JP7707869B2 - Cooling vessel, X-ray device and X-ray image processing method - Google Patents

Description

The present disclosure relates to a cooling vessel, an X-ray device, an X-ray image processing device, and an X-ray image processing method.

Takeya Satoshi, Development of Low-Temperature Phase Contrast X-Ray CT Technology, AIST TODAY, November 1, 2006, Vol. 6, No. 11, pp. 22-23 (Non-Patent Document 1), discloses a low-temperature sample container used in low-temperature phase contrast X-ray CT (Computed Tomography) measurement. This low-temperature sample container includes a metallic cooling container body and a metallic liquid container.

The cooling container body contains a refrigerant and a liquid container. The liquid container contains a liquid and a sample immersed in the liquid. The liquid container is provided with a heater that controls the temperature of the liquid. The heater is connected to a temperature regulator outside the low-temperature sample container through electrical wiring. A thermocouple is connected to the temperature regulator. The thermocouple is in contact with the liquid. The sample is suspended from a rotating shaft. The cooling container body and the liquid container are provided with X-ray transmitting windows. While the rotating shaft is rotated to rotate the sample, X-rays that pass through the X-ray transmitting windows of the cooling container body and the liquid container are irradiated onto the sample. The X-rays that have passed through the sample are detected.

Satoshi Takeya, Development of Low-Temperature Phase Contrast X-ray CT Technology, AIST TODAY, November 1, 2006, Vol. 6, No. 11, pp. 22-23

However, in the cryogenic sample container of Non-Patent Document 1, the sample is suspended from a rotating shaft. Since it is difficult to suspend a large sample from a rotating shaft, it is difficult to obtain an X-ray image of a large sample. In addition, the cryogenic sample container of Non-Patent Document 1 is provided with a heater and a thermocouple to control the temperature of the liquid. This increases the size of the cryogenic sample container and requires electrical wiring. It is difficult to incorporate the cryogenic sample container into an X-ray imaging device without modifying the X-ray imaging device.

The present disclosure has been made in consideration of the above problems, and the object of a first aspect of the present disclosure is to provide a cooling container that can more easily obtain X-ray images of a sample having a larger size at a low temperature lower than room temperature and can be easily incorporated into an X-ray imaging device. The object of a second aspect of the present disclosure is to provide an X-ray device that can more easily obtain X-ray images of a sample having a larger size at a low temperature lower than room temperature and can utilize an X-ray imaging device. The object of a third aspect of the present disclosure is to provide an X-ray image processing device and an X-ray image processing method that are suitable for the cooling container of the present disclosure.

The cooling container of the present disclosure includes a hollow container and an insulated container that houses the hollow container. The hollow container includes a tube and a bottom plate. The tube includes a first end and a second end opposite the first end. The bottom plate closes the first end of the tube. A sample and a cooling liquid that cools the sample are housed in a storage space defined by the tube and the bottom plate. The sample is immersed in the cooling liquid and is supported by at least one of the tube or the bottom plate. The tube and the insulated container are formed of a low-density material. The second end of the tube is provided with a first opening that communicates with the storage space. The insulated container is provided with a through hole that communicates with the first opening.

The X-ray device of the present disclosure includes an X-ray imaging device and the cooling container of the present disclosure. The X-ray imaging device includes an X-ray source that emits X-rays, a stage that supports the cooling container, and an X-ray detector that detects the X-rays that have passed through the cooling container.

The X-ray image processing device of the present disclosure includes an X-ray image receiving unit that receives an X-ray image of a sample acquired by imaging the sample for a first time at a low temperature lower than room temperature, and an image processing unit. The image processing unit inputs the X-ray image of the sample to a neural network to generate a corrected X-ray image of the sample that is clearer than the X-ray image of the sample. The neural network is generated by learning using a learning dataset. The learning dataset includes a first reference X-ray image acquired by imaging the reference sample for a second time without cooling the reference sample, and a second reference X-ray image acquired by imaging the reference sample for a third time without cooling the reference sample. The second time is 0.5 to 2.0 times the first time. The third time is longer than the first time and longer than the second time.

The X-ray image processing method of the present disclosure includes the steps of acquiring an X-ray image of a sample by imaging the sample for a first time at a low temperature lower than room temperature using the X-ray device of the present disclosure, and inputting the X-ray image of the sample into a neural network to generate a corrected X-ray image of the sample that is clearer than the X-ray image of the sample. The neural network is generated by learning using a training data set. The training data set includes a first reference X-ray image acquired by imaging the reference sample for a second time using the X-ray device of the present disclosure without cooling the reference sample, and a second reference X-ray image acquired by imaging the reference sample for a third time using the X-ray device of the present disclosure without cooling the reference sample. The second time is 0.5 to 2.0 times the first time. The third time is longer than the first time and longer than the second time.

The cooling container of the present disclosure makes it easier to obtain X-ray images of samples having larger sizes at temperatures lower than room temperature. The cooling container of the present disclosure can be easily incorporated into an X-ray imaging device.

The X-ray device disclosed herein makes it easier to obtain X-ray images of samples having larger sizes at temperatures lower than room temperature and allows the use of an X-ray imaging device.

The X-ray image processing device of the present disclosure is an X-ray image processing device suitable for the cooling container of the present disclosure, which shortens the X-ray imaging time at low temperatures lower than room temperature. The X-ray image processing method of the present disclosure is an X-ray image processing method suitable for the cooling container of the present disclosure, which shortens the X-ray imaging time at low temperatures lower than room temperature.

FIG. 1 is a schematic diagram of an X-ray device according to an embodiment of the present invention. FIG. 2 is a schematic enlarged cross-sectional view of a cooling vessel according to an embodiment that contains a sample, and shows a step in a method of acquiring an X-ray image of the sample. FIG. 3 is a schematic enlarged cross-sectional view of the cooling container according to the embodiment taken along the cross-sectional line III-III shown in FIG. FIG. 4 is a flowchart showing a method for acquiring an X-ray image of a sample cooled to a temperature lower than room temperature according to an embodiment. FIG. 5 is a diagram showing an example of an X-ray image of a sample. FIG. 6 is a schematic enlarged cross-sectional view showing a step of a method for acquiring a first reference X-ray image and a second reference X-ray image of a reference sample. FIG. 7 is a diagram showing an example of the first reference X-ray image. FIG. 8 is a diagram showing an example of the second reference X-ray image. FIG. 9 is a schematic diagram of an image processing system according to an embodiment. FIG. 10 is a schematic diagram illustrating an example of a hardware configuration of a server device according to an embodiment. FIG. 11 is a schematic diagram illustrating an example of a hardware configuration of a terminal device according to an embodiment. FIG. 12 is a functional block diagram illustrating a functional configuration of the server device according to the embodiment. FIG. 13 is a diagram illustrating a data structure of a learning data set according to the embodiment. FIG. 14 is a block diagram for explaining details of the process (learning function) of the learning unit according to the embodiment. FIG. 15 is a diagram showing an example of a provisionally corrected X-ray image. FIG. 16 is a flowchart of the learning process according to the embodiment. FIG. 17 is a functional block diagram illustrating a functional configuration of a terminal device according to an embodiment. FIG. 18 is a diagram showing an example of a corrected X-ray image of a sample. FIG. 19 is a flowchart of the X-ray image processing method according to the embodiment.

[Description of the embodiments of the present disclosure]
First, the embodiments of the present disclosure will be listed and described.

(1) The cooling container 5 of the present disclosure includes a hollow container 20 and an insulated container 35 that accommodates the hollow container 20. The hollow container 20 includes a tube 22 and a bottom plate 21. The tube 22 includes a first end 22a and a second end 22b opposite to the first end 22a. The bottom plate 21 closes the first end 22a of the tube 22. A sample 30 and a cooling liquid 33 that cools the sample 30 are accommodated in an accommodation space 22e defined by the tube 22 and the bottom plate 21. The sample 30 is immersed in the cooling liquid 33 and is supported by at least one of the tube 22 or the bottom plate 21. The tube 22 and the insulated container 35 are formed of a low-density material. A first opening 23 that communicates with the accommodation space 22e is provided at the second end 22b of the tube 22. The insulated container 35 has a through hole 39 that communicates with the first opening 23.

