JP7236236B2 - Medical image processing device, X-ray diagnostic device, and program - Google Patents

Description

TECHNICAL FIELD Embodiments of the present invention relate to a medical image processing apparatus, an X-ray diagnostic apparatus, and a program.

In recent years, in an X-ray diagnostic apparatus, from X-ray incidence directions including X-ray incidence directions on both sides of a predetermined X-ray incidence direction (hereinafter referred to as a reference X-ray irradiation axis), a plurality of imaging target regions of a subject are scanned. Tomosynthesis imaging is used in which a plurality of projection data with different X-ray incident directions are acquired by irradiating X-rays with a hologram. Also, recently, tomosynthesis imaging that secures a long field of view (hereinafter referred to as long tomosynthesis) imaging, which is a combination of tomosynthesis imaging and long imaging, has come into use. According to long tomosynthesis imaging, a long cross-sectional image (hereinafter referred to as a long tomosynthesis image) perpendicular to the reference X-ray irradiation axis can be obtained. Also, long tomosynthesis images can be generated orthogonal to any position within a predetermined range of heights along the reference x-ray exposure axis. According to this elongated tomosynthesis image, the user can easily confirm cross-sectional images over a wide range of the subject with a single image.

However, since the long tomosynthesis image is a cross-sectional image generated so as to be orthogonal to the position along the reference X-ray irradiation axis, one long tomosynthesis image includes the direction along the reference X-ray irradiation axis. cannot include a site located at a height different from that of the image. For example, it is known that the spine of a subject is located along the body axis of the subject while meandering in the anterior-posterior direction of the subject from the cervical vertebrae to the caudal vertebrae. Therefore, when long tomosynthesis imaging is performed with the reference X-ray irradiation axis aligned with the front-back direction of the subject and the body axis direction of the subject being set as the longitudinal direction, the reference X-ray irradiation axis (that is, the front-back direction of the subject) ), it is not possible to include all of the vertebral bodies that make up the spine in one long tomosynthesis image orthogonal to that location.

U.S. Patent Application Publication No. 2017/0076043

The problem to be solved by the present invention is to support the generation of a single image containing a plurality of regions with mutually different positions in the direction of the reference X-ray irradiation axis based on long-dimension tomosynthesis imaging in a predetermined reference X-ray irradiation axis. It is to be.

A medical image processing apparatus according to an embodiment includes an acquisition unit and a processing unit. The acquisition unit obtains a plurality of images with the direction of the reference X-ray irradiation axis as the slice direction, which is obtained based on the longitudinal tomosynthesis imaging of the subject performed with the direction perpendicular to the reference X-ray irradiation axis as the longitudinal direction. Acquire a slice image. Based on a plurality of slice images, the processing unit divides a plurality of divided ranges along the longitudinal direction into a plurality of divided ranges, and extracts characteristic regions of the subject included in the divided ranges perpendicular to the reference X-ray irradiation axis. Specific information is generated by inputting a plurality of slice images to a trained model that generates specific information indicating a position in a direction.

1 is a block diagram showing a configuration example of a medical image processing apparatus according to a first embodiment; FIG. The figure for demonstrating long tomosynthesis imaging|photography. FIG. 3 is an explanatory diagram showing an example of a plurality of slice images generated based on the long tomosynthesis imaging shown in FIG. 2; FIG. 4 is a diagram for explaining a divided range divided into a plurality of parts along the longitudinal direction; Explanatory drawing which shows an example of specific information. Explanatory drawing which shows an example of the synthetic image produced|generated based on specific information. Explanatory diagram showing an example of data flow during learning of a specific function. Explanatory diagram showing an example of a data flow when a specific function is operated. 1 is a block diagram showing a configuration example of an X-ray diagnostic apparatus including a medical image processing apparatus according to a second embodiment; FIG.

Hereinafter, embodiments of a medical image processing apparatus, an X-ray diagnostic apparatus, and a program will be described in detail with reference to the drawings.

(First embodiment)
FIG. 1 is a block diagram showing a configuration example of a medical image processing apparatus 10 according to the first embodiment. The medical image processing apparatus 10 has at least a memory circuit 13 and a processing circuit 15 . FIG. 1 shows an example in which the medical image processing apparatus 10 further includes an input interface 11, a display 12, and a network connection circuit .

The input interface 11 is composed of general input devices such as a trackball, switch buttons, mouse, keyboard, numeric keypad, etc., and outputs operation input signals corresponding to user's operations to the processing circuit 15 . The display 12 is composed of a general display output device such as a liquid crystal display or an OLED (Organic Light Emitting Diode) display.

