X-ray talbot imaging device and x-ray talbot imaging method

Abstract

The present invention can make a decrease in the accuracy of the direction and position of a subject less prone to occur, and increase the accuracy thereof. An X-ray Talbot imaging device 100 has a radiation irradiation axis in a substantially vertical direction and a lattice plane in a substantially horizontal direction, the X-ray Talbot imaging apparatus 100 comprising a CT rotation shaft 132 to which an imaging jig 131 for fixing a subject in a predetermined direction is attached, and the CT rotation shaft 132 being rotated and fixed in a predetermined direction in a substantially horizontal plane.

Description

X-ray Talbot imaging device and X-ray Talbot imaging method

[0001] The present invention relates to an X-ray Talbot imaging device and an X-ray Talbot imaging method.

[0002] Conventionally, there has been known an X-ray Talbot imaging device using a Talbot interferometer or a Talbot-Lau interferometer, which is capable of photographing a subject while changing its orientation in two or three dimensions. The X-ray Talbot imaging device can visualize features of the subject that cannot be confirmed from a general photograph alone from images that have been subjected to CT reconstruction or image processing to generate a three-dimensional CT (computed tomography) image from photographed images photographed while changing the orientation of the subject.

[0003] For example, Patent Document 1 describes an X-ray Talbot imaging device that uses a hollow jig that can fix a subject in any orientation. In this X-ray Talbot imaging device, the orientation of the subject is changed in three dimensions and radiation is irradiated horizontally onto the subject, thereby performing X-ray Talbot imaging.

[0004] U.S. Patent No. 11,143,603

[0005] To reconstruct a CT image from an image captured by an X-ray Talbot imaging device, the subject must be fixed to a jig capable of changing the subject's orientation, the jig must be placed at the tip of the CT rotation axis, and the image must be captured while the CT rotation axis is rotated. Because the CT rotation axis is rotatable, some play in the axis is unavoidable. Furthermore, repeated rotation of the CT rotation axis for imaging increases the play. Play in the CT rotation axis reduces the accuracy of the subject's orientation and position, thereby reducing the accuracy of the reconstructed image. In particular, the jig for the X-ray Talbot imaging device described in Patent Document 1 is placed on a CT rotation axis that extends vertically. Therefore, if play occurs in the CT rotation axis, the CT rotation axis and the jig may tilt, even in a random direction, further reducing the accuracy of the subject's orientation and position.

[0006] Therefore, an object of the present invention is to prevent the accuracy of the orientation and position of a subject from decreasing and to increase the accuracy of these.

[0007] In order to solve the above problems, the present invention provides an X-ray Talbot imaging device having a radiation irradiation axis in a substantially vertical direction and a grid surface in a substantially horizontal direction, the X-ray Talbot imaging device including a CT rotation axis to which an imaging jig for fixing a subject in a predetermined orientation is attached, the CT rotation axis being rotated in a predetermined orientation within a substantially horizontal plane and fixed. The imaging jig defines a direction orthogonal to the CT rotation axis as a first axis, and changes and fixes the orientation of the subject around the first axis. Alternatively, the imaging jig changes the orientation of the subject around the first axis, and then defines a direction orthogonal to the first axis as a second axis, and changes and fixes the orientation of the subject around the second axis.

[0008] Furthermore, an X-ray Talbot imaging method of the present invention is an X-ray Talbot imaging method using the above-mentioned X-ray Talbot imaging device, and includes: an imaging step of changing the orientation of the subject in two directions and rotating the CT rotation axis to obtain a plurality of images in each of the two directions; and a generation step of generating a tomographic image for each of the directions.

[0009] Further, an X-ray Talbot imaging method of the present invention is an X-ray Talbot imaging method using the above-mentioned X-ray Talbot imaging device, comprising: an imaging step of changing the orientation of the subject in a plurality of directions and rotating the CT rotation axis to obtain a plurality of images in a plurality of directions, and a generation step of generating a tomographic image for each of the directions.An X-ray Talbot imaging method using the above-mentioned X-ray Talbot imaging device, comprising: an imaging step of changing the orientation of the subject in a plurality of directions and rotating the CT rotation axis to obtain a plurality of images in a plurality of directions, and a generation step of generating a tomographic image for the plurality of directions collectively.

[0010] According to the present invention, the accuracy of the orientation and position of a subject can be prevented from decreasing, and these accuracies can be improved.

[0011] 3B is a diagram showing an example of the configuration of an X-ray Talbot imaging device. A plan view of a multi-slit. A perspective view of a subject table. A perspective view of a subject table. A cross-sectional view taken along line IV-IV in FIG. 3B. A perspective view of an imaging jig. A perspective view of an imaging jig. A schematic view of a rotating section and a holding section of the imaging jig. A block diagram showing the functional configuration of a controller. A diagram explaining the principle of a Talbot interferometer. An example of a subject. An example of a tomographic image. An example of a tomographic image. An example of a subject. An example of a weld line portion. A diagram showing fiber direction. An example of a scattering intensity tomographic image. An example of an orientation degree tomographic image. An example of an orientation angle tomographic image. A flowchart showing imaging processing.

[0012] Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, although the embodiments described below are subject to various technically preferable limitations for carrying out the present invention, the technical scope of the present invention is not limited to the following embodiments and illustrated examples.

[0013] (Configuration of Radiation Imaging System) FIG. 1 is a diagram schematically showing an X-ray Talbot imaging device 100 according to an embodiment of the present invention.

[0014] As shown in Fig. 1, the X-ray Talbot imaging device 100 includes a main body 1 and a controller 5. As shown in Fig. 1, the main body 1 includes a Talbot-Lau interferometer having a radiation source 11, a multi-slit 12, a subject table 13, a first grating 14, a second grating 15, and a radiation detector 16. The Talbot-Lau interferometer of the main body 1 is of a vertical type, and the radiation source 11, the multi-slit 12, the subject table 13, the first grating 14, the second grating 15, and the radiation detector 16 are arranged in this order in the z direction, which is a substantially vertical direction.

[0015] The radiation source 11 includes an X-ray tube that generates X-rays and irradiates them in the z-direction. Examples of the X-ray tube include a Coolidge X-ray tube and a rotating anode X-ray tube. The anode can be made of tungsten or molybdenum. The focal diameter of the radiation source 11 is preferably 0.03 to 3 mm, more preferably 0.1 to 1 mm. In the case of a vertical Talbot-Lau interferometer such as that shown in FIG. 1 , the radiation irradiation axis direction is substantially vertical, and the optical axis of the Talbot-Lau interferometer is indicated by 17 in FIG. 1 . The optical axis 17 is an imaginary line that passes substantially perpendicularly through the multi-slit 12, the first grating 14, the second grating 15, and the approximate center of the image area of ​​the radiation detector. While this embodiment will be described taking an example in which imaging is performed using X-rays, other radiation, such as neutron rays or gamma rays, may also be used.

[0016] The multi-slit 12 is a diffraction grating, and as shown in FIG. 2 , multiple slits are arranged at predetermined intervals in the x-direction, which is perpendicular to the z-direction, which is the radiation irradiation axis direction. The multi-slit 12 is formed of a material with high radiation shielding power, i.e., high radiation absorption, such as tungsten, lead, or gold, on a substrate made of a material with low radiation absorption, such as silicon or glass. Therefore, the multi-slit 12 may have a metallic luster. For example, a resist layer is masked in the shape of slits by photolithography, and UV light is irradiated to transfer the slit pattern onto the resist layer. A slit structure with the same shape as the pattern is obtained by exposure, and metal is filled between the slit structures by electroforming, thereby forming the multi-slit 12.

[0017] The slit period (grating period) of the multi-slit 12 is 1 to 60 (μm). As shown in FIG. 2, one slit period is the distance between adjacent slits. The slit width (the length of each slit in the slit period direction (x direction)) is 1 to 60 (%) of the slit period, and more preferably 10 to 40 (%). The slit height (height in the z direction) is 1 to 1500 (μm), and preferably 30 to 1000 (μm). Note that the multi-slit 12 may be omitted if the focus of the X-ray tube of the radiation source 11 is small, such as in a microfocus radiation source.