The tube 22 and the insulating container 35 are made of a low-density material, so that the tube 22 and the insulating container 35 can transmit X-rays 3a. The sample 30 is immersed in the cooling liquid 33 and is supported by at least one of the tube 22 or the bottom plate 21. Therefore, it is easier to obtain an X-ray image 60 of the sample 30 having a larger size at a low temperature lower than room temperature.

In the cooling container 5 of the present disclosure, the first opening 23 is provided at the second end 22b of the tube 22, and the through hole 39 is provided in the insulating container 35. During the X-ray fluoroscopic image of the sample 30, the temperature of the sample 30 and the cooling liquid 33 gradually rises, and even if a part of the cooling liquid 33 turns into gas, this gas is discharged to the outside of the cooling container 5 through the first opening 23 and the through hole 39. The cooling container 5 can be prevented from being damaged by this gas. Therefore, the temperature adjustment function can be omitted from the cooling container 5. The cooling container 5 of the present disclosure can be easily incorporated into the X-ray imaging device 2 without modifying the X-ray imaging device 2. The cost of the cooling container 5 can be reduced. In this specification, the cooling container 5 does not have a temperature adjustment function means that the cooling container 5 is not provided with an active temperature adjustment element (e.g., a Peltier element or a heater) that actively adjusts the temperature of the cooling liquid 33 and the temperature of the sample 30.

The cooling container 5 of the present disclosure includes a thermally insulated container 35. Therefore, even if the cooling container 5 does not have a temperature control function, the temperature rise of the sample 30 and the cooling liquid 33 during the capture of the X-ray image 60 of the sample 30 can be made gradual. The time for capturing the X-ray image 60 of the sample 30 at a low temperature lower than room temperature can be extended. The clarity of the X-ray image 60 of the sample 30 at a low temperature lower than room temperature can be improved.

(2) In the cooling vessel 5 of (1) above, the tube 22 has a cylindrical shape. The outer surface of the portion of the insulating vessel 35 that covers the tube 22 has a cylindrical outer surface shape.

The cooling vessel 5 can therefore be positioned closer to the X-ray source 3 of the X-ray imaging device 2. The magnification of the X-ray image 60 of the sample 30 can be increased.

(3) In the cooling vessel 5 of (1) or (2) above, the tube 22 is in contact with both the cooling liquid 33 and the insulated vessel 35.

Therefore, the cooling container 5 has a simple configuration. The cooling container 5 can be easily incorporated into the X-ray imaging device 2 without modifying the X-ray imaging device 2. The cost of the cooling container 5 can be reduced.

(4) In any of the cooling containers 5 described above in (1) to (3), the insulating container 35 covers the entire outer surface 22c of the tube 22.

As a result, the temperature rise of the sample 30 and the coolant 33 while the X-ray image 60 of the sample 30 is being taken can be made gentler. The time for taking the X-ray image 60 of the sample 30 at a low temperature lower than room temperature can be extended. The clarity of the X-ray image 60 of the sample 30 at a low temperature lower than room temperature can be improved.

(5) In any of the cooling containers 5 described above in (1) to (4), the insulating container 35 covers the entire outer surface of the hollow container 20.

As a result, the temperature rise of the sample 30 and the coolant 33 while the X-ray image 60 of the sample 30 is being taken can be made gentler. The time for taking the X-ray image 60 of the sample 30 at a low temperature lower than room temperature can be extended. The clarity of the X-ray image 60 of the sample 30 at a low temperature lower than room temperature can be improved.

(6) In any of the cooling containers 5 described above in (1) to (5), the insulated container 35 includes a first insulating member 36 in contact with the bottom plate 21 and a second insulating member 38 in contact with the tube 22 and the first insulating member 36.

Therefore, the hollow container 20 can be easily housed in the insulating container 35. The cooling container 5 can be easily incorporated into the X-ray imaging device 2 without modifying the X-ray imaging device 2.

(7) In the cooling container 5 of (6) above, a recess 37 is formed in the first insulating member 36. The second insulating member 38 is fitted into the recess 37.

Therefore, the hollow container 20 can be easily housed in the insulating container 35. The cooling container 5 can be easily incorporated into the X-ray imaging device 2 without modifying the X-ray imaging device 2.

(8) In any of the cooling containers 5 described above in (1) to (7), the tube 22 is made of a fluorine-based resin or a polyethylene resin.

Therefore, the tube 22 is formed of a low-density material that does not become embrittled when cooled with the cooling liquid 33. X-ray images 60 of samples 30 having larger sizes can be more easily obtained at temperatures lower than room temperature.

(9) In the cooling vessel 5 of (8) above, the tube 22 is made of polytetrafluoroethylene or perfluoroalkoxyalkane.

Therefore, the tube 22 is formed of a low-density material that does not become embrittled when cooled with the cooling liquid 33. X-ray images 60 of samples 30 having larger sizes can be more easily obtained at temperatures lower than room temperature.

(10) In any of the cooling containers 5 described above in (1) to (9), the insulating container 35 is made of polystyrene foam.

The insulated container 35 is made of inexpensive materials. This reduces the cost of the cooling container 5.

(11) In any of the cooling containers 5 described above in (1) to (10), the cooling liquid 33 is a mixture of dry ice and alcohol.

The sample 30 can be kept at a temperature lower than room temperature using an inexpensive cooling liquid 33. X-ray images 60 of the sample 30 having a larger size can be obtained more easily at a temperature lower than room temperature.

(12) In any of the cooling vessels 5 described above in (1) to (11), the hollow vessel 20 further includes a fixing member 25. The sample 30 is supported relative to the tube 22 via the fixing member 25. The fixing member 25 is provided with a second opening 27 that extends along the longitudinal direction of the tube 22 and passes through the fixing member 25.

Therefore, the fixing member 25 can more stably support a sample 30 having a larger size. It is easier to obtain an X-ray image 60 of a sample 30 having a larger size at a low temperature that is lower than room temperature.

(13) The X-ray device 1 of the present disclosure includes an X-ray imaging device 2 and any one of the cooling containers 5 described above in (1) to (11). The X-ray imaging device 2 includes an X-ray source 3 that emits X-rays, a stage 4 that supports the cooling container 5, and an X-ray detector 7 that detects the X-rays that have passed through the cooling container 5.

In the X-ray device 1 of the present disclosure, the sample 30 is cooled using the cooling container 5 of the present disclosure. Therefore, it is easier to obtain an X-ray image 60 of a sample 30 having a larger size at a low temperature lower than room temperature. The X-ray imaging device 2 can be used without modifying the X-ray imaging device 2.

(14) In the X-ray device 1 described above in (13), the stage 4 can rotate the cooling container 5 around the central axis 5c of the cooling container 5 that extends along the longitudinal direction of the tube 22, and can move along the longitudinal direction of the tube 22.

As a result, the X-ray device 1 can more easily obtain X-ray CT images of samples 30 having larger sizes.

(15) The X-ray image processing device (terminal device 200) of the present disclosure includes an X-ray image receiving unit 220 and an image processing unit 221. The X-ray image receiving unit 220 receives an X-ray image 60 of the sample 30 acquired by imaging the sample 30 for a first time at a low temperature lower than room temperature. The image processing unit 221 inputs the X-ray image 60 of the sample 30 to a neural network (trained model 113) to generate a corrected X-ray image 66 of the sample 30 that is clearer than the X-ray image 60 of the sample 30. The neural network is generated by learning using a learning dataset 115. The learning dataset 115 includes a first reference X-ray image 61 acquired by imaging the reference sample 31 for a second time without cooling the reference sample 31, and a second reference X-ray image 62 acquired by imaging the reference sample 31 for a third time without cooling the reference sample 31. The second time is 0.5 times or more and 2.0 times or less than the first time. The third time is longer than the first time and longer than the second time.

The temperatures of the sample 30 and the cooling liquid 33 gradually increase while the X-ray image 60 of the sample 30 is being captured. In order to capture the X-ray image 60 of the sample 30 at a low temperature lower than room temperature, it is necessary to shorten the first time. If the first time is shortened, the X-ray image 60 of the sample 30 cooled to a low temperature lower than room temperature becomes unclear. The X-ray image processing device of the present disclosure can make the X-ray image 60 of the sample 30 cooled to a low temperature lower than room temperature clearer. The X-ray image processing device of the present disclosure is suitable for the cooling container 5 of the present disclosure, which shortens the X-ray capture time (first time) at a low temperature lower than room temperature.