The storage circuit 13 has a configuration including a recording medium readable by a processor such as a RAM (Random Access Memory), a semiconductor memory device such as a flash memory, a hard disk, an optical disk, etc., and stores a program used by the processing circuit 15. , parameter data and other data. Some or all of the programs and data in the recording medium of the storage circuit 13 may be downloaded by communication via the network 100, or provided to the storage circuit 13 via a portable storage medium such as an optical disc. may

The storage circuit 13 also stores projection data based on long tomosynthesis imaging of the subject P acquired via the network 100 and a plurality of long tomosynthesis images (hereinafter referred to as slice images) reconstructed based on the projection data. ) may be stored.

The network connection circuit 14 implements various information communication protocols according to the form of the network 100 . The network connection circuit 14 connects with other electric devices via the network 100 according to these various protocols. The network 100 means a general information communication network using telecommunication technology, and includes a wireless/wired LAN such as a hospital backbone LAN (Local Area Network), an Internet network, a telephone communication network, an optical fiber communication network, and a cable communication network. Including networks and satellite communication networks.

The medical image processing apparatus 10 is connected to the X-ray diagnostic apparatus 101 and the image server 102 via the network 100 so as to be able to transmit and receive data to each other. The X-ray diagnostic apparatus 101 includes a simple X-ray imaging apparatus, a breast X-ray imaging apparatus (mammography apparatus), an X-ray TV apparatus, an X-ray angiography apparatus, and the like.

The processing circuit 15 implements a function of centrally controlling the medical image processing apparatus 10 . In addition, the processing circuit 15 reads out and executes the program stored in the storage circuit 13, so that the positions in the direction of the reference X-ray irradiation axis are different from each other based on the long tomosynthesis imaging in the predetermined reference X-ray irradiation axis. It is a processor that executes processing for supporting the generation of a single image containing a plurality of parts.
Here, the long tomosynthesis image will be briefly described.

FIG. 2 is a diagram for explaining long tomosynthesis imaging. FIG. 3 is an explanatory diagram showing an example of a plurality of slice images (long tomosynthesis images) generated based on the long tomosynthesis imaging shown in FIG.

A subject is placed in a supine position on the top plate 101c of the X-ray diagnostic apparatus 101, and longitudinal tomosynthesis imaging is performed with the front-rear direction of the subject as the reference X-ray irradiation axis direction and the head-foot direction of the subject as the longitudinal direction. Consider the case (see Figure 2). In this case, long tomosynthesis imaging is performed by moving the X-ray tube 101a and the X-ray detector 101b in parallel in the longitudinal direction while maintaining the positional relationship of facing each other with the subject interposed therebetween.

In the long tomosynthesis imaging, the X-ray tube 101a and the X-ray detector 101b, which are arranged to face each other with the subject interposed therebetween, intersect with each other on both sides of the subject in FIG. It may be performed by moving from left to right along the longitudinal direction while the other moves from right to left in the opposite direction.

In addition, the X-ray diagnostic apparatus 101 can perform long tomosynthesis imaging using a direction different from the front-rear direction of the subject, such as the left-right direction (side surface) of the subject, as a reference X-ray irradiation axis. For example, if the characteristic site is the spine and it is known that the subject's spine has a large degree of scoliosis, the reference X-ray irradiation axis may be set along the lateral direction of the subject.

In the following description, an example in which slice images (long tomosynthesis images) are obtained based on long tomosynthesis imaging performed with the longitudinal direction of the subject as the reference X-ray irradiation axis direction will be described.

On the other hand, as shown in FIG. 3, a plurality of slice images SL1, 2, . . . , SLk, . The Ys direction of the XsYs plane, which is the image plane of (integer from n to n), coincides with the longitudinal direction, and the slice direction Xs coincides with the reference X-ray irradiation axis direction.

However, as shown in FIG. 2, a plurality of vertebral bodies forming the spine of the subject extend in the head-foot direction of the subject while curving in the front-rear direction of the subject. Specifically, it is known that of the spine, the cervical spine curves forward, the thoracic spine curves backward, the lumbar spine curves forward, and the sacrum and tailbone curve backward. Therefore, the positions in the slice direction Zs of the multiple vertebral bodies forming the spine are different from each other. Therefore, one slice image (for example, slice image SL3) cannot include all vertebral bodies. That is, when there are a plurality of regions with different positions in the reference X-ray irradiation axis direction, one slice image generated based on tomosynthesis imaging in the reference X-ray irradiation axis includes all of these plurality of regions. difficult.

Therefore, the processor of the processing circuit 15 specifies the position of each of the plurality of parts in the reference X-ray irradiation axis direction based on the long tomosynthesis imaging from one direction (that is, in the predetermined reference X-ray irradiation axis). This supports the generation of a single image including a plurality of parts whose positions in the reference X-ray irradiation axis direction are different from each other.

Thus, as shown in FIG. 1, the processor of the processing circuitry 15 implements an acquisition function 21, a specification function 22, and an image synthesis function 23. FIG. Each of these functions is stored in the storage circuit 13 in the form of a program.

