CT image reconstruction device, CT image reconstruction method, and CT image reconstruction program

Abstract

To specify a position of an object area existing inside a sample even when the object area exists in a part different from a central part inside the sample, when reconfiguring a CT image inside the sample by using a measuring system for irradiating the sample with an X-ray from a part of all directions.SOLUTION: The CT image reconstruction system includes an X-ray source, a detector that detects X-rays emitted from the X-ray source and transmitted through a sample, and a CT image reconstruction device. The X-ray source irradiates a sample and a stage on which the sample is placed with X-rays from some of all directions. The stage includes a first member and a plurality of second members having lower X-ray transparency than the first member. The CT image reconstruction device reconstructs, for each second member, a CT image in which the second member appears at the center. The CT image reconstruction device detects the artifact shape from each of the plurality of CT images of each of the second members, and specifies the position of the target existing inside the sample based on the CT image in which the artifact shape is detected.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a CT image reconstruction device, a CT image reconstruction method, and a CT image reconstruction program. [Background technology]

[0002] Patent Document 1 discloses an X-ray system that can improve imaging speed and / or reduce radiation dose compared to conventional imaging techniques such as CT. The system disclosed in Patent Document 1 can identify a volume of interest within a subject.

[0003] Furthermore, Patent Document 2 discloses a CT image reconstruction system for obtaining CT images with reduced defects in a linear CT system. The reconstruction unit of the CT image reconstruction device in this CT image reconstruction system reconstructs a defective CT image of a sample based on detection data detected by multiple detectors. The defective CT image has a smaller amount of defect than a conventional defective CT image reconstructed based on detection data obtained by irradiating a sample with X-rays from a single X-ray source so that the sample receives X-rays from one direction. The image restoration unit of the CT image reconstruction device inputs the defective CT image into a trained learning device to obtain a restored CT image in which the defect has been restored.

[0004] Furthermore, Patent Document 3 discloses a tomography method and device for pipe inspection that can perform image reconstruction calculations only from incomplete projection data within a very small angle range, in cases such as on-site plant piping photography, where it is difficult to obtain the complete data required with conventional CT, and can create high-quality tomographic images that can visualize minute defects such as shallow wall thinning inside the object.

[0005] Furthermore, Patent Document 4 discloses a computed tomography apparatus that creates radiographic images of defects viewed from multiple directions based on the respective radiographic transmission data output by multiple detectors in a detector group, and also creates size information and position information of the defects.

[0006] Furthermore, Patent Document 5 discloses an X-ray CT system that realizes spectral CT that overcomes defects identified in the spectral CT configuration. This X-ray CT system uses a trained model to generate images with reduced streak artifacts.

[0007] Patent Document 6 also discloses a medical device that includes a processing circuit that acquires a sinogram representing the intensity of radiation detected by a detector, acquires a neural network, applies the acquired sinogram to the neural network, thereby outputting indicators of two or more artifacts from the neural network, determines whether or not a first artifact is present in the acquired sinogram and whether or not a second artifact is present in the acquired sinogram based on the indicators from the neural network, and, when it is determined that one or more artifacts are present in the acquired sinogram, generates a corrected sinogram in which the artifacts are corrected and performs reconstruction processing, and when it is determined that no artifacts are present, performs reconstruction processing using the sinogram.

[0008] Furthermore, Patent Document 7 discloses a method and system for CT imaging using a distributed X-ray source and interpolation-based reconstruction, which achieves high spatial resolution, high temporal resolution, good image quality, and sufficient imaging range using a standard or relatively small detector. [Prior art documents] [Patent documents]

[0009] [Patent Document 1] Special Publication No. 2023-503638 [Patent Document 2] Japanese Patent Application Publication No. 2022-158299 [Patent Document 3] Japanese Patent Application Laid-Open No. 2010-107298 [Patent Document 4] Japanese Patent Application Publication No. 3-100405 [Patent Document 5] Japanese Patent Publication No. 2020-99667 [Patent Document 6] Patent Publication No. 2021-13726 [Patent Document 7] Japanese Patent Application Laid-Open No. 2006-297083 Summary of the Invention [Problem to be solved by the invention]

[0010] Now, consider the case where a measurement system is used in which X-rays are irradiated onto the sample from some of the omnidirectional directions to obtain a CT image of the inside of the sample. In such a case, since the X-rays are irradiated onto the sample from some directions rather than all directions, it is not possible to accurately identify the location of voids or other objects inside the sample.

[0011] Obtaining a CT image with defects by irradiating a sample with X-rays from some of the directions is also called incomplete CT. Obtaining a CT image with defects by moving the sample in a straight line and irradiating X-rays from one direction is also called linear CT. Linear CT is an example of incomplete CT.

[0012] When attempting to obtain a CT image of the interior of a sample using linear CT, if there are defects such as voids or other objects inside the sample, it is difficult to determine the location of the defect. Furthermore, if the defect is located near the edge of the sample rather than in the center, the streak artifacts that appear in the CT image will have a unique shape, making it even more difficult to determine the location of the defect.

[0013] The above Patent Documents 1 to 7 disclose techniques for removing streak artifacts from a CT image of the inside of a sample to obtain a desired reconstructed image, but do not disclose a method for estimating the position of a target region, such as a void or other object, present inside the sample from the streak artifacts.