(16) The X-ray image processing method of the present disclosure includes the steps of acquiring an X-ray image 60 of the sample 30 by imaging the sample 30 for a first time at a low temperature lower than room temperature using the X-ray device 1 of (13) or (14) above (e.g., steps S1 to S6), and inputting the X-ray image 60 of the sample 30 into a neural network (trained model 113) to generate a corrected X-ray image 66 of the sample 30 that is clearer than the X-ray image 60 of the sample 30 (step S22). The neural network is generated by learning using a training data set 115. The training data set 115 includes a first reference X-ray image 61 acquired by imaging the reference sample 31 for a second time using the X-ray device 1 without cooling the reference sample 31, and a second reference X-ray image 62 acquired by imaging the reference sample 31 for a third time using the X-ray device 1 without cooling the reference sample 31. The second time is 0.5 to 2.0 times the first time. The third time is longer than the first time and longer than the second time.

The temperatures of the sample 30 and the cooling liquid 33 gradually increase while the X-ray image 60 of the sample 30 is being captured. In order to capture the X-ray image 60 of the sample 30 at a low temperature lower than room temperature, it is necessary to shorten the first time. If the first time is shortened, the X-ray image 60 of the sample 30 cooled to a low temperature lower than room temperature becomes unclear. The X-ray image processing method of the present disclosure can make the X-ray image 60 of the sample 30 cooled to a low temperature lower than room temperature clearer. The X-ray image processing method of the present disclosure is suitable for the cooling container 5 of the present disclosure, in which the X-ray capture time (first time) at a low temperature lower than room temperature is shortened.

[Details of the embodiment of the present disclosure]
Next, the details of the embodiments of the present disclosure will be described with reference to the drawings. Note that the same or corresponding parts in the following drawings are given the same reference numerals, and the description thereof will not be repeated. At least some of the configurations of the embodiments described below may be arbitrarily combined.

<X-ray device 1>
1 and 2, the X-ray device 1 includes an X-ray imaging device 2 and a cooling container 5. The X-ray device 1 is, for example, an X-ray CT (Computed Tomography) device.

The X-ray imaging device 2 includes an X-ray source 3, a stage 4, and an X-ray detector 7. The X-ray imaging device 2 may further include a stage driving device 6 and a computer 10.

The X-ray source 3 emits X-rays 3a toward the sample 30. The X-ray source 3 is, for example, an X-ray tube.

The stage 4 supports the cooling container 5. The cooling container 5 is placed on the sample placement surface 4a. The stage 4 can rotate around the rotation axis 4r of the stage 4. The rotation axis 4r of the stage 4 is, for example, an axis extending along the normal to the sample placement surface 4a. The stage 4 can move along the normal to the sample placement surface 4a (the rotation axis 4r of the stage 4). When the cooling container 5 is placed on the sample placement surface 4a of the stage 4, the stage 4 can rotate around the central axis 5c of the cooling container 5 extending along the longitudinal direction of the tube 22 of the cooling container 5, and can move along the longitudinal direction of the tube 22 of the cooling container 5 (the central axis 5c of the cooling container 5).

The stage driving device 6 includes a motor (not shown). The stage driving device 6 rotates the stage 4 around the rotation axis 4r of the stage 4. The stage driving device 6 moves the stage 4 along the normal to the sample placement surface 4a (the rotation axis 4r of the stage 4). When the cooling container 5 is placed on the sample placement surface 4a of the stage 4, the stage driving device 6 rotates the cooling container 5 around the central axis 5c of the cooling container 5 extending along the longitudinal direction of the tube 22 of the cooling container 5, and moves the cooling container 5 along the longitudinal direction of the tube 22 of the cooling container 5 (the central axis 5c of the cooling container 5).

The X-ray detector 7 detects the X-rays 3a that have passed through the sample 30. When the sample 30 is contained in the cooling container 5, the X-ray detector 7 detects the X-rays 3a that have passed through the sample 30 and the cooling container 5. The X-ray detector 7 may include a plurality of X-ray detection elements 8. Each of the plurality of X-ray detection elements 8 may include a photodiode (not shown) and a scintillator (not shown) provided on the photodiode. The scintillator converts the X-rays 3a into light. The photodiode detects the light converted from the X-rays by the scintillator. By irradiating the sample 30 with X-rays 3a while rotating the sample 30, the X-ray detector 7 captures an X-ray fluoroscopic image of the sample 30 from all directions around the sample 30.

The computer 10 includes a processor 11 and a storage 14 .
The processor 11 is a CPU (Central Processing Unit). The processor 11 realizes the functions of the control unit 12 and the image reconstruction unit 13 by executing a program stored in the storage 14. The control unit 12 controls the X-ray imaging device 2. The control unit 12 controls the stage driving device 6, so that the stage 4 rotates around the rotation axis 4r of the stage 4 and moves along the normal line (rotation axis 4r of the stage 4) of the sample placement surface 4a. The image reconstruction unit 13 reconstructs X-ray fluoroscopic images of the sample 30 from all directions around the sample 30 using a CT reconstruction algorithm such as back projection or iterative approximation, and generates an X-ray CT image of the sample 30 as an X-ray image 60 of the sample 30.

The storage 14 may be, for example, a non-volatile memory device such as a hard disk or a solid state drive (SSD). The X-ray image 60 of the sample 30 is stored in the storage 14.

<Cooling container 5>
1 to 3, the cooling container 5 is a container for cooling the sample 30. The cooling container 5 has a central axis 5c. When the cooling container 5 is placed on the sample mounting surface 4a of the stage 4, the central axis 5c of the cooling container 5 is coaxial with the rotation axis 4r of the stage 4. The cooling container 5 may have an elongated shape in the direction in which the central axis 5c of the cooling container 5 extends. The longitudinal direction of the cooling container 5 may be the direction in which the central axis 5c of the cooling container 5 extends. The cooling container 5 includes a hollow container 20 and a heat-insulating container 35.

The hollow container 20 includes a tube 22 and a bottom plate 21. The hollow container 20 may further include a fixing member 25.

The tube 22 is formed of a low-density material. In this specification, the low-density material means a material having a density of 2.8 g/cm ³ or less. The low-density material can transmit X-rays. The X-rays 3a transmit through the tube 22. The low-density material may have a density of 2.3 g/cm ³ or less. The low-density material may have a density of 0.9 g/cm ³ or more. The low-density material may have a density of 2.0 g/cm ³ or more. The low-density material may have a density of 0.9 g/cm ³ or more and 2.8 g/cm ³ or less. The low-density material may have a density of 2.0 g/cm ³ or more and 2.3 g/cm ³ or less. The tube 22 is formed of a material that does not become embrittled even when cooled with the cooling liquid 33. The tube 22 is formed of a material that does not become embrittled at a temperature of, for example, −70° C. or more. The tube 22 may be formed of a material that does not become embrittled at a temperature of −100° C. or more. The tube 22 is formed of, for example, a fluorine-based resin such as polytetrafluoroethylene (PTFE) or perfluoroalkoxyalkane (PFA), or a polyethylene resin.

The tube 22 has a cylindrical shape. The tube 22 includes an outer surface 22c and an inner surface 22d opposite the outer surface 22c. X-rays 3a pass through the outer surface 22c and the inner surface 22d. The tube 22 includes a first end 22a and a second end 22b opposite the first end 22a. The outer surface 22c and the inner surface 22d each extend from the first end 22a to the second end 22b. The outer surface 22c has, for example, the shape of a cylindrical outer surface. The inner surface 22d has, for example, the shape of a cylindrical inner surface.

The bottom plate 21 closes the first end 22a of the tube 22. The storage space 22e of the tube 22 is defined by the tube 22 and the bottom plate 21. The bottom plate 21 may be formed of a low-density material. The bottom plate 21 is formed of a material that does not become embrittled even when cooled with the cooling liquid 33. The bottom plate 21 is formed of a material that does not become embrittled at a temperature of, for example, -70°C or higher. The bottom plate 21 may be formed of a material that does not become embrittled at a temperature of -100°C or higher. The bottom plate 21 is formed of a fluorine-based resin such as polytetrafluoroethylene (PTFE) or perfluoroalkoxyalkane (PFA), or a polyethylene resin. The bottom plate 21 may be formed of the same material as the tube 22. A first opening 23 that communicates with the storage space 22e is provided at the second end 22b of the tube 22.