[0018] The subject table 13 includes an imaging jig 131, a CT rotation axis 132, a CT rotation axis support and drive unit 133, a rotation ring 134, a rotation base 135, a gear 136, a motor 137, a base unit 138, and the like. FIG. 3 is a perspective view of the subject table 13. FIG. 4 is a cross-sectional view taken along line IV-IV in FIG. 3. The imaging jig 131 is a jig that fixes the subject H in a predetermined orientation. Details will be described later. The CT rotation axis 132 is an axis on which the imaging jig 131 is attached. In the example of FIG. 1, the imaging jig 131 is attached to the tip of the CT rotation axis 132. The imaging jig 131 is rotatable around the CT rotation axis 132. The axial direction of the CT rotation axis 132 is approximately horizontal (y-axis direction). As shown in FIG. 4 , the CT rotation shaft support and drive unit 133 has a hole 133a through which the CT rotation shaft 132 passes, and supports the CT rotation shaft 132 by passing the CT rotation shaft 132 through the hole 133a. The hole 133a through which the CT rotation shaft 132 passes may penetrate the CT rotation shaft support and drive unit 133, or may be deep enough to support the CT rotation shaft 132. The CT rotation shaft 132 is configured to rotate when power from a CT rotation shaft motor MO provided in the CT rotation shaft support and drive unit 133 is transmitted to the CT rotation shaft 132 via a gear. When the hole 133a penetrates the CT rotation shaft support and drive unit 133, a weight W may be attached to the other end of the CT rotation shaft 132 opposite the end to which the imaging jig 131 is fixed. The weight of the weight W is preferably approximately the same as the combined weight of the imaging jig 131 and the subject H. By providing a weight W on the CT rotation shaft 132, the weights on the left and right sides of the plane of FIG. 4 become approximately equal. By attaching the weight W as a counterweight on the side opposite the subject H, the height of the rotation center of the CT rotation shaft 132 is less likely to change. The rotating ring 134 (rotation shaft rotating unit) supports the CT rotation shaft support / drive unit 133 and rotates in a plane approximately parallel to the grating plane around the optical axis 17, which is parallel to the z-axis direction. The rotating ring 134 also rotates and fixes the CT rotation shaft 132 in a predetermined direction in an approximately horizontal plane. The state in FIG. 3B is a state in which the rotating ring 134 has been rotated 90° clockwise as viewed from the radiation source 11 from the state in FIG. 3A. A groove 134a is provided on the side of the rotating ring 134 to receive power from a gear 136, which will be described later.The rotating base 135 is a base that supports the rotating ring 134. The rotating ring 134 and the rotating base 135 are fitted together to the extent that the rotating ring 134 can rotate. The gear 136 is connected to the motor 137 (described later) and transmits power from the motor 137 to the rotating ring 134. The motor 137 generates power to rotate the rotating ring 134. The base 138 supports the various components described above. As described above, the CT rotation shaft 132 is substantially horizontal, and therefore gravity acts perpendicular to the axial direction of the CT rotation shaft 132. Therefore, the CT rotation shaft 132 is stable and does not easily tilt within the horizontal plane or change in height. In addition, even if there is a difference in diameter between the CT rotation shaft 132 and the hole 133a through which the CT rotation shaft 132 passes (even if there is play), gravity acts perpendicular to the axial direction of the CT rotation shaft 132, thereby providing stability. If the diameter of the hole 133a through which the CT rotation shaft 132 passes expands due to repeated rotation of the CT rotation shaft 132, causing play, fine adjustments can be made by providing a mechanism that changes the angle of the CT rotation shaft support and drive unit 133 relative to the horizontal plane. Fine adjustments can also be made by inserting a shim between the CT rotation shaft support and drive unit 133 and the rotating ring 134. The rotating ring 134 is supported on the rotating base 135 in a substantially horizontal direction. Therefore, gravity acts vertically on the rotating ring 134, stabilizing the position of the rotating ring 134. This in turn stabilizes the CT rotation shaft 132, preventing it from shifting or tilting in the horizontal plane or changing in height.

[0019] The imaging jig 131 will now be described in detail with reference to FIGS. 5 and 6 . The imaging jig 131 (131a) shown in FIG. 5 is capable of fixing the subject H by changing the orientation of the subject H around a first axis, which is a direction perpendicular to the CT rotation axis 132 (here, the z-axis). The imaging jig 131 (131b) shown in FIG. 6 is capable of fixing the subject H by changing the orientation of the subject H around the first axis, as well as fixing the subject H by changing the orientation of the subject H around a second axis, which is a direction perpendicular to the first axis (here, the x-axis). The imaging jig 131 has a rotating unit R that fixes the subject H and is rotatable around a predetermined axis, and a holding unit K that holds the rotating unit R. The rotating unit R and the holding unit K are provided with positioning units L, and by aligning the positioning units L, the subject H is fixed in a predetermined orientation. In FIGS. 5 and 6 , the positioning portion L is a convex portion L1 provided on one of the rotating portion R and the holding portion K, and a concave portion L2 provided on the other of the rotating portion R and the holding portion K. The convex portion L1 and the concave portion L2 are aligned to fix the subject H in a predetermined orientation. As shown in FIG. 7 , the positioning portion L may be a pin L3 provided on one of the rotating portion R and the holding portion K, and a concave portion L4 provided on the other of the rotating portion R and the holding portion K. The pin L3 is movable back and forth (radially). When the rotating portion R rotates, it is retracted into the rotating portion R, and when the rotating portion R is fixed, it is aligned with the concave portion L4. The pin L3 may be moved back and forth manually, or it may be retracted automatically when the user rotates the rotating portion R. The rotating portion R has a base P on which the subject H is placed and fixed. For example, the subject H is fixed to the base P using an adhesive. Furthermore, for example, the subject H may be fixed using a chucking mechanism provided on the base P. As shown in FIG. 5 , markers M having a density greater than that of the base P may be attached to the base P. For example, the shape of the markers M may be triangular, rectangular, circular, or spherical. Since there is a density difference between the markers M and the base P, the markers M appear in the generated two-dimensional projection image or three-dimensional CT image, as described below. Therefore, the markers M can be used to align the three-dimensional CT image when generating a three-dimensional orientation analysis image.Specifically, the 3D CT images are aligned by rotating the image in the opposite direction by the angle of rotation of the subject H, and then fine-tuning and correcting the position and orientation of the image so that the subject H overlaps on the image (image alignment). The imaging jigs 131a and 131b have high mechanical accuracy and change the orientation of the subject H to increase the accuracy of the orientation of the subject H. Here, the marker M is used to achieve higher accuracy than mechanical accuracy or when accuracy decreases over time. The marker M may also be used to monitor whether accuracy has decreased over time. A specific method for using the marker M will be described later.

[0020] The imaging jig 131 is made of a uniformly low-density material such as resin or light metal. Note that resin does not include resin containing fillers, such as fiber-reinforced resin. Light metals include, for example, aluminum, magnesium, and alloys thereof. By making the imaging jig 131 using such a lightweight material, the CT rotation axis 132 can be made less likely to bend. Furthermore, by making the imaging jig 131 using such a material, these materials are less likely to absorb or scatter X-rays, so the visibility of the subject H is not hindered in various images, such as the reconstructed images described below.

[0021] Like the multi-slit 12, the first grating 14 is a diffraction grating in which a plurality of slits are arranged in the x-direction perpendicular to the z-direction, which is the radiation irradiation axis direction. Like the multi-slit 12, the first grating 14 may be formed by UV photolithography. Alternatively, the first grating 14 may be formed solely from silicon by deep-drilling a silicon substrate with fine wires using the so-called ICP method. The slit period of the first grating 14 is 1 to 20 μm. The slit width is 20 to 70% of the slit period, preferably 35 to 60%. The slit height is 1 to 100 μm.