[0014] The present disclosure has been made in consideration of the above circumstances, and provides a CT image reconstruction device, a CT image reconstruction method, and a CT image reconstruction program that can identify the position of a target area inside a sample when reconstructing a CT image inside the sample using a measurement system that irradiates X-rays onto the sample from some of the all directions, even if the target area is located in a part other than the central part inside the sample. [Means for solving the problem]

[0015] A first aspect of the present disclosure is a CT image reconstruction device in a CT image reconstruction system including an X-ray source, a detector that detects X-rays irradiated from the X-ray source and transmitted through a sample, and a CT (Computed Tomography) image reconstruction device, wherein the X-ray source irradiates X-rays from some of all directions toward a sample and a stage on which the sample is placed, and the stage is configured to include a first member and a plurality of second members that are less X-ray transparent than the first member, and the CT image reconstruction device includes: a reconstruction unit that reconstructs, based on detection data detected by the detector, CT images having defects and in which the second member is centered for each of the second members; a detection unit that detects an artifact shape determined according to a measurement system representing the X-ray source and the detector from each of the plurality of CT images for each of the second members; and an identification unit that identifies the position of an object present inside the sample based on the CT images in which the artifact shape has been detected.

[0016] The second aspect is a CT image reconstruction method executed by a CT image reconstruction device in a CT image reconstruction system including an X-ray source, a detector that detects X-rays irradiated from the X-ray source and transmitted through a sample, and a CT (Computed Tomography) image reconstruction device, wherein the X-ray source irradiates X-rays from some of all directions toward the sample and a stage on which the sample is placed, and the stage is configured to include a first member and a plurality of second members that are less X-ray transparent than the first member, and the CT image reconstruction device executes the following processing based on detection data detected by the detector: reconstructs a CT image having defects and in which the second member is centered for each of the second members; detects an artifact shape determined according to a measurement system representing the X-ray source and the detector from each of the plurality of CT images for each of the second members; and identifies the position of an object present inside the sample based on the CT image in which the artifact shape has been detected.

[0017] A third aspect is a CT image reconstruction program executed by a CT image reconstruction device in a CT image reconstruction system including an X-ray source, a detector that detects X-rays irradiated from the X-ray source and transmitted through a sample, and a CT (Computed Tomography) image reconstruction device, wherein the X-ray source irradiates X-rays from some of all directions toward the sample and a stage on which the sample is placed, and the stage is configured to include a first member and a plurality of second members that are less X-ray transparent than the first member, and the CT image reconstruction program causes a computer constituting the CT image reconstruction device to execute the following processes: reconstruct a CT image having a defect and in which the second member is centered for each of the second members based on detection data detected by the detector; detect an artifact shape determined according to a measurement system representing the X-ray source and the detector from each of the plurality of CT images for each of the second members; and identify the position of an object present inside the sample based on the CT image in which the artifact shape has been detected. [Effects of the Invention]

[0018] According to the present disclosure, when a CT image of the inside of a sample is reconstructed using a measurement system that irradiates the sample with X-rays from some of the all directions, the position of the target area inside the sample can be identified even if the target area is located in a part other than the central part inside the sample. [Brief explanation of the drawings]

[0019] [Figure 1] FIG. 1 is a diagram illustrating an example of a CT image reconstruction system according to an embodiment. [Figure 2] FIG. 2 is a block diagram showing an example of a functional configuration of CT image reconstruction according to the embodiment. [Figure 3] FIG. 2 is a block diagram showing an example of the configuration of a computer that functions as a CT image reconstruction device according to the embodiment. [Figure 4] FIG. 10 is a diagram for explaining a transmission image of a linear CT. [Figure 5] FIG. 10 is a diagram for explaining streak artifacts that occur in a reconstructed image obtained by linear CT. [Figure 6] FIG. 10 is a diagram for explaining streak artifacts in more detail. [Figure 7] FIG. 10 is a diagram for explaining streak artifacts in more detail. [Figure 8] FIG. 10 is a diagram for explaining a method for calculating the size of a rhombus. [Figure 9] FIG. 10 is a diagram showing an example of an image obtained by calculating the difference between an image showing an ideal sample and an actually captured reconstructed image. [Figure 10] FIG. 10 is a diagram showing an example of a streak artifact that appears in a shape other than a diamond. [Figure 11] FIG. 2 is a diagram for explaining a stage on which a sample is placed in the present embodiment. [Figure 12] FIG. 1 is a diagram for explaining a method of obtaining a transmission image by irradiating a sample and a stage with X-rays. [Figure 13]FIG. 1 is a diagram for explaining a method of obtaining a transmission image by irradiating a sample and a stage with X-rays. [Figure 14] FIG. 1 is a diagram for explaining a method of obtaining a transmission image by irradiating a sample and a stage with X-rays. [Figure 15] 10A and 10B are diagrams illustrating an example of a method for arranging spheres embedded in a stage. [Figure 16] FIG. 2 is a diagram for explaining an image of a sphere portion and an image of a sample portion in a CT image. [Figure 17] 10 is a flowchart showing an example of the processing flow of a CT image reconstruction program according to the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0020] Hereinafter, embodiments of the technology of the present disclosure will be described in detail with reference to the drawings. Note that in the present embodiment, components and processes that perform the same actions and functions are assigned the same reference numerals throughout the drawings, and duplicated descriptions may be omitted as appropriate. Each drawing is merely a schematic illustration to allow a sufficient understanding of the technology of the present disclosure. Therefore, the technology of the present disclosure is not limited to only the illustrated examples. Furthermore, in the present embodiment, descriptions of configurations that are not directly related to the technology of the present disclosure or well-known configurations may be omitted.