The sample 30 and the cooling liquid 33 are contained in the storage space 22e. The sample 30 is supported by at least one of the tube 22 or the bottom plate 21. The sample 30 may be supported by both the tube 22 and the bottom plate 21.

The sample 30 may be supported relative to the tube 22 via a fixing member 25. The fixing member 25 is in contact with the inner surface 22d of the tube 22. The fixing member 25 may be fitted to the tube 22. A through hole 26 is provided in the fixing member 25. The sample 30 may be fitted into the through hole 26. A second opening 27 is provided in the fixing member 25. The second opening 27 extends along the longitudinal direction of the tube 22 and penetrates the fixing member 25. The fixing member 25 is formed of a low-density material. The fixing member 25 is formed of a material that does not become embrittled even when cooled with a cooling liquid 33. The fixing member 25 is formed of a material that does not become embrittled at a temperature of, for example, -70°C or higher. The fixing member 25 may be formed of a material that does not become embrittled at a temperature of -100°C or higher. The fixing member 25 is formed of, for example, polystyrene foam. The sample 30 may be supported relative to the tube 22 via a plurality of fixing members 25. The multiple fixing members 25 may support, for example, the upper portion of the sample 30 and the lower portion of the sample 30. Therefore, even if the size of the sample 30 is large, the sample 30 is stably supported by the multiple fixing members 25.

The sample 30 is immersed in the cooling liquid 33. Specifically, a part of the sample 30 may be immersed in the cooling liquid 33, or the entire sample 30 may be immersed in the cooling liquid 33. The cooling liquid 33 cools the sample 30 to a low temperature lower than room temperature. In this specification, room temperature means a temperature of 15°C or higher and 30°C or lower. The low temperature may be, for example, a temperature below 0°C, a temperature of -10°C or lower, a temperature of -20°C or lower, a temperature of -30°C or lower, or a temperature of -40°C or lower. The cooling liquid 33 contacts the inner surface 22d of the tube 22. The cooling liquid 33 is, for example, a mixture of dry ice and alcohol (e.g., ethanol).

The insulating container 35 accommodates the hollow container 20. The insulating container 35c is in contact with the outer surface 22c of the pipe 22. The insulating container 35 may cover the entire outer surface 22c of the pipe 22. The insulating container 35 may cover the entire outer surface of the hollow container 20. The insulating container 35 may include a first insulating member 36 and a second insulating member 38. The first insulating member 36 is in contact with the bottom plate 21. A recess 37 may be formed in the first insulating member 36. Specifically, the first insulating member 36 may include a base 36a and an annular protruding portion 36b protruding from the base 36a. The recess 37 is the inner space of the annular protruding portion 36b. The second insulating member 38 is in contact with the pipe 22 (outer surface 22c) and the first insulating member 36. The second insulating member 38 is fitted into the recess 37.

The insulating container 35 is formed of a low-density material. Therefore, the insulating container 35 can transmit the X-rays 3a. The X-rays 3a transmit through the insulating container 35. The insulating container 35 may be formed of a material that does not become brittle even when cooled with the cooling liquid 33. The insulating container 35 may be formed of a material that does not become brittle at a temperature of -70°C or higher. The insulating container 35 may be formed of a material that does not become brittle at a temperature of -100°C or higher. The insulating container 35 is formed of, for example, polystyrene foam. The insulating container 35 includes an outer surface 35s. The X-rays 3a transmit through a portion of the outer surface 35s that covers the tube 22 (for example, the outer surface of the second insulating member 38). The portion of the outer surface 35s that covers the tube 22 has the shape of a cylindrical outer surface. The insulating container 35 is provided with a through hole 39 that communicates with the first opening 23. The through hole 39 is provided in, for example, the second insulating member 38.

The cooling container 5 does not need to have a temperature control function. Therefore, while an X-ray fluoroscopic image of the sample 30 is being captured, the temperatures of the sample 30 and the cooling liquid 33 gradually rise. Part of the cooling liquid 33 turns into gas. This gas passes through the second opening 27 of the fixing member 25, the first opening 23 of the tube 22, and the through hole 39 of the insulating container 35 and is discharged to the outside of the cooling container 5. The second opening 27 of the fixing member 25, the first opening 23 of the tube 22, and the through hole 39 of the insulating container 35 function as gas vents for the gas, preventing the cooling container 5 from being damaged by the gas.

<Method of acquiring X-ray image 60 of sample 30>
1 to 4, an example of a method for acquiring an X-ray image 60 of a sample 30 cooled to a temperature lower than room temperature using an X-ray device 1. When the X-ray device 1 is an X-ray CT device, the X-ray image 60 of the sample 30 is an X-ray CT image of the sample 30.

The method for acquiring an X-ray image 60 of a sample 30 in this embodiment includes placing the sample 30 and a cooling liquid 33 in a cooling container 5 (step S1). The sample 30 is supported by at least one of the tube 22 or the bottom plate 21. The sample 30 may be supported by both the tube 22 and the bottom plate 21. The sample 30 may be supported relative to the tube 22 via a fixing member 25. The cooling liquid 33 cools the sample 30 to a low temperature that is lower than room temperature.

The method for acquiring the X-ray image 60 of the sample 30 in this embodiment includes placing the cooling container 5 on the stage 4 (step S2). Specifically, the cooling container 5 is placed on the sample placement surface 4a of the stage 4. The central axis 5c of the cooling container 5 is coaxial with the rotation axis 4r of the stage 4.

The method for acquiring the X-ray image 60 of the sample 30 in this embodiment includes moving the cooling container 5 in the longitudinal direction of the tube 22 (step S3). Specifically, the control unit 12 controls the stage driving device 6. The stage 4 moves along the longitudinal direction of the tube 22. In this way, the portion of the sample 30 for which an X-ray fluoroscopic image is desired is positioned on the path of the X-rays 3a.

The method for acquiring the X-ray image 60 of the sample 30 in this embodiment includes capturing an X-ray fluoroscopic image of the sample 30 while rotating the cooling container 5 (step S4). Specifically, the control unit 12 controls the stage driving device 6. The stage 4 rotates around the rotation axis 4r of the stage 4 (the central axis 5c of the cooling container 5). The cooling container 5 placed on the sample placement surface 4a of the stage 4 rotates around the central axis 5c of the cooling container 5 (the rotation axis 4r of the stage 4). The sample 30 contained in the cooling container 5 rotates around the central axis 5c of the cooling container 5 (the rotation axis 4r of the stage 4). By irradiating the sample 30 with X-rays 3a while rotating the sample 30, the X-ray detector 7 captures X-ray fluoroscopic images of the sample 30 from all directions around the sample 30. The X-ray fluoroscopic images of the sample 30 are stored in the storage 14.

The method of acquiring the X-ray image 60 of the sample 30 in this embodiment includes judging whether or not the acquisition of X-ray fluoroscopic images of all parts of the sample 30 for which X-ray fluoroscopic images are to be obtained has been completed (step S5). For example, the control unit 12 executes step S5. If the acquisition of X-ray fluoroscopic images of some parts of all parts of the sample 30 for which X-ray fluoroscopic images are to be obtained has not been completed (NO in step S5), steps S3 and S4 are executed. In this manner, steps S3 and S4 are repeated until X-ray fluoroscopic images of all parts of the sample 30 for which X-ray fluoroscopic images are to be obtained are obtained.

When the capturing of X-ray fluoroscopic images of all parts of the sample 30 for which X-ray fluoroscopic images are to be obtained has been completed (YES in step S5), the X-ray fluoroscopic images of the sample 30 are reconstructed to generate an X-ray CT image of the sample 30 as an X-ray image 60 of the sample 30 (step S6). Specifically, the image reconstruction unit 13 reconstructs X-ray fluoroscopic images of the sample 30 from all directions around the sample 30 using a CT reconstruction algorithm such as back projection or iterative approximation to generate an X-ray CT image of the sample 30. The X-ray image 60 of the sample 30 is stored in the storage 14.