[0022] Similar to the multi-slit 12, the second grating 15 is a diffraction grating having a plurality of slits arranged in the x-direction perpendicular to the z-direction, which is the radiation irradiation axis direction. The second grating 15 can also be formed by photolithography. The slit period of the second grating 15 is 1 to 20 μm. The slit width is 30 to 70% of the slit period, preferably 35 to 60%. The slit height is 1 to 100 μm. A moving mechanism 15a that moves the second grating 15 in the x-direction is provided adjacent to the second grating 15. The moving mechanism 15a may have any configuration as long as it can linearly move the second grating 15 in the x-direction by driving a motor or the like.

[0023] The radiation detector 16 has conversion elements arranged two-dimensionally, which generate electrical signals in response to irradiated radiation, and reads the electrical signals generated by the conversion elements as image signals. The pixel size of the radiation detector 16 is 10 to 300 (μm), and more preferably 50 to 200 (μm). The radiation detector 16 is preferably fixed to a base portion so as to abut against the second grating 15. This is because the greater the distance between the second grating 15 and the radiation detector 16, the more blurred the moiré fringe image obtained by the radiation detector 16.

[0024] A flat panel detector (FPD) can be used as the radiation detector 16. FPDs include an indirect conversion type that converts radiation into an electrical signal using a photoelectric conversion element via a scintillator, a direct conversion type that directly converts radiation into an electrical signal, and a photon counting type that counts each photon of radiation. Any of these types may be used. Alternatively, a radiation detector that provides the intensity modulation effect of the second grating 15 may be used as the radiation detector 16. For example, a slit scintillator detector may be used as the radiation detector 16. Grooves may be carved in the scintillator to provide blind areas in the scintillator with the same period and width as the slits in the second grating 15, forming a grid-shaped scintillator. In this case, the radiation detector 16 combines the functions of both the second grating 15 and the radiation detector 16, eliminating the need for a separate second grating 15. In other words, providing a slit scintillator detector is the same as providing both the second grating 15 and the radiation detector 16.

[0025] The Talbot-Lau interferometer of the main body 1 has been described as being configured to irradiate X-rays from the radiation source 11 provided on the upper side toward the subject H located below (so-called vertical type). However, the Talbot-Lau interferometer of the main body 1 is not limited to this, and may be configured to irradiate X-rays from the radiation source 11 provided on the lower side toward the subject H located above.

[0026] As shown in FIG. 8 , the controller 5 includes a control unit 51, an operation unit 52, a display unit 53, a communication unit 54, and a storage unit 55. The control unit 51 includes a CPU (Central Processing Unit), a RAM (Random Access Memory), and the like. The control unit 51 is connected to each unit of the main body 1 (e.g., the radiation source 11, the radiation detector 16, the moving mechanism 15a, and the like) and controls the operation of each unit. The control unit 51 also executes various processes, including an imaging process described below, in cooperation with a program stored in the storage unit 55. The control unit 51 functions as an imaging unit that changes the orientation of the subject in multiple directions and rotates the rotation axis to obtain multiple images in multiple directions. The captured images include moiré fringe images (described later), two-dimensional projection images (absorption images, small-angle scattering images, differential phase images), three-dimensional CT images (absorption tomographic images, small-angle scattering tomographic images, phase tomographic images), etc. The control unit 51 functions as a generation unit that generates tomographic images for each direction or all at once.

[0027] The operation unit 52 includes a group of keys used for inputting the exposure switch, imaging conditions, etc., as well as a touch panel integrated with the display of the display unit 53, and generates operation signals according to these operations and outputs them to the control unit 51. The display unit 53 displays an operation screen, the operating status of the main body 1, etc. on the display in accordance with the display control of the control unit 51.

[0028] The communication unit 54 includes a communication interface and communicates with external devices on the network.

[0029] The storage unit 55 is configured with a non-volatile semiconductor memory, a hard disk, or the like, and stores various information such as programs executed by the control unit 51 and data required for executing the programs. The storage unit 55 may also be provided with a DB (Data Base) separate from the controller 5, and various information may be managed in the DB.

[0030] (Photography Using a Talbot Interferometer and a Talbot-Lau Interferometer) Here, a photographing method using a Talbot interferometer and a Talbot-Lau interferometer will be described. Photographing using a Talbot interferometer or a Talbot-Lau interferometer is referred to as Talbot photographing. Talbot photographing also includes Talbot CT (computed tomography) photographing, which will be described later. As shown in FIG. 9 , when X-rays irradiated from a radiation source 11 pass through a first grating 14, the transmitted X-rays form images at regular intervals in the z direction. These images are called self-images, and the phenomenon of forming self-images is called the Talbot effect. A second grating 15 is disposed at the position where the self-images are formed, generally parallel to the self-images. A Moiré fringe image (indicated by Mo in FIG. 9 ) is obtained by the X-rays transmitted through the second grating 15. That is, the first grating 14 forms a periodic pattern, and the second grating 15 converts the periodic pattern into Moiré fringes. If a subject H is present between the radiation source 11 and the first grating 14, the subject H will cause a phase shift in the X-rays, and the moiré fringes on the moiré fringe image will be distorted at the edge of the subject H, as shown in Figure 9. This distorted moiré fringe can be detected by processing the moiré fringe image, and the subject image can be visualized. This is the principle of the Talbot interferometer.

[0031] In the main body 1, a multi-slit 12 is disposed between the radiation source 11 and the first grating 14 and close to the radiation source 11, and X-ray imaging is performed using a Talbot-Lau interferometer. The Talbot interferometer is based on the premise that the radiation source 11 is an ideal point radiation source, but in actual imaging, a focal spot with a relatively large focal spot diameter is used, and therefore the multi-slit 12 produces an effect as if X-rays were being irradiated from a series of multiple point radiation sources. This is the X-ray imaging method using a Talbot-Lau interferometer, and even when the focal spot diameter is relatively large, a Talbot effect similar to that of a Talbot interferometer can be obtained.

[0032] Two-dimensional projection images generated by processing the fringe scanning method based on the moiré fringe image include absorption images, differential phase images, and small-angle scattering images. Absorption images visualize the absorption of X-rays by the subject and are equivalent to conventional plain X-ray images. Differential phase images visualize the refraction of X-rays by the subject, with the signal value increasing as the X-ray refraction increases. Absorption images have lower sensitivity for lighter elements, but differential phase images maintain high sensitivity even for light elements, making it possible to capture changes in materials that are difficult to capture with absorption images. Small-angle scattering images visualize the scattering of X-rays in microstructures, with the signal value increasing as the X-ray scattering increases. Small-angle scattering images can capture microstructure assemblies of several to tens of micrometers, which are smaller than the pixel size.

[0033] In addition, image processing can be performed on the three types of 2D projection images described above to generate even more types of images. For example, small-angle scattering images captured at multiple (three or more) grating facing angles can be used. By aligning each image, i.e., rotating the image in reverse by the angle corresponding to the grating facing angle, and then fine-tuning the position and orientation of the object on the image, a sine wave fitting can be performed for each pixel, and fitting parameters can be extracted. The sine wave graph has the relative angle between the sample and the grating on the horizontal axis and the small-angle scattering signal value of a given pixel on the vertical axis. The amplitude, average, and phase of the sine wave are obtained as fitting parameters. The image representing the amplitude value for each pixel is called the orientation image, the image showing the average value for each pixel is called the scattering intensity image, and the image showing the phase for each pixel is called the orientation angle image. Note that the fitting method is not limited to a sine wave. Hereinafter, the orientation image, scattering intensity image, and orientation angle image, which are two-dimensional images generated by performing image processing on the 2D projection images, are collectively referred to as the orientation analysis image.