[0021] FIG. 1 is a diagram showing an example of a CT image reconstruction system 10 for linear CT. As shown in FIG. 1, the CT image reconstruction system 10 includes an X-ray source 11, a detector 12, a conveyor belt 13, and a CT image reconstruction device 14. The X-ray source 11, the detector 12, and the CT image reconstruction device 14 are communicatively connected. The z-axis in FIG. 1 is set along the direction from the back to the front of the paper. d in FIG. 1 sd represents the distance between the X-ray source 11 and the detector 12. so represents the distance between the X-ray source 11 and the conveyor belt 13. o represents the length of the conveyor belt 13.

[0022] The X-ray source 11 irradiates the sample S with X-rays R.

[0023] The detector 12 detects the X-rays that are emitted from the X-ray source 11 and transmitted through the sample S.

[0024] The conveyor belt 13 operates to move the sample S in a direction D.

[0025] As shown in Fig. 1, an X-ray source 11 irradiates a sample S with X-rays R from some of all directions. Therefore, the measurement system representing the X-ray source 11 and detector 12 in the CT image reconstruction system 10 shown in Fig. 1 is a measurement system that acquires an image with so-called defects (hereinafter also simply referred to as an incomplete CT image).

[0026] The CT image reconstruction device 14 identifies the position of an object present inside the sample S based on the detection data detected by the detector 12. The object present inside the sample S may be, for example, a void or other object. Hereinafter, the other object will also be referred to as an inclusion.

[0027] 2 is a block diagram showing an example of the functional configuration of the CT image reconstruction device 14 according to the embodiment. As shown in FIG. 2, the CT image reconstruction device 14 functionally includes a data storage unit 140, a reconstruction unit 142, a detection unit 144, and an identification unit 146.

[0028] 3, the CT image reconstruction device 14 is realized by a computer 50 including a CPU 51, a memory 52 as a temporary storage area, and a non-volatile storage unit 53. The computer 50 of the CT image reconstruction device 14 also includes an input / output interface (I / F) 54 to which external devices, output devices, etc. are connected, and a read / write (R / W) unit 55 that controls reading and writing of data from and to a recording medium 59. The computer also includes a network I / F 56 that is connected to a network such as the Internet. The CPU 51, memory 52, storage unit 53, input / output I / F 54, R / W unit 55, and network I / F 56 are connected to one another via a bus 57.

[0029] The storage unit 53 can be realized by a hard disk drive (HDD), a solid state drive (SSD), a flash memory, etc. The storage unit 53 as a storage medium stores a program for causing the computer to function. The CPU 51 reads the program from the storage unit 53, loads it into the memory 52, and sequentially executes the processes contained in the program.

[0030] In the measurement system shown in Figure 1, X-rays R are irradiated onto the sample S from one direction of the sample S while the sample S is moving in a linear motion. For this reason, the image of the inside of the sample S obtained by such a measurement system is also called linear CT.

[0031] A measurement system using linear CT, such as that shown in Figure 1, can only obtain an incompletely reconstructed image of the interior of the sample S. As a result, the shape of voids or inclusions inside the sample S is unclear in the incompletely reconstructed image of the interior of the sample S.

[0032] FIG. 4 is a diagram illustrating the transmission images of the linear CT. IM1 shown in FIG. 4 is a transmission image based on detection data obtained when the sample S was located at the left end of the conveyor belt 13. IM2 shown in FIG. 4 is a transmission image based on detection data obtained when the sample S was located at the center of the conveyor belt 13. IM3 shown in FIG. 4 is a transmission image based on detection data obtained when the sample S was located at the right end of the conveyor belt 13. As shown in FIG. 4, the transmission images IM1 and IM3 are less clear than the transmission image IM2. An image of the interior of the sample S is reconstructed using these less clear transmission images. However, if a void or inclusion exists inside the sample S, a streak artifact appears in the reconstructed image of the linear CT. Therefore, even if only these transmission images are used, it is not possible to detect the shape of a void in the depth direction (Y direction), which is the direction in which the X-rays are irradiated.

[0033] An artifact is a phenomenon in which an image that does not actually exist inside the sample S is captured in the reconstructed image. A streak artifact is a phenomenon in which noise appears in a consistent direction in the reconstructed image due to insufficient information when generating the reconstructed image.

[0034] Figure 5 is a diagram illustrating streak artifacts that occur in reconstructed images obtained by linear CT. Image A shown in Figure 5 is a cross-sectional image of a true numerical phantom inside a sample, and image B shown in Figure 5 is a reconstructed image of a real sample using linear CT. A void is present in the center of image A, and a void also exists in the center of the real sample. However, the shape of the void in the center of image B is distorted and significantly different from the shape of the void in the cross-sectional image of the true numerical phantom.

[0035] 6 and 7 are diagrams for explaining the streak artifact in more detail. Inside the sample S shown in FIG. 6, there are voids T A ,T B The xyz coordinate system in Fig. 6 corresponds to the xyz coordinate system in Fig. 1. The Z cross section in Fig. 6 is B It is positioned so that it slices the image into slices.