The X-ray image 60 of the sample 30 is obtained by imaging the sample 30 for a first time at a low temperature lower than room temperature. In this embodiment, the cooling container 5 does not have a temperature adjustment function. Therefore, the temperature of the sample 30 and the cooling liquid 33 gradually increases while the X-ray image 60 of the sample 30 is being imaged. In order to image the X-ray image 60 of the sample 30 at a low temperature lower than room temperature, it is necessary to shorten the first time. The first time is defined by a temperature change ΔT that is acceptable for X-ray imaging of the sample 30 at a low temperature lower than room temperature. The acceptable temperature change ΔT may be, for example, 10° C. or less, 8° C. or less, 5° C. or less, 3° C. or less, or 1° C. or less. If the first time is shortened, the X-ray image 60 of the sample 30 cooled to a low temperature lower than room temperature becomes unclear.

Referring to FIG. 5, an example of an X-ray image 60 of sample 30 is shown. In FIG. 5, sample 30 is an optical fiber cable including a plurality of optical fibers and a sheath covering the plurality of optical fibers. X-ray image 60 of sample 30 is an X-ray CT image of the optical fiber cable obtained by irradiating it with X-rays for a first time at a low temperature of -30°C to -40°C. The white areas in the X-ray image 60 of sample 30 shown in FIG. 5 are the plurality of optical fibers, and the black background area in the X-ray image 60 of sample 30 shown in FIG. 5 is the sheath.

(Modification of X-ray device 1 and method for acquiring X-ray image 60 of sample 30)
The X-ray device 1 may be, for example, an X-ray fluoroscopic image capturing device from which the image reconstruction unit 13 is omitted, and the X-ray image 60 may be, for example, an X-ray fluoroscopic image. When an X-ray fluoroscopic image is captured using an X-ray fluoroscopic image capturing device, the step of rotating the cooling container 5 in step S4 and the step S6 are omitted from the method of capturing the X-ray image 60 of the sample 30 shown in FIG.

(Method of acquiring multiple first reference X-ray images 61 and multiple second reference X-ray images 62)
1, 6 and 7, the multiple first reference X-ray images 61 are acquired by imaging multiple sites of the reference sample 31 for a second time using the X-ray device 1 without cooling the reference sample 31. Imaging the reference sample 31 using the X-ray device 1 without cooling the reference sample 31 means, for example, imaging an X-ray image of the reference sample 31 in the cooling container 5 without putting the cooling liquid 33 in the cooling container 5 (see FIG. 6), or imaging an X-ray image of the reference sample 31 by placing the reference sample 31 on the sample mounting surface 4a of the stage 4 without using the cooling container 5. The second time is approximately the same as the first time, which is the imaging time of the X-ray image 60 of the sample 30 at a low temperature lower than room temperature. The second time is, for example, 0.5 times or more and 2.0 times or less of the first time. The multiple first reference X-ray images 61 may be acquired, for example, by irradiating the reference sample 31 with X-rays 3a for a second time equal to the first time at room temperature using the X-ray device 1. The reference sample 31 may be different from or the same as the sample 30. The multiple first reference X-ray images 61 are stored in the storage 14.

Referring to FIG. 7, an example of a first reference X-ray image 61 is shown. In FIG. 7, the reference sample 31 is an optical fiber cable including a plurality of optical fibers and a sheath covering the plurality of optical fibers. The first reference X-ray image 61 is an X-ray CT image of the optical fiber cable obtained by irradiating the cable with X-rays at room temperature for a second time period that is the same as the first time period. Because the second time period is comparable to the first time period, the first reference X-ray image 61 is unclear.

1, 6 and 8, the multiple second reference X-ray images 62 are acquired by imaging multiple locations of the reference sample 31 for a third time period using the X-ray device 1 without cooling the reference sample 31. The multiple second reference X-ray images 62 may be acquired, for example, by irradiating the reference sample 31 with X-rays 3a for a third time period at room temperature using the X-ray device 1. The multiple second reference X-ray images 62 are stored in the storage 14.

The imaging conditions of the first reference X-ray image 61 and the second reference X-ray image 62 differ mainly in the imaging time (irradiation time of the X-ray 3a to the reference sample 31). That is, the third time which is the imaging time of the second reference X-ray image 62 is longer than the second time which is the imaging time of the first reference X-ray image 61. The third time may be, for example, twice or more than the second time, 2.5 times or more than the second time, 4 times or more than the second time, 5 times or more than the second time, 8 times or more than the second time, or 10 times or more than the second time. Since the third time is longer than the second time, the second reference X-ray image 62 is clearer than the first reference X-ray image 61.

The third time, which is the imaging time of the second reference X-ray image 62, is longer than the first time, which is the imaging time of the X-ray image 60 of the sample 30. The third time may be, for example, four times or more than the first time, five times or more than the first time, eight times or more than the first time, or ten times or more than the first time.

Referring to FIG. 8, an example of the second reference X-ray image 62 is shown. In FIG. 8, the reference sample 31 is the optical fiber cable described above. The second reference X-ray image 62 is an X-ray CT image of the optical fiber cable obtained by irradiating it with X-rays for a third time period at room temperature. Because the third time period is longer than the second time period, the second reference X-ray image 62 is clearer than the first reference X-ray image 61.

Although the temperature of the reference sample 31 when the second reference X-ray image 62 is captured is preferably equal to the temperature of the reference sample 31 when the first reference X-ray image 61 is captured, the temperature of the reference sample 31 when the second reference X-ray image 62 is captured may be different from the temperature of the reference sample 31 when the first reference X-ray image 61 is captured. If the temperature of the reference sample 31 when the second reference X-ray image 62 is captured is different from the temperature of the reference sample 31 when the first reference X-ray image 61 is captured, the difference between the temperature of the reference sample 31 when the multiple second reference X-ray images 62 are captured and the temperature of the reference sample 31 when the multiple first reference X-ray images 61 are captured is smaller than the difference between the temperature of the reference sample 31 when the multiple first reference X-ray images 61 are captured and the temperature (low temperature) of the sample 30 when the X-ray image 60 of the sample 30 is captured. The difference between the temperature of the reference sample 31 when the multiple second reference X-ray images 62 are taken and the temperature of the reference sample 31 when the multiple first reference X-ray images 61 are taken may be less than half, less than one-third, or less than one-quarter of the difference between the temperature of the reference sample 31 when the multiple first reference X-ray images 61 are taken and the temperature (low temperature) of the sample 30 when the X-ray image 60 of the sample 30 is taken.

The difference between the temperature of the reference sample 31 when the multiple second reference X-ray images 62 are captured and the temperature of the reference sample 31 when the multiple first reference X-ray images 61 are captured may be, for example, 5°C or less, 3°C or less, or 1°C or less. The difference between the temperature of the reference sample 31 when the multiple first reference X-ray images 61 are captured and the temperature (low temperature) of the sample 30 when the X-ray image 60 of the sample 30 is captured may be, for example, 10°C or more, 20°C or more, 30°C or more, or 40°C or more.

A reference image ID (not shown) assigned to each portion of the reference sample 31 is stored in the storage 14 together with the first reference X-ray image 61 and the second reference X-ray image 62 corresponding to each portion of the reference sample 31.

<Image Processing System 90>
As already described, the temperatures of the sample 30 and the cooling liquid 33 gradually increase while the X-ray image 60 of the sample 30 is being captured. In order to capture the X-ray image 60 of the sample 30 at a low temperature lower than room temperature, it is necessary to shorten the first time for capturing the X-ray image 60 of the sample 30. Therefore, the X-ray image 60 of the sample 30 cooled to a low temperature lower than room temperature becomes unclear. Therefore, image processing is required to make the X-ray image 60 of the sample 30 cooled to a low temperature lower than room temperature clear. The image processing to make the X-ray image 60 of the sample 30 clear is, for example, super-resolution processing to improve the resolution of the X-ray image 60 of the sample 30. With reference to FIG. 9, an image processing system 90 for making the X-ray image 60 of the sample 30 cooled to a low temperature lower than room temperature clear will be described.