[0034] (CT Imaging Using a Talbot Interferometer and a Talbot-Lau Interferometer) The above describes Talbot imaging, which generates various two-dimensional images using a Talbot interferometer and a Talbot-Lau interferometer. Here, we will describe Talbot CT imaging, which acquires various three-dimensional images (tomographic images) using a Talbot-Lau interferometer. Talbot CT imaging is a three-dimensional extension of Talbot imaging, which generates various two-dimensional images. In Talbot CT imaging, the control unit 51 rotates the CT rotation axis 132 by a predetermined angle (e.g., 1°) to acquire Moiré fringe images for 180° or 360°, and performs fringe scanning processing to generate two-dimensional projection images (absorption image, small-angle scattering image, differential phase image). Next, the control unit 51 calculates the signal value of each voxel using the two-dimensional projection images (absorption image, small-angle scattering image, differential phase image) to generate three-dimensional CT images (absorption tomographic image, small-angle scattering tomographic image, phase tomographic image) corresponding to the two-dimensional projection images. The control unit 51 also generates three-dimensional orientation analysis images (orientation degree tomographic image, orientation angle tomographic image, and scattering intensity image) by performing CT reconstruction and image processing using small-angle scattering images captured in multiple directions relative to the CT rotation axis 132. Note that if the subject H contains a highly anisotropic structure, such as a fiber-reinforced resin, artifacts are likely to occur in the small-angle scattering tomographic image. In this case, when generating a small-angle scattering tomographic image or a three-dimensional orientation analysis image based on the small-angle scattering tomographic image, it is desirable to orthogonalize the CT rotation axis 132 to the lattice direction (y direction in FIG. 2 ). Note that this is because, if the subject contains a highly anisotropic structure, if the CT rotation axis 132 is parallel to the lattice direction, the scattering signal detected by the lattice in the highly anisotropic structure will change as the CT rotation axis rotates, resulting in artifacts in the small-angle scattering tomographic image. When the CT rotation axis 132 is oriented perpendicular to the grid direction, the scattering signal detected by the grid is less likely to change, and artifacts are less likely to occur in the small-angle scattering tomographic image (U.S. Patent No. 11,143,603). On the other hand, the differential phase image used to generate a phase tomographic image is an image differentiated in a direction perpendicular to the grid direction (y direction in FIG. 2). Therefore, when generating a phase tomographic image, a process of integrating the differential phase image in a direction perpendicular to the grid direction is performed in conjunction with the reconstruction.Therefore, the CT rotation axis must be set to the lattice direction (y direction in FIG. 2 ). Therefore, when generating phase tomographic images and small-angle scattering tomographic images if the subject H, such as a fiber-reinforced resin, does not contain a highly anisotropic structure, the CT rotation axis is set to the lattice direction (y direction in FIG. 2 ). On the other hand, when generating phase tomographic images, small-angle scattering tomographic images, or three-dimensional orientation analysis images if the subject H, such as a fiber-reinforced resin, contains a highly anisotropic structure, both imaging with the CT rotation axis set to the lattice direction and imaging with the CT rotation axis set to a direction perpendicular to the lattice direction (x direction in FIG. 2 ) are performed. Therefore, the orientation of the CT rotation axis 132 must be changed frequently, and it is important to maintain accuracy when changing the orientation of the CT rotation axis.

[0035] Here, a specific example of Talbot CT imaging will be described. (Case 1) Assume that subject H is an object with a clear fiber direction, as shown in FIG. 10 . The fibers of subject H shown in FIG. 10 are woven in the long-side direction LD and the short-side direction SS. The fiber direction of subject H shown in the top view of FIG. 10 is the long-side direction LD. Talbot imaging is performed using the imaging jig 131 shown in FIG. 5 in two directions (0° / 90°) with the long-side direction LD and the short-side direction SS aligned with the lattice direction. This results in small-angle scattering tomographic images for each direction based on multiple two-dimensional projection images obtained by processing multiple moiré fringe images in the two directions using a fringe scanning method. These two-directional images are then averaged to obtain scattering intensity tomographic images representing the scattering intensity independent of direction. FIG. 11A shows small-angle scattering tomographic images (cross-section, side, and top) when the lattice direction is aligned with the long-side direction LD. 11B shows small-angle scattering tomographic images (cross section, side, top) when the lattice direction is aligned with the short side direction SS. In Case 1, as shown in FIG. 11A and B, orientation information in the direction (0° / 90° direction) corresponding to the lattice direction of the subject H is obtained.

[0036] More specifically, in the case of the imaging example (Case 1), the rotating ring 134 is rotated so that the orientation of the CT rotation axis 132 is perpendicular to the grid direction. Then, using the imaging jig 131 shown in FIG. 5 , imaging is performed in two directions (0° / 90°) with the long side direction LD and the short side direction SS aligned with the grid direction, followed by processing using the fringe scanning method. CT reconstruction is performed using a method known as the Feldkamp method, which can be solved analytically without iterative calculations. Another CT reconstruction method may be an iterative approximation method, in which a desired 3D CT image is created from a certain initial value, a forward projection (predictive calculation as if an actual image had been captured) is performed on this image to obtain a 2D projection image, and the 3D CT image is corrected based on the deviation from the actual 2D projection image. This correction is then sequentially and iteratively calculated to obtain the desired 3D CT image. As described above, after CT reconstruction of small-angle scattering tomographic images of a 3D CT image is performed from the two-dimensional projection images of the small-angle scattering images using these methods, one of the small-angle scattering tomographic images is rotated 90° so that the orientations of the resulting 0° / 90° small-angle scattering tomographic images are the same. This small-angle scattering tomographic image is the image shown in FIG. 11 . Note that, if markers M are attached to the base P of the imaging jig 131 in these small-angle scattering tomographic images, the marker positions on the images will approximately match, but this is due to mechanical precision. To achieve higher precision than mechanical precision, or in cases where precision decreases over time, for one of these small-angle scattering tomographic images, the entire small-angle scattering tomographic image is shifted in the required direction and position in any of x, y, z, θx, θy, and θz, and the image is adjusted so that the marker positions match. This image registration using markers may be performed on any of the 3D CT images: small-angle scattering tomographic images, phase tomographic images, and absorption tomographic images. Furthermore, because there is a density difference between the marker M and the base P, it is more preferable to use an absorption tomographic image obtained from an absorption image that is more likely to produce contrast. In this case, the shift amounts in x, y, z, θx, θy, and θz required to match the positions of the markers in the absorption tomographic image are determined, and the phase tomographic image and the absorption tomographic image are shifted by these shift amounts. Furthermore, by averaging the images in these two directions, a scattering intensity tomographic image that represents the scattering intensity independent of direction can be obtained.While Figure 11 shows a small-angle scattering tomographic image, absorption tomographic images of 3D CT images can also be obtained from absorption images of 2D projection images in a similar manner. Since absorption images do not detect the anisotropy of the subject, CT reconstruction is performed from 2D projection images in only one of the above-mentioned 0° / 90° directions. Furthermore, after rotating the rotating ring 134 so that the CT rotation axis 132 is parallel to the lattice direction, 2D projection images of differential phase images are acquired, and then integration in a direction perpendicular to the lattice direction is performed in conjunction with the CT reconstruction to obtain phase tomographic images of 3D CT images. Since differential phase images do not detect the anisotropy of the subject, unidirectional imaging is performed, with either the long side direction (LD) or the short side direction (SS) aligned with the lattice direction.

[0037] (Case 2) Talbot photography is performed using the photography jig 131 shown in Figure 5, with the orientation of the object H relative to the lattice direction in three directions (0° / 60° / 120°) or four directions (0° / 45° / 90° / 135°) in a plane parallel to the lattice plane. In Case 2, unlike Case 1, two-dimensional projection images are obtained by fringe scanning processing based on moiré fringe images in multiple directions, and then CT reconstruction and image processing are performed for each direction or all at once to obtain small-angle scattering tomographic images, scattering intensity tomographic images, orientation angle tomographic images, and orientation degree tomographic images. Figure 12A shows the object. The object is a test piece made of carbon fiber and PC (polycarbonate) created by injection molding. An image of a weld line (Figure 12B) that occurred where resin flows from two directions collide was taken. Figure 12C is a diagram showing the fiber direction in an orientation angle tomographic image. As shown in the scattering intensity tomographic image of Fig. 12D, the orientation degree tomographic image of Fig. 12E, and the orientation angle tomographic image of Fig. 12F, it is possible to obtain fiber orientation information that changes in the thickness direction of the weld line portion of the test piece, which is subject H. In case 2, as shown in the orientation angle tomographic image, it is possible to obtain continuous orientation information in a plane parallel to the lattice plane of subject H.