[0036] Figure 7 shows an example of a reconstructed image of the linear CT of the sample S in Figure 6. The images shown in the left column of Figure 7 are cross-sectional images of the correct numerical phantom inside the sample, and the images shown in the right column of Figure 7 are reconstructed images of the actual sample S by linear CT.

[0037] As a result of carrying out a simulation as shown in Fig. 7, the inventors of the present invention found that streak artifacts occur not only at the outer boundary of the sample S but also at the boundary of voids or inclusions present inside the sample S. As shown in image B of Fig. 5, it can be seen that streak artifacts occur around voids. Note that, as shown in image B of Fig. 5, streak artifacts occur along the depth direction (y direction) in which X-rays are irradiated into the sample.

[0038] Therefore, in this embodiment, the shape of the streak artifact is regarded as a diamond, and the size in the depth direction of the voids or inclusions present inside the sample S is calculated. Although the shape of the streak artifact is, strictly speaking, different from a diamond, for ease of explanation, hereinafter the shape of the streak artifact will also be simply referred to as a "diamond."

[0039] Fig. 8 is a diagram for explaining a method for calculating the size of a diamond. A diamond-shaped streak artifact that appears in a reconstructed image of a linear CT can be decomposed into its components as shown in Fig. 8.

[0040] For this reason, first, the length c of the longer diagonal of the rhombus shown in Figure 8 is measured using known image processing technology. Then, the length a of the shorter diagonal of the rhombus shown in Figure 8 is measured using known image processing technology. Note that the acute angle θ of the rhombus shown in Figure 8 corresponds to the imaging angle θ of the linear CT shown in Figure 8. Therefore, the following equation (1) holds between the length c of the longer diagonal of the rhombus, the length a of the shorter diagonal of the rhombus, and the acute angle θ of the rhombus. Therefore, the dimension b of a predetermined point of the rhombus shown in Figure 8 is calculated according to the following equation (1).

[0041]

number

[0042] The following formula (2) holds between the length c of the longer diagonal of the rhombus, the depth length d of the void or inclusion in the rhombus, and the acute angle θ of the rhombus. Therefore, the depth length d of the void or inclusion in the rhombus shown in Figure 8 is calculated according to the following formula (2).

[0043]

number

[0044] In order to make the diamonds in the reconstructed image clearer, the diamonds in the reconstructed image may be detected by calculating the difference between an image of an ideal sample that does not contain voids or inclusions inside and an actually captured reconstructed image.

[0045] Fig. 9 shows an example of an image obtained by calculating the difference between an image of an ideal sample and an actually captured reconstructed image. As shown in Fig. 9, the diamond-shaped portion is emphasized, making it possible to more accurately calculate the depth length d of the void or inclusion within the diamond.

[0046] The values ​​of each diamond shown in Figure 9 are as follows:

[0047]

number

[0048] As described above, by identifying the diamonds in the CT images taken by the linear CT, it is possible to calculate the size of the voids or inclusions present inside the sample S. However, if a void or inclusion exists in the central part of the sample, the streak artifact will appear as a diamond, but if the void or inclusion exists in a part other than the central part of the sample, the streak artifact will appear as a shape other than a diamond.

[0049] FIG. 10 shows an example of a streak artifact that appears in a shape other than a diamond. The top image in FIG. 10 is a cross-sectional image of a true numerical phantom inside the sample, and the middle image in FIG. 10 is a CT image of the actual sample obtained by linear CT. As shown in the middle image in FIG. 10, when a void or inclusion is present in a location other than the center of the sample S, the streak artifact appears in a crescent shape. Although the shape of such a streak artifact is, strictly speaking, different from a crescent shape, for ease of explanation, hereinafter, the streak artifact shape shown in the middle image in FIG. 10 will also be simply referred to as a "crescent shape."

[0050] As will be described later, the CT image reconstruction device 14 of this embodiment detects diamond-shaped streak artifacts H1, H2, etc. as shown in the lower image of FIG. 10, rather than crescent-shaped streak artifacts as shown in the middle image of FIG. 10.

[0051] FIG. 11 is a diagram illustrating a stage G on which a sample S is placed in this embodiment. As shown in FIG. 11, in this embodiment, when a CT image of the sample S is reconstructed by linear CT, the sample S is placed on the stage G as shown in FIG. 11. The stage G includes a first member M1 and multiple second members M2. As shown in FIG. 11, the sample S and the stage G are placed on a conveyor belt 13 and move in a straight line in accordance with the movement of the conveyor belt 13. At this time, X-rays are irradiated from an X-ray source 11 toward the sample S and the stage G. A detector 12 (not shown in FIG. 11) detects a transmission image of X-rays transmitted through the sample S and the stage G as detection data. Each of the multiple second members M2 is made of a material with lower X-ray transmittance than the first member M1. Hereinafter, one second member will be simply referred to as a "sphere."

[0052] 12 to 14 are diagrams for explaining a method for irradiating X-rays toward the sample S and the stage G to obtain a transmission image.

[0053] As shown in Figure 12, generally, when reconstructing a CT image of a sample S, X-rays are irradiated from an X-ray source 11 when the sample S is positioned at J1 to J2, and a CT image of the inside of the sample S is reconstructed based on the detection data of the transmission image detected by a detector 12 at that time.

[0054] On the other hand, if it is desired to reconstruct a CT image of the right 1 / 3 of the sample S, it is necessary to reconstruct a CT image of the inside of the sample S based on the detection data of the transmission image obtained when the sample S is positioned at J3 to J4, as shown in Fig. 13. It can be seen that the transmission image used to reconstruct the CT image is shifted to the left side compared to the normal measurement shown in Fig. 13.