The image processing system 90 includes a server device 100 and a terminal device 200. The image processing system 90 is connected to a computer 10 (see FIG. 1) of the X-ray device 1 via a communication network 50 such as the Internet.

(Server device 100)
The server device 100 obtains a reference image ID (not shown), a plurality of first reference X-ray images 61, and a plurality of second reference X-ray images 62 from the computer 10 of the X-ray device 1 (see Figure 1) via the communication network 50.

The server device 100 generates a trained model 113 for performing image processing to make the X-ray image 60 of the sample 30 clear by a learning process using the first reference X-ray image 61 and the second reference X-ray image 62. The generated trained model 113 is stored in the storage 110 (see FIG. 10 ) and is also transmitted (distributed) to the terminal device 200.

(Terminal device 200)
The terminal device 200 acquires the trained model 113 from the server device 100 via the communication network 50. The trained model 113 is stored in the storage 210 (see FIG. 11). The terminal device 200 acquires the X-ray image 60 of the sample 30 from the computer 10 of the X-ray device 1. The X-ray image 60 of the sample 30 is stored in the storage 210. The terminal device 200 inputs the X-ray image 60 of the sample 30 to the trained model 113. The X-ray image 60 of the sample 30 is image-processed by the trained model 113. The trained model 113 outputs a corrected X-ray image 66 of the sample 30 (see FIGS. 11 and 17) that is clearer than the X-ray image 60 of the sample 30. The corrected X-ray image 66 of the sample 30 is stored in the storage 210 or displayed on the display 204 (see FIG. 11).

<Hardware Configuration>
(Server device 100)
Referring to FIG. 10, the server device 100 includes an input device 101 , a processor 102 , a memory 103 , a display 104 , a network controller 106 , a storage medium drive 107 , and a storage 110 .

The input device 101 accepts various input operations. The input device 101 is, for example, a keyboard, a mouse, or a touch panel.

The display 104 displays information necessary for processing in the server device 100. The display 104 may display, for example, the first reference X-ray image 61, the second reference X-ray image 62, and the provisionally corrected X-ray image 65. The display 104 is, for example, an LCD (Liquid Crystal Display) or an organic EL (Electroluminescence) display.

The processor 102 executes the programs described below to perform the processes required to realize the functions of the server device 100. The processor 102 is composed of, for example, one or more CPUs or GPUs (Graphics Processing Units). A CPU or GPU having multiple cores may be used as the processor 102. In the server device 100, it is preferable to employ a GPU or the like that is suitable for the learning process for generating the trained model 113.

The memory 103 provides a storage area for temporarily storing program code or work memory when the processor 102 executes programs including the data preprocessing program 111 and the learning program 112. The memory 103 is, for example, a volatile memory device such as a DRAM (Dynamic Random Access Memory) or an SRAM (Static Random Access Memory).

The network controller 106 transmits and receives programs or data to and from any device including the X-ray device 1 and the terminal device 200 via the communication network 50 (see FIG. 9). For example, the network controller 106 receives data such as a reference image ID (not shown), a plurality of first reference X-ray images 61, and a plurality of second reference X-ray images 62 from the X-ray device 1 via the communication network 50. The network controller 106 transmits the trained model 113 to the terminal device 200 via the communication network 50. The network controller 106 supports any communication method such as Ethernet (registered trademark), wireless LAN (Local Area Network), or Bluetooth (registered trademark).

The storage medium drive 107 is a device that reads out a program or data stored in the storage medium 108. The storage medium drive 107 may also be a device that writes a program or data to the storage medium 108. The storage medium 108 is a non-transitory storage medium that stores the program or data in a non-volatile manner. The storage medium 108 is, for example, an optical storage medium such as an optical disk (e.g., a CD-ROM or a DVD-ROM), a semiconductor storage medium such as a flash memory or a USB (Universal Serial Bus) memory, a magnetic storage medium such as a hard disk, a FD (Flexible Disk) or a storage tape, or a magneto-optical storage medium such as an MO (Magneto-Optical) disk.

The storage 110 stores programs executed by the processor 102 (including, for example, a data preprocessing program 111 and a learning program 112), a reference image ID (not shown), a plurality of first reference X-ray images 61, a plurality of second reference X-ray images 62, a trained model 113, and a training dataset 115. The data preprocessing program 111 is a program for generating the training dataset 115. The training program 112 is a program for generating the trained model 113 using the training dataset 115. The storage 110 is, for example, a non-volatile memory device such as a hard disk or SSD.

Programs for realizing the functions of the server device 100, including the data preprocessing program 111 and the learning program 112, may be stored in a non-transient storage medium, distributed, and installed in the storage 110. Programs for realizing the functions of the server device 100 may be downloaded to the server device 100 via the Internet or an intranet.

In this embodiment, an example is shown in which a general-purpose computer (processor 102) executes programs including a data preprocessing program 111 and a learning program 112 to realize the functions of the server device 100, but this is not limited to the above, and all or part of the functions of the server device 100 may be realized using an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field-Programmable Gate Array).

(Terminal device 200)
Referring to FIG. 11, the terminal device 200 includes an input device 201 , a processor 202 , a memory 203 , a display 204 , a network controller 206 , a storage medium drive 207 , and a storage 210 .

The input device 201 accepts various input operations. The input device 201 is, for example, a keyboard, a mouse, or a touch panel.

The display 204 displays information necessary for processing in the terminal device 200. The display 204 displays, for example, a corrected X-ray image 66 of the sample 30. The display 204 is, for example, an LCD (Liquid Crystal Display) or an organic EL (Electroluminescence) display.

The processor 202 executes the programs described below to perform the processes required to realize the functions of the terminal device 200. The processor 202 is composed of, for example, one or more CPUs or GPUs. A CPU or GPU having multiple cores may be used as the processor 202.

The memory 203 provides a storage area for temporarily storing program code or work memory when the processor 202 executes a program. The memory 203 is, for example, a volatile memory device such as a DRAM or SRAM.

The network controller 206 transmits and receives programs or data between any device including the X-ray device 1 and the server device 100 via the communication network 50 (see FIG. 9). For example, the network controller 206 receives data such as an X-ray image 60 of the sample 30 from the X-ray device 1 via the communication network 50. The network controller 206 receives the trained model 113 from the server device 100 via the communication network 50. The network controller 206 is compatible with any communication method such as Ethernet (registered trademark), wireless LAN, or Bluetooth (registered trademark).

The storage medium drive 207 is a device that reads out a program or data stored in the storage medium 208. The storage medium drive 207 may also be a device that writes a program or data to the storage medium 208. The storage medium 208 is a non-transitory storage medium that stores the program or data in a non-volatile manner. The storage medium 208 is, for example, an optical storage medium such as an optical disk (e.g., a CD-ROM or a DVD-ROM), a semiconductor storage medium such as a flash memory or a USB memory, a magnetic storage medium such as a hard disk, a FD or a storage tape, or a magneto-optical storage medium such as an MO disk.

The storage 210 stores programs executed by the processor 202 (including, for example, the image processing program 211), the trained model 113, the X-ray image 60 of the sample 30, and the corrected X-ray image 66 of the sample 30. The image processing program 211 is a program for inputting the X-ray image 60 of the sample 30 into the trained model 113 to generate the corrected X-ray image 66 of the sample 30. The storage 210 is, for example, a non-volatile memory device such as a hard disk or SSD.

Programs for realizing the functions of the terminal device 200, including the image processing program 211, may be stored in a non-transient storage medium, distributed, and installed in the storage 210. Programs for realizing the functions of the terminal device 200 may be downloaded to the terminal device 200 via the Internet or an intranet.

In this embodiment, an example is shown in which a general-purpose computer (processor 202) executes a program including an image processing program 211 to realize the functions of the terminal device 200, but this is not limited to the above, and all or part of the functions of the terminal device 200 may be realized using an integrated circuit such as an ASIC or FPGA.

<Generation of trained model 113>
The learning process (a method for generating the trained model 113) of the present embodiment will be described with reference to Fig. 9 and Fig. 12 to Fig. 16. The learning process of the present embodiment is executed by, for example, the server device 100. The server device 100 functions as a trained model generation device.