[0038] More specifically, in the case of the imaging example (Case 2), the rotation ring 134 is rotated so that the orientation of the CT rotation axis 132 is perpendicular to the lattice direction. Then, using the imaging jig 131 shown in FIG. 5 , imaging is performed in a plane parallel to the lattice plane, with the orientation of the subject H relative to the lattice direction set to three directions (0° / 60° / 120°) or four directions (0° / 45° / 90° / 135°), and then processing using the fringe scanning method is performed. CT reconstruction and image processing can be performed either by reconstructing images for each imaging direction to generate a tomographic image, or by reconstructing images simultaneously from images taken in each direction to generate a tomographic image.

[0039] First, a method for generating tomographic images by reconstructing images for each imaging direction will be described. CT reconstruction can be performed using a method called the Feldkamp method, which allows for analytical solutions without iterative calculations. Another CT reconstruction method may be an iterative approximation method, in which the desired 3D CT image is created starting from a certain initial value, a forward projection (predictive calculation as if actual imaging had been performed) of this image is performed to obtain a 2D projection image, the 3D CT image is corrected based on the deviation from the actually captured 2D projection image, and this is sequentially and iteratively calculated to obtain the desired 3D CT image. These methods allow CT reconstruction of a small-angle scattering tomographic image of a 3D CT image from a 2D projection image of a small-angle scattering image. Thereafter, using one of the obtained small-angle scattering tomographic images in three directions, or one of the obtained small-angle scattering tomographic images in four directions as a reference, the orientations of the other small-angle scattering tomographic images are rotated so that the orientation of the subject in this small-angle scattering tomographic image is the same as the orientation of the subject in the other small-angle scattering tomographic images. That is, when the 0° direction is used as the reference in three-direction imaging, the orientations of the small-angle scattering tomographic images for the remaining two images are rotated by -60° and -120°, respectively. Note that if markers M are attached to the base P of the imaging jig 131 in these small-angle scattering tomographic images, the marker positions will approximately match on the images, but this is due to mechanical precision. To achieve higher accuracy than mechanical accuracy, or in cases where accuracy decreases over time, these small-angle scattering tomographic images are rotated in the 60° or 120° direction, and the entire small-angle scattering tomographic image is shifted in the required direction and position in any of x, y, z, θx, θy, and θz to adjust the image so that the marker positions match. This image alignment using markers may be performed on any 3D CT image, including small-angle scattering tomographic images, phase tomographic images, and absorption tomographic images. Because there is a density difference between the marker M and the base P, it is preferable to use an absorption tomographic image obtained from an absorption image that is more likely to have contrast. In this case, the shift amounts in x, y, z, θx, θy, and θz required to match the marker positions in the absorption tomographic image are determined, and the phase tomographic image and absorption tomographic image are shifted by these shift amounts. The same applies to four-directional imaging.Furthermore, using the small-angle scattering tomographic image and a rotated small-angle scattering tomographic image, the following image processing is performed on each tomographic plane parallel to the lattice plane to obtain orientation information. That is, a sine wave is fitted to each pixel, and fitting parameters are extracted. The sine wave graph is a graph in which the horizontal axis represents the relative angle between the sample and the lattice, and the vertical axis represents the small-angle scattering signal value of a given pixel. The amplitude, average, and phase of the sine wave are obtained as fitting parameters. An image representing the amplitude value for each pixel is called an orientation tomographic image, an image showing the average value for each pixel is called a scattering intensity tomographic image, and an image showing the phase for each pixel is called an orientation angle tomographic image. Note that the fitting method is not limited to a sine wave. In this way, a three-dimensional orientation analysis image representing continuous orientation information in a plane parallel to the lattice plane of the subject H can be obtained.

[0040] Next, a method for generating a tomographic image by simultaneously reconstructing images taken in each direction will be described. The small-angle scattering tomographic image of the desired 3D CT image has a structure in which multiple scattering directions are defined for each voxel in a plane parallel to the lattice plane, and the scattering intensity for each direction is stored in each voxel. The small-angle scattering tomographic image of the desired 3D CT image is started from a certain initial value, and for this small-angle scattering tomographic image of the 3D CT image, 3D images are generated in multiple directions so that the orientation (angle) of the subject H relative to the lattice direction is the same as in the actual imaging. Small-angle scattering images of 2D projection images obtained by forward projection (predictive calculation as if actual imaging had been performed) on each 3D image are obtained, and the small-angle scattering tomographic image of the 3D CT image is corrected based on the deviation from the small-angle scattering images of the 2D projection images actually taken in multiple directions. This is sequentially and iteratively calculated to obtain the desired small-angle scattering tomographic image of the 3D CT image, using an iterative approximation method. When generating tomographic images simultaneously in this manner, small-angle scattering tomographic images for multiple directions are not created, as would be the case if a tomographic image were generated for each direction. This reduces the data volume, thereby reducing the load on the control unit 51 of the controller 5 and the capacity of the memory unit 55. When generating three-dimensional images of multiple directions in which the orientation (angle) of the subject H relative to the lattice direction is the same as that of the actual image, a shift amount previously determined using markers may be used. Because there is a density difference between the markers M and the base P, an absorption tomographic image obtained from an absorption image that is more likely to produce contrast is used. In the small-angle scattering tomographic image reconstructed by CT, the orientation degree, orientation angle, and scattering intensity at each voxel are calculated from the scattering intensity in multiple directions in a plane parallel to the lattice plane at each voxel, thereby obtaining an orientation degree tomographic image, an orientation angle tomographic image, and a scattering intensity tomographic image. In this way, a three-dimensional orientation analysis image representing continuous orientation information in a plane parallel to the lattice plane of the subject H can be obtained. 12 shows the orientation degree tomographic image, the orientation angle tomographic image, and the scattering intensity tomographic image, but the absorption tomographic image of the 3D CT image can also be obtained from the absorption image of the 2D projection image in the same way as the small-angle scattering tomographic image of the 3D CT image described above. Note that since the absorption image does not detect the anisotropy of the subject, CT reconstruction is performed from the 2D projection image in only one of the multiple directions described above.After rotating the rotating ring 134 so that the CT rotation axis 132 is parallel to the grid direction, a two-dimensional projection image of the differential phase image is acquired, and then integration in a direction perpendicular to the grid direction is performed in combination with the above-mentioned CT reconstruction to obtain a phase tomographic image of a three-dimensional CT image. Note that, since a phase image does not detect the anisotropy of the subject, imaging is performed in one of the above-mentioned multiple directions.

[0041] (Case 3) Talbot photography uses the photography jig 131 shown in FIG. 6 . Coordinate axes (x'-axis, y'-axis, z'-axis; not shown) representing the orientation of the subject H are set, and photography is performed in multiple directions such that three orthogonal directions (x'-axis, y'-axis, and z'-axis) and four diagonal directions of a cube with the x'-axis, y'-axis, and z'-axis as three sides are aligned with the CT rotation axis 132, i.e., perpendicular to the lattice direction. Unlike Case 1, two-dimensional projection images are obtained by fringe scanning processing based on moiré fringe images from multiple directions, and then CT reconstruction and image processing are performed for each direction or all at once to obtain small-angle scattering tomographic images, scattering intensity tomographic images, orientation angle tomographic images, and orientation degree tomographic images. In Case 3, continuous orientation information can be obtained for the subject H in all three directions.

[0042] More specifically, in the case of the imaging example (Case 3), the rotating ring 134 is rotated so that the direction of the CT rotation axis 132 is perpendicular to the grid direction, and then imaging is performed using the imaging jig 131 shown in Figure 6, after which processing using the fringe scanning method is performed. Coordinate axes (x'-axis, y'-axis, z'-axis) representing the orientation of the subject H are set, and the orientation of the subject H is set using the imaging jig 131 shown in Figure 6 so that three orthogonal directions (x'-axis, y'-axis, and z'-axis) and four diagonal directions of a cube with the x'-axis, y'-axis, and z'-axis as three sides are perpendicular to the direction of the CT rotation axis 132, i.e., the grid direction. CT reconstruction and image processing can then be performed by a method in which tomographic images are generated by collectively reconstructing images taken in each direction. Furthermore, if the orientation of the subject H during imaging is limited to three or more directions within the x'y' plane, y'z' plane, and z'x' plane on the coordinate axes representing the orientation of the subject (a total of nine or more directions), a method can also be used in which a tomographic image is generated by reconstructing the image for each imaging direction.