[0055] Furthermore, when it is desired to reconstruct a CT image of the left 1 / 3 of the sample S, it is necessary to reconstruct a CT image of the inside of the sample S based on the detection data of the transmission image obtained when the sample S is positioned at J5 to J6, as shown in Fig. 14. It can be seen that the transmission image used to reconstruct the CT image is shifted to the right compared to the normal measurement shown in Fig. 14.

[0056] In this embodiment, as shown in Figures 12 to 14, time-series data of the transmission image is reconstructed for each point in the linear direction of the sample S, with the point at the center, so that streak artifacts that appear in the CT image and are caused by voids or inclusions within the point are made diamond-shaped.

[0057] The diamond-shaped streak artifact H1 shown in the lower part of Fig. 10 is a CT image obtained from a transmission image centered on a void or inclusion present on the right side of the numerical phantom image in the upper part. Also, the diamond-shaped streak artifact H2 shown in the lower part of Fig. 11 is a CT image obtained from a transmission image centered on two voids or inclusions present on the right side of the numerical phantom image in the upper part.

[0058] Thus, with the conventional method, the streak artifacts of voids or inclusions, which appear as crescents as shown in the middle image of Figure 10, appear as diamonds, with the void or inclusion present on the right side of the sample at the center, as shown in the bottom image of Figure 10. Therefore, the depth dimension (y direction) of the void or inclusion can also be estimated from the diamond shape.

[0059] The positions of the spheres embedded in the stage G (the distance between the spheres in the linear direction) are known. Therefore, if the shape of an artifact present in the image of the sample S portion of a CT image reconstructed around a certain sphere is diamond-shaped, it can be determined that a void or inclusion exists inside the sample S at the position of that sphere.

[0060] In the examples shown in Figures 12 to 14, the sample S is divided into three equal parts in the traveling direction, and CT images are reconstructed with each part as the center. However, if the sample S is long in the x direction, which is the traveling direction, or if it is desired to examine the position of voids or inclusions in more detail, the above reconstruction may be performed by equally dividing the scan range into parts smaller than the number of parts into which the sample S is divided (or the spacing between spheres).

[0061] As described above, to obtain a CT image of the interior of the sample S, it is necessary to reconstruct the CT image from transmission images obtained in the linear CT from the point where the edge of the sample S enters the X-ray imaging range to the point where it leaves the imaging range. Therefore, as described above, in this embodiment, in order to determine whether the sample S is within the imaging range of the CT image and to determine which area of ​​the sample S is the center of the CT image reconstruction even if there are no voids or inclusions within the sample S, spheres made of a material with low X-ray transparency are embedded evenly from end to end inside the stage G, as shown in FIG. 11 . When a sample is placed on this stage G and measured using the linear CT, the first spheres can be seen in the transmission image, indicating that the sample S has entered and left the imaging range. This eliminates the need for empty imaging, thereby minimizing the calculation cost for reconstruction.

[0062] Furthermore, by referring to the cross sections in which the spheres appear in the multiple reconstructed CT images and observing the shape of the artifacts resulting from those shapes, it is possible to determine at what position the void or inclusion is located and is causing the diamond-shaped artifact to appear. Note that the number of spheres embedded in the stage G may be increased or decreased according to the number of equal divisions of the scan range.

[0063] Furthermore, as shown in FIG. 15, by embedding the sphere on the side of the stage G closer to the detector 12, the sphere embedded in the stage G can be detected in a wider scanning range.

[0064] Figure 16 is a diagram illustrating an image of a sphere portion and an image of a sample S portion in a CT image. Figure 16(a) is a numerical phantom image of a sphere embedded in stage G, Figure 16(b) is a CT image reconstructed from a transmission image obtained when sphere B1 embedded in stage G is the center, and Figure 16(c) is a CT image reconstructed from a transmission image obtained when sphere B2 embedded in stage G is the center. Figure 16(d) is a numerical phantom image of a void or inclusion present in the sample, Figure 16(e) is an image corresponding to the sample portion in the CT image when sphere B1 embedded in stage G is the center, and Figure 16(f) is an image corresponding to the sample portion in the CT image when sphere B2 embedded in stage G is the center.

[0065] As shown in Figure 16(e), the CT image was reconstructed around the central sphere B1 embedded in the stage G, and only the void or inclusion H1 located approximately directly above it appears as a diamond. On the other hand, in Figure 16(c), the CT image was reconstructed around the sphere B2, which is one sphere away from the central sphere B1, and so the void or inclusion H2 located directly above sphere B2 appears as a diamond. Therefore, if a diamond is detected in a CT image of the inside of a sample, it is possible to identify the location of the void or inclusion inside the sample by identifying which sphere the CT image was reconstructed around.

[0066] Each part of the CT image reconstruction device 14 will be described below.

[0067] The data storage unit 140 stores the detection data detected by the detector 12.

[0068] The reconstruction unit 142 reconstructs a CT image with a defect using a known method based on the detection data stored in the data storage unit 140. A CT image with a defect is a CT image obtained by irradiating a sample with X-rays from only some of the omnidirectional directions. Artifacts occur in a CT image with a defect. Specifically, the reconstruction unit 142 reconstructs a CT image for each sphere based on the detection data, with the sphere at the center. It is possible to identify which position on the sample the reconstructed CT image is centered around, based on the shape of the artifacts that appear in the CT image and are caused by the spheres embedded in the stage G.