Referring to FIG. 12, the server device 100 includes a reference X-ray image receiving unit 120, a data preprocessing unit 121, a learning unit 122, and an output unit 127. The learning unit 122 includes a learning model 123 and a learning program 125. The learning model 123 is composed of a neural network structure 123N and parameters 123P. The neural network structure 123N is a neural network structure classified as a deep neural network (DNN), and includes, for example, a convolutional neural network (CNN) structure. The neural network structure 123N may include multiple convolutional layers. The neural network structure 123N is constructed in advance and stored in the server device 100 (storage 110).

The reference X-ray image receiving unit 120 receives a reference image ID, a plurality of first reference X-ray images 61, and a plurality of second reference X-ray images 62 from the X-ray device 1. The reference X-ray image receiving unit 120 stores the reference image ID, the plurality of first reference X-ray images 61, and the plurality of second reference X-ray images 62 in the storage 110.

The data preprocessing unit 121 reads out a reference image ID, a plurality of first reference X-ray images 61, and a plurality of second reference X-ray images 62 from the storage 110. The data preprocessing unit 121 generates a learning data set 115 by associating the reference image ID, the first reference X-ray image 61 corresponding to the reference image ID among the plurality of first reference X-ray images 61, and the second reference X-ray image 62 corresponding to the reference image ID among the plurality of second reference X-ray images 62. As shown in FIG. 13, the learning data set 115 includes a plurality of data sets (learning data 116). Each of the plurality of data sets (each of the learning data 116) includes a reference image ID, a first reference X-ray image 61 corresponding to the reference image ID, and a second reference X-ray image 62 corresponding to the reference image ID. To improve the accuracy of the learning of the learning model 123, the data preprocessing unit 121 may upscale the multiple first reference X-ray images 61 and the multiple second reference X-ray images 62 to a desired size using a bicubic interpolation method or the like.

Referring to FIG. 12 and FIG. 14, the learning unit 122 generates a trained model 113. The learning unit 122 updates the values of the parameters 123P of the training model 123 through machine learning using the training dataset 115.

Specifically, the learning unit 122 updates the value of the parameter 123P by using the learning program 125. The learning program 125 inputs the first reference X-ray image 61 of the learning data set 115 to the learning model 123. The learning model 123 outputs the provisionally corrected X-ray image 65. An example of the provisionally corrected X-ray image 65 is shown with reference to FIG. 15. In FIG. 15, the reference sample 31 is the optical fiber cable described above. The learning program 125 repeatedly updates and optimizes the parameter 123P so as to minimize the error between the provisionally corrected X-ray image 65 and the corresponding second reference X-ray image 62. Specifically, the parameter 123P of the learning model 123 is repeatedly updated and optimized so as to minimize the error between the provisionally corrected X-ray image 65 obtained by inputting the first reference X-ray image 61 with a reference image ID of "1" to the learning model 123 and the second reference X-ray image 62 with a reference image ID of "1". Similarly, for each of the other reference image IDs, the parameters 123P of the learning model 123 are repeatedly updated and optimized to minimize the error between the provisionally corrected X-ray image 65 and the corresponding second reference X-ray image 62. When the learning of the learning model 123 is completed, the learned model 113 is generated.

Referring to FIG. 17, the trained model 113 has a neural network structure 113N and trained parameters 113P. The neural network structure 113N is a neural network structure classified as a deep neural network (DNN), and includes, for example, a convolutional neural network (CNN) structure. The neural network structure 113N may include multiple convolutional layers. The trained parameters 113P are parameters 123P for which updating has been completed.

Referring to FIG. 12, the output unit 127 outputs the trained model 113. Specifically, the output unit 127 stores the trained model 113 in the storage 110. The output unit 127 transmits the trained model 113 stored in the storage 110 to the terminal device 200 via the network controller 206.

The learning processing method of this embodiment will be described with reference to FIG. 16. The learning processing method of this embodiment is executed, for example, in the server device 100. Each step shown in FIG. 16 is realized, for example, by the processor 102 of the server device 100 executing a program including a data preprocessing program 111 (see FIG. 10) and a learning program 112 (see FIG. 10).

The learning processing method of this embodiment includes receiving a reference image ID, a plurality of first reference X-ray images 61, and a plurality of second reference X-ray images 62 from the X-ray device 1 (step S11). The reference X-ray image receiving unit 120 (see FIG. 12) executes step S11. The reference X-ray image receiving unit 120 stores the reference image ID, the plurality of first reference X-ray images 61, and the plurality of second reference X-ray images 62 in the storage 110.

The learning processing method of this embodiment includes generating a learning dataset 115 (step S12). The learning dataset 115 is generated by associating a reference image ID with a first reference X-ray image 61 corresponding to the reference image ID among the multiple first reference X-ray images 61 and a second reference X-ray image 62 corresponding to the reference image ID among the multiple second reference X-ray images 62. The data preprocessing unit 121 (see FIG. 12) reads out the reference image ID, the multiple first reference X-ray images 61, and the multiple second reference X-ray images 62 stored in the storage 110, and executes step S12. In order to improve the accuracy of learning of the learning model 123, the data preprocessing unit 121 may upscale the multiple first reference X-ray images 61 and the multiple second reference X-ray images 62 to a desired size using a bicubic interpolation method or the like. The data preprocessing unit 121 stores the learning dataset 115 in the storage 110.

The learning processing method of this embodiment includes inputting the learning data set 115 into the learning model 123 to generate a learned model 113 (step S13). The learning unit 122 (see FIG. 12) executes step S13. Specifically, step S13 includes steps S14 and S15. In step S14, the first reference X-ray image 61 included in the learning data set 115 is input into the learning model 123 to generate a provisionally corrected X-ray image 65. In step S14, the parameters 123P of the learning model 123 are repeatedly updated and optimized so as to minimize the error between the provisionally corrected X-ray image 65 and the corresponding second reference X-ray image 62. For example, the parameters 123P of the learning model 123 are repeatedly updated and optimized so as to minimize the error between the provisionally corrected X-ray image 65 obtained by inputting the first reference X-ray image 61 having a reference image ID of "1" into the learning model 123 and the second reference X-ray image 62 having a reference image ID of "1". Similarly, for each of the other reference image IDs, the parameters 123P of the learning model 123 are repeatedly updated and optimized so as to minimize the error between the provisionally corrected X-ray image 65 and the corresponding second reference X-ray image 62. In this way, the learned model 113 is generated from the learning model 123.

The learning processing method of this embodiment includes outputting the learned model 113 (step S16). The output unit 127 (see FIG. 12) executes step S16. Step S16 includes, for example, steps S17 and S18. In step S17, the output unit 127 stores the learned model 113 in the storage 210. In step S18, the output unit 127 transmits the learned model 113 to the terminal device 200. Specifically, the output unit 127 reads out the learned model 113 from the storage 210 and transmits the learned model 113 to the terminal device 200 via the network controller 206. The learned model 113 is stored in the storage 210 (see FIG. 11). In this way, the learning processing method of this embodiment is completed.

<Image processing using trained model 113>
9 and 17 to 19, the image processing of the present embodiment using the trained model 113 will be described. The image processing of the present embodiment is executed by, for example, the terminal device 200. The terminal device 200 functions as an X-ray image processing device.

As shown in FIG. 17, the terminal device 200 includes an X-ray image receiving unit 220, an image processing unit 221, and an output unit 227. The image processing unit 221 includes the trained model 113.

The X-ray image receiving unit 220 receives the X-ray image 60 of the sample 30 from the X-ray device 1. The X-ray image receiving unit 220 stores the X-ray image 60 of the sample 30 in the storage 210.

The image processing unit 221 inputs the X-ray image 60 of the sample 30 into the learned model 113 to generate a corrected X-ray image 66 of the sample 30. The learned model 113 is learned by the first reference X-ray image 61 captured over a second time period that is approximately the same length as the first time period, and the second reference X-ray image 62 captured over a third time period that is longer than the first time period and the second time period. Therefore, the clarity of the corrected X-ray image 66 of the sample 30 is improved compared to the clarity of the X-ray image 60 of the sample 30. For example, the corrected X-ray image 66 of the sample 30 has a clarity equivalent to that of the second reference X-ray image 62. The corrected X-ray image 66 of the sample 30 has a clarity equivalent to that of the X-ray image 60 of the sample 30 acquired by capturing an image of the sample 30 over a third time period that is the capturing time of the second reference X-ray image 62 at a low temperature lower than room temperature. An example of the corrected X-ray image 66 is shown with reference to FIG. 18. In Figure 18, sample 30 is the optical fiber cable described above.