[0043] First, a method for generating tomographic images by reconstructing images for each imaging direction will be described. CT reconstruction can be performed using a method called the Feldkamp method, which allows for analytical solutions without iterative calculations. Another CT reconstruction method may be an iterative method, in which a desired 3D CT image is generated starting from a certain initial value, a forward projection (predictive calculation as if an actual image had been captured) is performed on this image to obtain a 2D projection image, the 3D CT image is corrected based on the deviation from the actual 2D projection image, and this is then sequentially and iteratively calculated to obtain the desired 3D CT image. After CT reconstruction of small-angle scattering tomographic images of a three-dimensional CT image is performed from two-dimensional projection images of small-angle scattering images using these methods, if the orientation of the subject H is within the x'y', y'z', or z'x' plane of the coordinate axes representing the orientation of the subject with respect to the direction of the CT rotation axis 132, i.e., the direction perpendicular to the lattice direction, one of the obtained small-angle scattering tomographic images in one direction is used as a reference, and the orientations of the other small-angle scattering tomographic images are rotated so that the orientations of the other small-angle scattering tomographic images are the same. Note that if markers M are attached to the base P of the imaging jig 131 in these small-angle scattering tomographic images, the marker positions on the images will approximately match, but this is due to mechanical precision. To achieve higher precision than mechanical precision, or in cases where precision decreases over time, the rotated small-angle scattering tomographic images are shifted in the required direction and position in any of x, y, z, θx, θy, and θz, and the images are adjusted so that the marker positions match. This image registration using markers may be performed on any of the 3D CT images, such as small-angle scattering tomographic images, phase tomographic images, or absorption tomographic images. Because there is a density difference between the marker M and the base P, it is preferable to use an absorption tomographic image obtained from an absorption image that is more likely to have contrast. In this case, the amounts of shift in x, y, z, θx, θy, and θz required to align the marker positions on the absorption tomographic image are determined, and the phase tomographic image and absorption tomographic image are shifted by these amounts of shift.Furthermore, for each case where the orientation of the subject H is within the x'y', y'z', or z'x' plane of the coordinate axes representing the orientation of the subject relative to the direction of the CT rotation axis 132, i.e., the direction perpendicular to the grid direction, the reference small-angle scattering tomographic image and the rotated small-angle scattering tomographic image are used to perform sinusoidal fitting for each pixel on each tomographic plane parallel to the grid plane, and fitting parameters are extracted. The sinusoidal wave graph has the relative angle between the sample and the grid on the horizontal axis and the small-angle scattering signal value of a given pixel on the vertical axis. The amplitude, average, and phase of the sinusoidal wave are obtained as fitting parameters. The image representing the amplitude value for each pixel is called the orientation tomographic image, the image showing the average value for each pixel is called the scattering intensity tomographic image, and the image showing the phase for each pixel is called the orientation angle tomographic image. The fitting method is not limited to sinusoidal fitting. As a result of the above, three sets of orientation tomographic images, orientation angle tomographic images, and scattering intensity tomographic images are obtained, with the planes parallel to the x'y', y'z', and z'x' planes on the coordinate axes representing the orientation of the subject as tomographic planes. These three sets of tomographic images may be used as is, or a set of orientation tomographic images, orientation angle tomographic images, and scattering intensity tomographic images may be generated by combining vectors in three orthogonal directions. In this way, a three-dimensional orientation analysis image can be obtained that represents continuous orientation information of the subject H in all three directions.

[0044] Next, a method for generating a tomographic image by simultaneously reconstructing images taken in each direction will be described. The small-angle scattering tomographic image of the desired 3D CT image has a structure in which multiple three-dimensional scattering directions are defined for each voxel, and the scattering intensity for each direction is stored in each voxel. The small-angle scattering tomographic image of the desired 3D CT image is started from a certain initial value, and 3D images of this small-angle scattering tomographic image of the 3D CT image are generated in multiple directions so that the orientation (angle) of the small-angle scattering tomographic image with respect to the lattice direction of the subject H is the same as that of the actual image. Small-angle scattering images of 2D projection images obtained by forward projection (predictive calculation as if actual image capture had been performed) on each 3D image are obtained, and the small-angle scattering tomographic image of the 3D CT image is corrected based on the deviation from the small-angle scattering images of the 2D projection images actually taken in multiple directions. This is sequentially and iteratively calculated to obtain the small-angle scattering tomographic image of the desired 3D CT image, using an iterative approximation method. When generating tomographic images simultaneously in this manner, small-angle scattering tomographic images for multiple directions are not created, as would be the case if a tomographic image were generated for each direction. This reduces the data volume, thereby reducing the load on the control unit 51 of the controller 5 and the capacity of the memory unit 55. When generating three-dimensional images of multiple directions in which the orientation (angle) of the subject H relative to the lattice direction is the same as that of the actual image, a shift amount previously determined using markers may be used. Because there is a density difference between the markers M and the base P, an absorption tomographic image obtained from an absorption image that is more likely to produce contrast is used. In the small-angle scattering tomographic image reconstructed by CT, the orientation degree, orientation angle, and scattering intensity at each voxel are calculated from the multiple three-dimensional scattering intensities at each voxel, and an orientation degree tomographic image, an orientation angle tomographic image, and a scattering intensity tomographic image are obtained. In this way, a three-dimensional orientation analysis image representing continuous orientation information in all three-dimensional directions of the subject H can be obtained. Furthermore, an absorption tomographic image of a three-dimensional CT image can also be obtained from the absorption image of a two-dimensional projection image in a manner similar to that used to obtain a small-angle scattering tomographic image of a three-dimensional CT image. Since the absorption image does not detect the anisotropy of the subject, CT reconstruction is performed from the two-dimensional projection image in only one of the above multiple directions.After rotating the rotating ring 134 so that the CT rotation axis 132 is parallel to the grid direction, a two-dimensional projection image of the differential phase image is acquired, and then integration in a direction perpendicular to the grid direction is performed in combination with the above-mentioned CT reconstruction to obtain a phase tomographic image of a three-dimensional CT image. Note that, since a phase image does not detect the anisotropy of the subject, imaging is performed in one of the above-mentioned multiple directions.

[0045] (Operation of X-ray Talbot imaging device 100) Next, a description will be given of the operation of the X-ray Talbot imaging device 100. Fig. 13 is a flowchart showing imaging processing (Talbot CT imaging) executed by cooperation between the control unit 51 of the X-ray Talbot imaging device 100 and a program stored in the storage unit 55.

[0046] First, the control unit 51 controls various components of the X-ray Talbot imaging device 100 to perform Talbot imaging (step S1). Specifically, as described above, the motor 137 is first driven to rotate the rotating ring 134, so that the CT rotation axis 132 is oriented either parallel or perpendicular to the lattice direction. When generating absorption tomographic images or small-angle scattering tomographic images of a sample with no orientation, either orientation is acceptable. However, when generating orientation degree tomographic images, orientation angle tomographic images, scattering intensity tomographic images, and small-angle scattering tomographic images for generating these images, the CT rotation axis 132 should be oriented perpendicular to the lattice direction. When generating phase tomographic images, the CT rotation axis 132 should be oriented parallel to the lattice direction. The control unit 51 then rotates the CT rotation axis 132 by a predetermined angle to generate multiple absorption, differential phase, and small-angle scattering images as two-dimensional projection images using a fringe scanning method based on multiple moiré fringe images covering 180° or 360°.