[0069] The detection unit 144 detects an artifact shape determined according to a measurement system for linear CT from each of the multiple CT images for each sphere obtained by the reconstruction unit 142. Specifically, the detection unit 144 detects a diamond from each image of the sample S portion in the CT image in which the sphere is captured at the center.

[0070] The measurement system of this embodiment is a linear CT measurement system that generates a CT image by irradiating X-rays from one direction while the sample S travels straight ahead. As described above, the artifact shape determined by the linear CT measurement system is a diamond. For example, the detection unit 144 detects the diamond artifact shape using the following method.

[0071] <Example of image processing procedure for mechanically extracting diamonds> As described above, if it is possible to prepare a numerical phantom image that is an image of an ideal sample that does not contain voids or inclusions, it is possible to extract the diamond shape with higher accuracy using the numerical phantom image. Therefore, if it is possible to prepare a numerical phantom image, for example, the detection unit 144 detects the diamond shape, which is an artifact shape, using the following method.

[0072] [When the difference between the numerical phantom and the reconstructed image can be calculated] First, the detection unit 144 thins the lines in the difference image by performing erosion processing, a morphological operation, on the difference image between the numerical phantom image and the CT image with defects. Then, the detection unit 144 extracts the contours of the thinned difference image using the marching squares method or the like, thereby making it possible to highlight the locations of voids or inclusions.

[0073] If it is not possible to prepare a numerical phantom image, the detection unit 144 detects the diamond shape, which is an artifact shape, using the following method, for example.

[0074] [When the difference between the numerical phantom and the reconstructed image cannot be calculated] (Example 1) The detection unit 144 performs Canny binarization on the CT image with defects, and then extracts, by probabilistic straight-line Hough detection, a diamond-shaped shape or a line of a streak artifact that occurs depending on the imaging angle θ of the incomplete CT.

[0075] (Example 2) A template image containing a diamond shape expected to occur due to a streak artifact is prepared in advance. Then, the detection unit 144 scans the template against a CT image containing a defect and determines that a void or inclusion exists in a location where the similarity is higher than a threshold (so-called template matching).

[0076] (Example 3) The detection unit 144 performs semantic segmentation using deep learning on the CT image with defects and extracts diamond-shaped streak artifacts that are thought to contain voids.

[0077] The identifying unit 146 identifies the position (position in the x direction) of the sphere corresponding to the CT image in which the diamond is detected by the detecting unit 144 as the position (position in the x direction) of a void or inclusion present inside the sample S. Furthermore, as described above, the identifying unit 146 may calculate the length d of the void or inclusion in the diamond in the depth direction by identifying the length of each point of the diamond using a known image processing technique.

[0078] As described above, the spacing between the multiple spheres embedded in the stage G may be determined in advance according to the size of voids or inclusions that may exist inside the sample S. In addition, the number of spheres embedded in the stage G may be determined in advance according to the size of voids or inclusions that may exist inside the sample S. Note that, since the shape of a streak artifact is a diamond, the smallest size at which it can be detected is 3 voxels.

[0079] Next, the operation of the CT image reconstruction device 14 according to the embodiment will be described with reference to Fig. 17. Fig. 17 is a flowchart showing an example of the processing flow of the CT image reconstruction program according to the embodiment.

[0080] First, in step S100, the reconstruction unit 142 reconstructs a CT image having a defect using a known method based on the detection data stored in the data storage unit 140. Specifically, the reconstruction unit 142 reconstructs a CT image in which, for each sphere embedded in the stage G, the sphere is imaged at the center based on the detection data.

[0081] In step S102, the detection unit 144 detects a diamond from the Z cross-sectional image among the multiple CT images obtained in step S100.

[0082] In step S104, the identification unit 146 identifies the position (position in the x direction) of the sphere corresponding to the CT image in which the diamond was detected in step S102 as the position (position in the x direction) of a void or inclusion present inside the sample S.

[0083] In step S106, the identification unit 146 outputs the position of the diamond identified in step S104 as a result, and ends the process.

[0084] As described above, the CT image reconstruction system of the embodiment includes an X-ray source, a detector that detects X-rays emitted from the X-ray source and transmitted through a sample, and a CT (Computed Tomography) image reconstruction device. The X-ray source irradiates X-rays from some of all directions toward the sample and a stage on which the sample is placed. The stage includes a first member and a plurality of second members that are less X-ray transparent than the first member. The CT image reconstruction device reconstructs, based on detection data detected by the detector, CT images that include defects and in which the second member is centered for each second member. The CT image reconstruction device detects an artifact shape determined according to a measurement system representing the X-ray source and the detector from each of the plurality of CT images for each second member. The CT image reconstruction device identifies the position of an object present inside the sample based on the CT images in which the artifact shape is detected. This makes it possible to identify the position of the target area inside the sample when reconstructing a CT image of the inside of the sample using a measurement system that irradiates the sample with X-rays from some of the omnidirectional directions, even if the target area is located in a part other than the center of the sample.

[0085] Furthermore, as described above, when the measurement system is a measurement system for linear CT, the CT image reconstruction device detects an artifact shape determined according to the measurement system for linear CT from each image of the sample portion in a CT image in which the second member is at the center, and identifies the position of the second member corresponding to the CT image in which the artifact shape was detected as the position of an object present inside the sample. This makes it possible to identify the position of a void or inclusion inside the sample when capturing a CT image of the inside of the sample using linear CT.