The output unit 227 outputs the corrected X-ray image 66 of the sample 30. For example, the output unit 227 stores the corrected X-ray image 66 of the sample 30 in the storage 210. The output unit 227 displays the corrected X-ray image 66 of the sample 30 on the display 204 (see FIG. 11).

The X-ray image processing method of this embodiment will be described with reference to FIG. 19. The X-ray image processing method of this embodiment is executed, for example, in the terminal device 200. Each step shown in FIG. 19 is realized, for example, by the processor 202 of the terminal device 200 executing a program including the image processing program 211 (see FIG. 11).

The X-ray image processing method of this embodiment includes receiving an X-ray image 60 of the sample 30 from the X-ray device 1 (step S21). The X-ray image receiving unit 220 executes step S21. The X-ray image receiving unit 220 stores the X-ray image 60 of the sample 30 in the storage 210.

The X-ray image processing method of this embodiment includes inputting the X-ray image 60 of the sample 30 into the trained model 113 to generate a corrected X-ray image 66 (step S22). The image processing unit 221 executes step S22. The trained model 113 is trained using a first reference X-ray image 61 captured over a second time period that is approximately the same length as the first time period, and a second reference X-ray image 62 captured over a third time period that is longer than the first time period and the second time period. Therefore, the clarity of the corrected X-ray image 66 of the sample 30 is improved compared to the clarity of the X-ray image 60 of the sample 30.

The X-ray image processing method of this embodiment includes outputting the corrected X-ray image 66 (step S23). Step S23 includes, for example, storing the corrected X-ray image 66 in the storage 210 (step S24) and displaying the corrected X-ray image 66 on the display 204 (step S25). In this way, the X-ray image processing method of this embodiment is completed.

(Modifications of Image Processing System 90 and Image Processing Method)
In the image processing system 90 of the first modification, the server device 100 and the terminal device 200 may not be communicatively connected to the X-ray device 1 via the communication network 50. In the image processing system 90 of the first modification, the reference image ID, the plurality of first reference X-ray images 61, and the plurality of second reference X-ray images 62 may be transferred from the X-ray device 1 (storage 14) to the server device 100 (storage 110) using a non-transient storage medium. In the image processing system 90 of the first modification, the X-ray image 60 of the sample 30 acquired by imaging the sample 30 for a first time at a low temperature lower than room temperature may be transferred from the X-ray device 1 (storage 14) to the terminal device 200 (storage 210) using a non-transient storage medium.

The image processing system 90 of the second modified example includes a server device 100, in which the trained model 113 is generated and the image processing that generated the trained model 113 is performed. The image processing method of the second modified example is executed by the server device 100. The image processing system 90 of the third modified example includes a terminal device 200, in which the trained model 113 is generated and the image processing that generated the trained model 113 is performed. The image processing method of the third modified example is executed by the terminal device 200.

The embodiments disclosed herein are illustrative in all respects and should not be considered limiting. The scope of the present disclosure is indicated by the claims rather than the above-described embodiments, and is intended to include all modifications within the scope of the claims and meanings equivalent thereto.

1 X-ray device 2 X-ray imaging device 3 X-ray source 3a X-ray 4 Stage 4a Sample placement surface 4r Rotation axis 5 Cooling container 5c Central axis 6 Stage driving device 7 X-ray detector 8 X-ray detection element 10 Computer 11 Processor 12 Control unit 13 Image reconstruction unit 14 Storage 20 Hollow container 21 Bottom plate 22 Tube 22a First end 22b Second end 22c Outer surface 22d Inner surface 22e Storage space 23 First opening 25 Fixing member 26 Through hole 27 Second opening 30 Sample 31 Reference sample 33 Cooling liquid 35 Insulating container 35s Outer surface 36 First insulating member 36a Base 36b Annular protrusion 37 Recess 38 Second insulating member 39 Through hole 50 Communication network 61 First reference X-ray image 62 Second reference X-ray image 60 X-ray image 65 Temporary corrected X-ray image 66 Corrected X-ray image 90 Image processing system 100 Server device 101 Input device 102 Processor 103 Memory 104 Display 106 Network controller 107 Storage medium drive 108 Storage medium 110 Storage 111 Pre-processing program 112, 125 Learning program 113 Learned model 113N, 123N Neural network structure 113P Learned parameters 115 Learning data set 116 Learning data 120 Reference X-ray image receiving unit 121 Pre-processing unit 122 Learning unit 123 Learning model 123P Parameters 127, 227 Output unit 200 Terminal device 201 Input device 202 Processor 203 Memory 204 Display 206 Network controller 207 Storage medium drive 208 Storage medium 210 Storage 211 Image processing program 220 Reception unit 221 Image processing unit

Claims (15)

A cooling vessel, the cooling vessel comprising:
a hollow vessel including a tube having a first end and a second end opposite the first end, and a bottom plate closing the first end;
and a heat-insulating container that houses the hollow container,
a sample and a cooling liquid for cooling the sample are accommodated in an accommodation space defined by the tube and the bottom plate;
the sample is immersed in the cooling liquid and supported by at least one of the tube or the bottom plate;
The tube and the insulating container are made of a material that is transparent to X-rays and has a density of 2.8 g/cm3 or less ^;
A first opening communicating with the accommodation space is provided at the second end,
The heat-insulating container has a through hole communicating with the first opening. the tube has a cylindrical shape;
The cooling vessel according to claim 1 , wherein an outer surface of the portion of the insulating vessel that covers the tube has a cylindrical outer surface shape. The cooling vessel of claim 1 or claim 2, wherein the tube is in contact with both the cooling liquid and the insulating vessel. The cooling vessel according to any one of claims 1 to 3, wherein the insulating vessel covers the entire outer surface of the tube. The cooling container according to any one of claims 1 to 4, wherein the insulating container covers all of the outer surfaces of the hollow container. The cooling container according to any one of claims 1 to 5, wherein the insulated container includes a first insulating member in contact with the bottom plate and a second insulating member in contact with the tube and the first insulating member. A recess is formed in the first heat insulating member,
The cooling vessel according to claim 6 , wherein the second insulating member is fitted into the recess. The cooling container according to any one of claims 1 to 7, wherein the tube is made of a fluororesin or a polyethylene resin. The cooling vessel of claim 8, wherein the tube is formed of polytetrafluoroethylene or perfluoroalkoxyalkane. The cooling container according to any one of claims 1 to 9, wherein the insulating container is made of polystyrene foam. The cooling container according to any one of claims 1 to 10, wherein the cooling liquid is a mixture of dry ice and alcohol. The hollow vessel further includes a fixing member,
the sample is supported relative to the tube via the fixing member;
The cooling vessel according to claim 1 , wherein the fixing member is provided with a second opening extending along the longitudinal direction of the tube and penetrating the fixing member. An X-ray imaging device;
The cooling container according to any one of claims 1 to 11,
The X-ray imaging device is an X-ray device including an X-ray source that emits X-rays, a stage that supports the cooling container, and an X-ray detector that detects the X-rays that have passed through the cooling container. The X-ray device according to claim 13, wherein the stage is capable of rotating the cooling vessel around a central axis of the cooling vessel that extends along the longitudinal direction of the tube, and is also capable of moving along the longitudinal direction of the tube. obtaining an X-ray image of the sample by imaging the sample at a cryogenic temperature below room temperature for a first period of time using the X-ray device of claim 13 or 14;
inputting the X-ray image of the sample into a neural network to generate a corrected X-ray image of the sample that is sharper than the X-ray image of the sample;
The neural network is generated by training using a training data set;
the training data set includes a first reference X-ray image acquired by imaging the reference sample with the X-ray device for a second time period without cooling the reference sample, and a second reference X-ray image acquired by imaging the reference sample with the X-ray device for a third time period without cooling the reference sample;
the second time period is 0.5 to 2.0 times the first time period,
The X-ray image processing method, wherein the third time period is longer than the first time period and longer than the second time period.