[0047] Next, the control unit 51 determines whether there has been a change in the angle of the subject H relative to the CT rotation axis 132 (step S2). If there has been a change (step S2; YES), the control unit 51 displays a message on the display unit 53 prompting the user to manually rotate the rotating part R of the imaging jig 131 to a predetermined angle, or goes into a standby state and advances the imaging process to step S1 if it receives information indicating that the rotating part R of the imaging jig 131 has been manually rotated to the predetermined angle. If there has been no change (step S2; NO), the control unit 51 advances the imaging process to step S3. Note that by using a pre-programmed robot hand to rotate the rotating part R of the imaging jig 131 to a predetermined angle, multiple Talbot CT imaging sessions can be performed automatically without human intervention.

[0048] Next, the control unit 51 generates three-dimensional CT images (absorption tomographic images, differential phase tomographic images, and small-angle scattering tomographic images) using the absorption images, differential phase images, and small-angle scattering images of the multiple two-dimensional projection images generated in step S1 (step S3). The control unit 51 then terminates the imaging process. Furthermore, if imaging of the subject has been performed in multiple directions relative to the CT rotation axis 132, the control unit 51 generates orientation analysis tomographic images using the multiple small-angle scattering images captured in multiple directions relative to the CT rotation axis 132. The orientation analysis tomographic images include an orientation degree tomographic image, a scattering intensity tomographic image, and an orientation angle tomographic image.

[0049] As described above, in the imaging process (Talbot CT imaging), image alignment is not performed to fine-tune the orientation and position of the subject on the image for the Moiré fringe images, two-dimensional projection images (absorption images, small-angle scattering images, differential phase images), and three-dimensional CT images (absorption tomography images, small-angle scattering tomography images, and phase tomography images). In other words, the image is rotated in the reverse direction by the angle of the subject rotation, but the orientation and position of the subject on the image is not fine-tuned. This is because the CT rotation axis support / drive unit and the rotating ring 134 accurately switch the CT rotation axis 132 between parallel and perpendicular to the lattice direction, with the optical axis 17 as the center. The orientation of the subject H relative to the CT rotation axis 132 is accurately set to the desired orientation by the positioning unit L of the imaging jig described above. When performing image alignment to correct the orientation of the subject H on the image using the marker M, image alignment can be completed in a short time because the original orientation of the subject is highly accurate. This makes it possible to further improve the accuracy of the orientation of the subject on the image, and to improve the accuracy of the orientation of the subject on the image even if the accuracy decreases due to changes over time.

[0050] (Effects) As described above, the X-ray Talbot imaging device 100 is an X-ray Talbot imaging device in which the radiation irradiation axis is substantially vertical and the grid plane is substantially horizontal, and includes a CT rotation axis 132 to which an imaging jig 131 for fixing the subject in a predetermined orientation is attached. The CT rotation axis 132 rotates and is fixed in a predetermined orientation within a substantially horizontal plane. Furthermore, a CT rotation axis support and drive unit that supports and rotates the CT rotation axis 132 is mounted on a rotating ring 134 that rotates around the optical axis 17. The CT rotation axis support and drive unit rotates the CT rotation axis 132 within a substantially horizontal plane to switch the direction of the CT rotation axis 132 between a direction parallel to the grid direction and a direction perpendicular to the grid direction, depending on the type of tomographic image to be generated. This makes it possible to prevent a decrease in the accuracy of the orientation and position of the subject, and to increase these accuracies, as described above. Specifically, even if rattle occurs in the subject table 13 (particularly, the CT rotation axis support / drive unit 133, the rotating ring 134, and the rotating base 135), gravity acts perpendicularly on the CT rotation axis 132 relative to the CT rotation axis support / drive unit, or on the rotating ring 134 relative to the rotating base 135, so the position is stable. This also makes artifacts less likely to occur. Furthermore, if the subject includes a highly anisotropic structure, rotating the rotating ring 134 to orthogonalize the CT rotation axis 132 to the lattice direction makes it less likely that the scattering signal detected by the lattice will change in the highly anisotropic structure, making it less likely that artifacts will occur in the small-angle scattering tomographic image. This also makes it less likely that artifacts will occur in the three-dimensional orientation analysis image generated from small-angle scattering tomographic images in multiple directions.

[0051] The imaging jig 131a can fix the subject H by changing the orientation of the subject H around the first axis, which is orthogonal to the CT rotation axis 132, so that the axial direction is fixed in one direction and the orientation of the subject is defined. Therefore, the accuracy of the orientation of the subject is high.

[0052] The imaging jig 131b is not only capable of changing and fixing the orientation of the subject H around a first axis, which is a direction perpendicular to the CT rotation axis 132, but also capable of changing and fixing the orientation of the subject H around a second axis, which is a direction perpendicular to the first axis. In other words, the axial direction is fixed in two directions. Therefore, the orientation of the subject is defined, resulting in high accuracy of the orientation of the subject.

[0053] The photographing jig 131 has a rotating part R that fixes the subject H and is rotatable around a predetermined axis, and a holding part K that holds the rotating part R. The rotating part R and the holding part K are provided with positioning parts L, and the subject H is fixed in a predetermined orientation by the positioning parts L. Therefore, the orientation of the subject is highly accurate.

[0054] A convex portion L1 is provided on one of the positioning portions L of the rotating portion R and the holding portion K of the photographing jig 131, and a concave portion L2 is provided on the other of the positioning portions L of the rotating portion R and the holding portion K, so that by aligning these, the orientation of the subject H is changed to a predetermined orientation and fixed. Therefore, the orientation of the subject is highly accurate.

[0055] The rotating part R of the imaging jig 131 has a base P on which the subject H is placed and fixed, and markers M, each having a triangular, rectangular, circular, or spherical shape, are attached to the base P. The density of the markers M is greater than the density of the base P. This density difference causes the markers M to appear in the generated 2D projection image and 3D CT image. Because the imaging jig 131 has high mechanical precision, the markers M can be used to align the 3D CT image. However, since the imaging jig 131 already has high mechanical precision, this can be done in a short time. Therefore, this is particularly effective when the orientation of the subject H is highly accurate, but the precision is higher than the mechanical precision, or when the mechanical precision decreases over time. Furthermore, because the markers M are attached to the base P of the imaging jig 131, unlike when markers M are attached to the subject H, there is no need to prepare and attach markers M for each imaging session.

[0056] The imaging jig 131 is made of lightweight materials such as filler-free resin or light metal, which makes it difficult for the CT rotation axis 132 to bend, resulting in high accuracy in the orientation and position of the subject. Furthermore, these materials are less likely to absorb or scatter X-rays, so they do not impede the visibility of the subject H and are less likely to produce artifacts.

[0057] The X-ray Talbot imaging method is an X-ray Talbot imaging method using the X-ray Talbot imaging device 100. The X-ray Talbot imaging method includes an imaging step (step S1) of acquiring multiple images in two directions by changing the orientation of the subject H in two directions (case 1) and rotating the CT rotation axis 132, and a generation step (step S2) of generating tomographic images for each direction. As a result, as described above, since the orientation and position of the subject H are highly accurate in step S1, artifacts are less likely to occur in step S3, and it is possible to obtain 3D CT images in two directions with high accuracy and in the desired orientation.

[0058] The X-ray Talbot imaging method is an X-ray Talbot imaging method using the X-ray Talbot imaging device 100. The X-ray Talbot imaging method includes an imaging step (step S1) in which the orientation of the subject H is changed in two directions (case 1) and the CT rotation axis 132 is rotated to acquire images in two directions, and a generation step (step S2) in which a tomographic image is generated for each direction. As a result, as described above, the accuracy of the orientation and position of the subject H is high in step S1. Furthermore, when the image alignment step of correcting the orientation and position of the subject H on the image using the marker M is performed in step S3, the accuracy of the original orientation and position of the subject H is high, so that the process can be completed in a short time. The accuracy of the orientation and position of the subject H on the image can be further improved, or even if the accuracy decreases due to changes over time, the accuracy of the orientation and position of the subject H on the image can be improved, making it less likely that artifacts will occur.