[0086] Furthermore, the spacing between the multiple second members embedded in the stage is determined in advance according to the size of voids or inclusions that may exist inside the sample. Furthermore, the number of multiple second members embedded in the stage is determined in advance according to the size of voids or inclusions that may exist inside the sample. This allows the location of voids or inclusions that may exist inside the sample to be identified with greater accuracy.

[0087] As described above, the method for determining the position and depth (y-direction) dimensions of voids or inclusions from the shape of streak artifacts using elementary geometry requires that the resulting artifacts be diamond-shaped. However, artifacts generated by voids or inclusions at the left and right ends of the sample have shapes other than diamonds. In contrast, in this embodiment, by devising the method for photographing the sample and reconstructing the CT images, the artifacts generated by these voids or inclusions can be made diamond-shaped. This makes it possible to identify the position of voids or inclusions inside the sample and measure their depth (y-direction) dimensions.

[0088] In addition, by taking transmission images using linear CT, with the center set at an arbitrary position other than the center of the sample, and by changing the center in the reconstruction calculations based on those transmission images, it is possible to obtain a reconstructed image centered at any position on the sample. If there is a void or inclusion at that center, the streak artifacts that occur due to the sparse number of transmission images can be made into a symmetrical diamond shape.

[0089] [Other embodiments] The above describes an embodiment using a CT image reconstruction device as an example. The embodiment may be in the form of a program that causes a computer to function as each unit of the CT image reconstruction device. The embodiment may be in the form of a computer-readable storage medium that stores the program.

[0090] In the above embodiment, the incomplete CT is a linear CT, but the present invention is not limited to this. For example, the present invention can be applied to dimensional measurement in a direction with low resolution in the case of a CT measurement method called laminography or circular tomosynthesis, which is different from linear CT.

[0091] Furthermore, the configuration of the CT image reconstruction device described in the above embodiment is merely an example, and may be changed depending on the situation without departing from the spirit of the invention.

[0092] Furthermore, the processing flow of the program described in the above embodiment is merely an example, and unnecessary steps may be deleted, new steps may be added, or the processing order may be changed without departing from the scope of the invention. You can also swap the order.

[0093] In the above embodiment, the processing according to the embodiment is realized by a software configuration using a computer by executing a program, but the present invention is not limited to this. The embodiment may be realized by, for example, a hardware configuration or a combination of a hardware configuration and a software configuration.

[0094] Furthermore, the processes in the above-described embodiments may be stored as a program on a storage medium such as an optical disk and distributed.

[0095] In the above embodiment, the term "CPU" refers to a processor in a broad sense, including general-purpose processors and dedicated processors (e.g., GPU: Graphics Processing Unit, ASIC: Application Specific Integrated Circuit, FPGA: Field Programmable Gate Array, programmable logic device, etc.).

[0096] In this case, the operations of the processors may not only be performed by a single processor, but may also be performed by multiple processors located at physically separate locations working together. Furthermore, the order of the operations of the processors is not limited to the order described in the above embodiments, and may be changed as appropriate.

[0097] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[0098] (Implementation) The technology of the present disclosure may be implemented in the dependent relationships shown in the following appendices.

[0099] (Appendix 1) an X-ray source, a detector that detects X-rays that are irradiated from the X-ray source and transmitted through a sample, and a CT (Computed Tomography) image reconstruction device; A CT image reconstruction device in a CT image reconstruction system comprising: the X-ray source irradiates the sample and a stage on which the sample is placed with X-rays from some of all directions; the stage is configured to include a first member and a plurality of second members having lower X-ray transmittance than the first member, The CT image reconstruction device a reconstruction unit that reconstructs a CT image having a defect and in which the second member is imaged at the center for each of the second members based on the detection data detected by the detector; a detection unit that detects an artifact shape determined according to a measurement system representing the X-ray source and the detector from each of the plurality of CT images for each of the second members; an identifying unit that identifies a position of an object present inside the sample based on the CT image in which the artifact shape is detected; CT image reconstruction device including: (Appendix 2) When the measurement system is a measurement system for linear CT that causes the sample to travel in a straight line and irradiates X-rays from one direction to generate the CT image, the detection unit detects an artifact shape determined according to a measurement system related to the linear CT from each image of a sample portion in a CT image in which the second member is captured at the center; the identifying unit identifies the position of the second member corresponding to the CT image in which the artifact shape is detected as the position of an object present inside the sample. 10. The CT image reconstruction device according to claim 1. (Appendix 3) The identification unit further identifies the size of an object present inside the sample. CT image reconstruction device according to appendix 1 or appendix 2. (Appendix 4) The object present inside the sample is a void or a body; CT image reconstruction device according to any one of Supplementary Note 1 to Supplementary Note 3. (Appendix 5) The spacing between the plurality of second members embedded in the stage is determined in advance according to the size of the object that may be present inside the sample. 10. The CT image reconstruction device according to claim 1, wherein the CT image reconstruction device is a CT image reconstruction device. (Appendix 6) The number of the second members embedded in the stage is predetermined depending on the size of the object that may be present inside the sample. 10. The CT image reconstruction device according to claim 1, wherein the first and second sub-components are arranged in a plane parallel to each other. (Appendix 7) an X-ray source, a detector that detects X-rays that are irradiated from the X-ray source and transmitted through a sample, and a CT (Computed Tomography) image reconstruction device; A CT image reconstruction method executed by a CT image reconstruction device in a CT image reconstruction system comprising: the X-ray source irradiates the sample and a stage on which the sample is placed with X-rays from some of all directions; the stage is configured to include a first member and a plurality of second members having lower X-ray transmittance than the first member, The CT image reconstruction device reconstructing a CT image having a defect and in which the second member is imaged at the center for each of the second members based on the detection data detected by the detector; detecting an artifact shape determined according to a measurement system representing the X-ray source and the detector from each of the plurality of CT images for each of the second members; Identifying a position of an object present inside the sample based on the CT image in which the artifact shape is detected. CT image reconstruction method to perform processing. (Appendix 8) an X-ray source, a detector that detects X-rays that are irradiated from the X-ray source and transmitted through a sample, and a CT (Computed Tomography) image reconstruction device; A CT image reconstruction program executed by a CT image reconstruction device in a CT image reconstruction system comprising: the X-ray source irradiates the sample and a stage on which the sample is placed with X-rays from some of all directions; the stage is configured to include a first member and a plurality of second members having lower X-ray transmittance than the first member, A computer constituting the CT image reconstruction device includes: reconstructing a CT image having a defect and in which the second member is imaged at the center for each of the second members based on the detection data detected by the detector; detecting an artifact shape determined according to a measurement system representing the X-ray source and the detector from each of the plurality of CT images for each of the second members; Identifying a position of an object present inside the sample based on the CT image in which the artifact shape is detected. CT image reconstruction program for performing processing. [Explanation of symbols]

