry and evaluating the simulated sinogram of the reconstructed second volume (2) by a similarity index between the sinogram or the corrected sinogram. Equipped with, Herein, the simulated sinogram is calculated from the reconstructed second volume (2) using the modified projection geometry, and herein, the following data, namely, (a) Sinogram from step (S1), (b) Corrected sinogram from step (S3), (c) Simulated sinogram and (d) A method characterized in that at least a partial region of the intermediate result for calculating the similarity index derived from at least one of the sinograms (a) to (c) is evaluated such that it is weighted differently from the rest of the data during the calculation of the similarity index, and the partial region includes the metal region from step (S2).

2. The method according to claim 1, characterized in that, after step (S5), step (S6) is a step of repeating steps (S4) through (S5).

3. The method according to claim 1 or 2, characterized in that, after step (S5), step (S7) is a step of repeating steps (S2) to (S5) or steps (S2) to (S6).

4. The method according to claim 1 or 2, further comprising the step (S8) of reconstructing the third volume (3) after step (S5) by taking into account the estimated projection geometry and applying metal artifact correction.

5. The detection of the metal region in the sinogram in step (S2) is (a) Detection of the metal region in the first volume (1), (b) Projection of the detected metal region onto the sinogram and The method according to claim 1 or 2, characterized by comprising:

6. The detection of the metal region in the sinogram in step (S2) is (a') To generate a simulated sinogram of the first volume (1), The method according to claim 1 or 2, characterized by comprising (b') detecting the metal region in the simulated sinogram of step (S2)(a') and transferring the detected metal region to the sinogram.

7. The method according to claim 6, further comprising the step (a) after step (S4) determining the metal region in the second volume (2) using the detected metal region in the sinogram.

8. The method according to claim 5, further comprising the step (b) correcting the metal region in the second volume (2) after step (S4), wherein step (b) comprises one or more of the following steps: (b1) filling the metal region in the second volume (2) with a value from the reconstructed first volume (1) or an artificial value representing the metal region; and (b2) mixing the value from step (S4) (b1) with the value from the second volume (2) while weighting them.

9. The method according to claim 1 or 2, characterized in that the correction of the metal region in the sinogram in step (S3) comprises one or more of the following steps: (a) filling the metal region in the corrected sinogram with new pixel values, wherein the step is either calculated from adjacent pixels or corresponds to artificial pixel values; and (b) mixing the new pixel values ​​from step (S3)(a) with the pixel values ​​of the sinogram in a weighted manner.

10. The method according to claim 1 or 2, characterized in that the weighted evaluation of at least a portion of the data during the calculation of the similarity index in step (S5) comprises pixel-by-pixel weights in the range of 0 to 1.

11. The method according to claim 1 or 2, characterized in that the evaluation of the similarity index in step (S5) with different weights for at least a portion of the data during calculation comprises local filtering.

12. A computer program comprising computer-readable code, which, when executed by a computer-assisted DVT system (1), causes the computer-assisted DVT system (1) to perform the method step of the method according to claim 1 or 2.

13. A computerized DVT system comprising an X-ray device (2) and a computing unit (8) configured to execute the computer program described in claim 12.

14. A method for using a dataset for visualization provided by the method of claim 1 or 2.

Images