[0059] The X-ray Talbot imaging method is an X-ray Talbot imaging method using the X-ray Talbot imaging device 100. The X-ray Talbot imaging method includes an imaging step (step S1) in which the orientation of the subject is changed in multiple directions (cases 2 and 3) and the CT rotation axis 132 is rotated to acquire images taken in multiple directions. The X-ray Talbot imaging method further includes a generation step (step S3) in which a tomographic image is generated for each of the directions. As a result, as described above, the orientation and position of the subject H are highly accurate. Therefore, artifacts are less likely to occur when generating a three-dimensional orientation analysis image. Furthermore, the generation step (step S3) in which tomographic images are generated is a generation step in which tomographic images are generated collectively for the multiple directions. As a result, small-angle scattering tomographic images for multiple directions are not created when tomographic images are generated for each of the directions, thereby reducing the data volume and the load on the control unit 51 of the controller 5 and the capacity of the storage unit 55.

[0060] The X-ray Talbot imaging method is an X-ray Talbot imaging method using the X-ray Talbot imaging device 100. The X-ray Talbot imaging method includes an imaging step (step S1) in which the orientation of the subject is changed in multiple directions (cases 2 and 3) and the CT rotation axis 132 is rotated to acquire images taken in multiple directions. The X-ray Talbot imaging method further includes a generation step (step S3) in which a tomographic image is generated for each of the directions. This results in high accuracy in the orientation and position of the subject H, as described above. When the image alignment step S3, in which the orientation and position of the subject H on the image are corrected using the marker M, is performed in a short time because the original orientation and position of the subject H are highly accurate. The accuracy of the orientation and position of the subject H on the image can be further improved, or even if the accuracy decreases due to changes over time, the accuracy of the orientation and position of the subject H on the image can be improved, thereby making it less likely that artifacts will occur. Furthermore, the generation step (step S3) in which tomographic images are generated collectively for the multiple directions is used. As a result, when generating a tomographic image for each of the above directions, small-angle scattering tomographic images for multiple directions are not created, so the data volume can be reduced, and the load on the control unit 51 of the controller 5 and the capacity of the memory unit 55 can be reduced.

[0061] The embodiment and its modifications of the present invention have been described above, but the description of the above-described embodiment is a preferred example of the present invention, and the present invention is not limited to this.

[0062] For example, in the above embodiment, an inspection device using a Talbot-Lau interferometer in which the second grating 15 is moved relative to the multi-slit 12 and the first grating 14 during imaging using the fringe scanning method has been described as an example. However, the present invention may also be applied to an inspection device using a Talbot-Lau interferometer in which either the multi-slit 12, the first grating 14, or the second grating 15, or two of these gratings, are moved during imaging using the fringe scanning method. The present invention may also be applied to an inspection device using a Talbot interferometer in which either the first grating 14 or the second grating 15 is moved relative to the other grating. The present invention may also be applied to an inspection device using a Law interferometer in which either the multi-slit 12 or the first grating 14 is moved relative to the other grating. The present invention may also be applied to a Talbot-Lau interferometer, Talbot interferometer, or Law interferometer that uses the Fourier transform method, which does not require the fringe scanning method.

[0063] Furthermore, for example, in the above description, examples have been disclosed in which a hard disk or a semiconductor nonvolatile memory is used as a computer-readable medium for the program according to the present invention, but the present invention is not limited to this example. Other computer-readable media include portable recording media such as CD-ROMs. Furthermore, carrier waves are also applicable as a medium for providing the program data according to the present invention via a communication line.

[0064] In addition, the detailed configuration and detailed operation of each device constituting the X-ray Talbot imaging device can be modified as appropriate without departing from the spirit of the invention.

[0065] The present disclosure can be used in an X-ray Talbot imaging device and an X-ray Talbot imaging method.

[0066] 100 X-ray Talbot imaging device 1 Main body 11 Radiation source 12 Multi-slit 13 Subject table 131 Imaging jig 132 CT rotation axis 133 CT rotation axis support / drive unit 134 Rotating ring (rotation axis rotating unit) 135 Rotating base 136 Gear 137 Motor 14 First grating 15 Second grating 15a Moving mechanism 16 Radiation detector 5 Controller 51 Control unit (imaging unit, generating unit) 52 Operation unit 53 Display unit 54 Communication unit 55 Storage unit

Claims

1. An X-ray Talbot imaging device in which the radiation irradiation axis is approximately vertical and the grid surface is approximately horizontal, and which is equipped with a CT rotation axis to which an imaging jig that fixes a subject in a specified orientation is attached, and the CT rotation axis rotates in a specified orientation within an approximately horizontal plane and is fixed.

2. An X-ray Talbot imaging device as described in claim 1, further comprising a rotation axis rotating unit capable of changing the angle between the CT rotation axis and the grid direction of the grid surface within a plane approximately parallel to the grid surface, to a direction perpendicular to or parallel to the grid.

3. An X-ray Talbot imaging device according to claim 1, wherein the imaging jig is capable of fixing the subject by changing the orientation of the subject around a first axis that is perpendicular to the CT rotation axis.

4. An X-ray Talbot imaging device according to claim 3, wherein the imaging jig is capable of fixing the subject by changing the orientation of the subject around a second axis that is perpendicular to the first axis.

5. The X-ray Talbot imaging device according to claim 1, wherein the imaging tool has a rotating section that fixes the subject and is rotatable around a predetermined axis, and a holding section that holds the rotating section, and the rotating section and the holding section are provided with positioning sections, and the subject is fixed in a predetermined orientation by aligning the positioning sections.

6. The X-ray Talbot imaging device according to claim 5, wherein a convex portion is provided on the positioning portion of one of the rotating portion and the holding portion, and a concave portion is provided on the positioning portion of the other of the rotating portion and the holding portion.

7. The X-ray Talbot imaging device according to claim 5, wherein the rotating section has a base on which the subject is placed and fixed, markers are attached to the base, and the density of the markers is greater than the density of the base.

8. The X-ray Talbot imaging device according to claim 7, wherein the shape of the marker is triangular, rectangular, circular or spherical.

9. The X-ray Talbot imaging device according to claim 1, wherein said imaging jig is made of resin.

10. The X-ray Talbot imaging device according to claim 1, wherein said imaging jig is made of light metal.

11. An X-ray Talbot imaging method using the X-ray Talbot imaging device according to claim 1, comprising: an imaging step of changing the orientation of the subject in two directions and rotating the CT rotation axis to obtain a plurality of images in each of the two directions; and a generation step of generating a tomographic image for each of the directions.

12. An X-ray Talbot imaging method using the X-ray Talbot imaging device described in claim 7, comprising: an imaging step of changing the orientation of the subject in two directions and rotating the CT rotation axis to obtain a plurality of images in each of the two directions; and a generation step of generating a tomographic image for each of the directions, wherein the generation step includes an image alignment step of correcting the orientation and position of the subject on the image using the markers.

13. An X-ray Talbot imaging method using the X-ray Talbot imaging device described in claim 1, comprising: an imaging step of acquiring a plurality of images in a plurality of directions by changing the orientation of the subject in a plurality of directions and rotating the CT rotation axis; and a generating step of generating a tomographic image for each of the directions.

14. An X-ray Talbot imaging method using the X-ray Talbot imaging device described in claim 1, comprising: an imaging step of acquiring a plurality of images in a plurality of directions by changing the orientation of the subject in a plurality of directions and rotating the CT rotation axis; and a generating step of generating a tomographic image by combining the plurality of directions at once.

15. An X-ray Talbot imaging method using the X-ray Talbot imaging device described in claim 7, comprising: an imaging step of acquiring a plurality of images in a plurality of directions by changing the orientation of the subject in a plurality of directions and rotating the CT rotation axis; and a generation step of generating a tomographic image for each of the directions, wherein the generation step includes an image alignment step of correcting the orientation and position of the subject on a three-dimensional CT image using the markers.

16. An X-ray Talbot imaging method using the X-ray Talbot imaging device described in claim 7, comprising: an imaging step of acquiring a plurality of images in a plurality of directions by changing the orientation of the subject in a plurality of directions and rotating the CT rotation axis; and a generation step of generating a tomographic image by combining the plurality of directions at once, wherein the generation step includes an image alignment step of correcting the orientation and position of the subject on a three-dimensional CT image using the markers.

Images