[0100] 10 CT image reconstruction system 11 X-ray source 12 Detectors 13 Conveyor Belt 14 CT image reconstruction device 140 Data storage unit 142 Reconstruction part 144 Detector 146 Specific part

Claims

1. an X-ray source, a detector that detects X-rays that are irradiated from the X-ray source and transmitted through a sample, and a CT (Computed Tomography) image reconstruction device; A CT image reconstruction device in a CT image reconstruction system comprising: the X-ray source irradiates the sample and a stage on which the sample is placed with X-rays from some of all directions; the stage is configured to include a first member and a plurality of second members having lower X-ray transmittance than the first member, The CT image reconstruction device comprises: a reconstruction unit that reconstructs, based on the detection data detected by the detector, a CT image having a defect and in which the second member is imaged at the center for each of the second members; a detection unit that detects an artifact shape determined according to a measurement system representing the X-ray source and the detector from each of the plurality of CT images for each of the second members; an identifying unit that identifies a position of an object present inside the sample based on the CT image in which the artifact shape is detected; A CT image reconstruction device comprising:

2. When the measurement system is a measurement system for linear CT that generates the CT image by irradiating the sample with X-rays from one direction while the sample is moved in a straight line, the detection unit detects an artifact shape determined according to a measurement system related to the linear CT from each image of a sample portion in a CT image in which the second member is located at the center; the identifying unit identifies a position of the second member corresponding to the CT image in which the artifact shape is detected as a position of an object present inside the sample.

2. The CT image reconstruction device according to claim 1.

3. The identification unit further identifies the size of an object present inside the sample.

3. A CT image reconstruction device according to claim 1.

4. The object present inside the sample is a void or a body; 3. A CT image reconstruction device according to claim 1.

5. The spacing between the plurality of second members embedded in the stage is determined in advance according to the size of the object that may be present inside the sample.

3. A CT image reconstruction device according to claim 1.

6. the number of the second members embedded in the stage is predetermined depending on the size of the object that may be present inside the sample; 3. A CT image reconstruction device according to claim 1.

7. an X-ray source, a detector that detects X-rays that are irradiated from the X-ray source and transmitted through a sample, and a CT (Computed Tomography) image reconstruction device; A CT image reconstruction method executed by a CT image reconstruction device in a CT image reconstruction system comprising: the X-ray source irradiates the sample and a stage on which the sample is placed with X-rays from some of all directions; the stage is configured to include a first member and a plurality of second members having lower X-ray transmittance than the first member, The CT image reconstruction device comprises: reconstructing a CT image having a defect and in which the second member is imaged at the center for each of the second members based on the detection data detected by the detector; detecting an artifact shape determined according to a measurement system representing the X-ray source and the detector from each of the plurality of CT images for each of the second members; identifying a position of an object present inside the sample based on the CT image in which the artifact shape is detected; A CT image reconstruction method for performing the process.

8. an X-ray source, a detector that detects X-rays that are irradiated from the X-ray source and transmitted through a sample, and a CT (Computed Tomography) image reconstruction device; A CT image reconstruction program executed by a CT image reconstruction device in a CT image reconstruction system comprising: the X-ray source irradiates the sample and a stage on which the sample is placed with X-rays from some of all directions; the stage is configured to include a first member and a plurality of second members having lower X-ray transmittance than the first member, A computer constituting the CT image reconstruction device includes: reconstructing a CT image having a defect and in which the second member is imaged at the center for each of the second members based on the detection data detected by the detector; detecting an artifact shape determined according to a measurement system representing the X-ray source and the detector from each of the plurality of CT images for each of the second members; identifying a position of an object present inside the sample based on the CT image in which the artifact shape is detected; A CT image reconstruction program for executing the process.

Images