The acquisition function 21 obtains slices in the direction of the reference X-ray irradiation axis obtained based on long tomosynthesis imaging (see FIG. 2) of the subject performed with the direction perpendicular to the reference X-ray irradiation axis as the longitudinal direction. A plurality of slice images SL1, 2, . . . , SLn (see FIG. 3) are obtained. Acquisition function 21 is an example of an acquisition unit.

Based on a plurality of slice images, the specifying function 22 calculates a reference X of a characteristic region (for example, a vertebral body) of the subject included in each divided range divided into a plurality of sections along the longitudinal direction. Specific information is generated by inputting a plurality of slice images to a trained model that generates specific information indicating a position in a direction perpendicular to the line irradiation axis. The identification information is information indicating the position of each of the plurality of characteristic regions in the reference X-ray irradiation axis direction, and is, for example, information that identifies a slice image including the characteristic region in the divided range. Information specifying a slice image includes, for example, a slice number. The specific function 22 is an example of a processing unit.

FIG. 4 is a diagram for explaining a plurality of divided ranges along the longitudinal direction. Moreover, FIG. 5 is explanatory drawing which shows an example of specific information. Also, FIG. 6 is an explanatory diagram showing an example of a composite image generated based on the specific information.

The specifying function 22 divides the slice image plane parallel to the XsYs plane into a plurality of divided ranges Di (where i is a positive integer) along the longitudinal direction, and the characteristic regions ( specific information indicating the position of the vertebral body) in the reference X-ray irradiation axis direction; As the specific information, for example, as shown in FIG. 5, a matrix of division ranges Di and slice numbers SLk having different components (1 or 0 in FIG. 5) depending on whether feature information is included can be used.

In this example, only the component of slice number 3 is 1 when looking at the slice direction of the divided range D5, for example. In this case, if a partial image corresponding to the Ys range of the division range D5 of the slice SL3 is extracted as the image of the division range D5 having the XsYs plane as the image plane, the image of the division range D5 can be reliably included in the characteristic region. can be done.

In the following description, an example in which the characteristic site is a vertebral body will be shown. However, the characteristic site, such as the femur, extends substantially linearly when viewed from one direction, while when viewed from another direction, the characteristic site extends almost linearly. Then, it is preferable to extend in a curved shape. Also, the characteristic site may be a site that extends in a curved shape and does not have a linear shape, regardless of the perspective from any direction.

The image synthesizing function 23 extracts a partial image corresponding to the divided range of the slice image containing the characteristic part of the subject for each divided range Di, arranges these extracted images in the longitudinal direction, and synthesizes them. A composite image 72 is generated and displayed on the display 12 (see FIG. 6). As a result, for example, if the feature sites are vertebral bodies, one composite image 72 can include all of the plurality of vertebral bodies. At least a portion of each of the plurality of vertebral bodies included in the composite image 72 is rendered, and preferably all of them are rendered. The image synthesizing function 23 is an example of an image synthesizing unit.

Note that if the vertebral body extends over a plurality of slice images within the division range, the image synthesizing function 23 selects these slice images or a combination of these slice images as partial images corresponding to the division range. A simple average of the slice images before and after that may be used, or a weighted average may be used.

In addition, even if the vertebral body is included in only one slice image within the divided range, the image synthesizing function 23 performs A partial image may be smoothed by performing simple averaging or weighted averaging of slice images before and after the slice image. In addition to the weighted average method, for example, the minimum intensity projection method used for image processing in X-ray computed tomography or the like may be used for smoothing the tomographic image.
Next, the operation of the specific function 22 will be explained.

The processing of the specific function 22 is preferably performed using machine learning. In the following description, the identification function 22 includes a neural network 31 and uses deep learning to obtain multiple slice images based on long tomosynthesis imaging in a predetermined reference X-ray irradiation axis (i.e. from one direction). An example of generating specific information is shown. In this case, the specific function 22 sequentially updates the parameter data 32 by learning with a large number of input training data.
FIG. 7 is an explanatory diagram showing an example of data flow during learning of the specific function 22. As shown in FIG.

The specific function 22 successively updates the parameter data 32 by receiving a large number of training data and performing learning. The training data are a plurality of slice images 411, a plurality of slice images 412, a plurality of slice images 413, . and ideal specific information 511, 512, 513, . . .

The specific function 22 uses the neural network 31 to process a plurality of slice images 411, a plurality of slice images 412, a plurality of slice images 413, . So-called learning is performed by updating the parameter data 32 so as to approach 511, 512, 513, . In general, when the change rate of the parameter data 32 converges within a threshold value, it is determined that learning has ended. Hereinafter, the parameter data 32 after learning is particularly referred to as learned parameter data 32a. Note that the type of training input data and the type of input data during operation shown in FIG. 8 should match.

Here, a method for generating the ideal specific information 511 among the ideal specific information 511, 512, 513, . . . will be described. Since the method of generating the other ideal specific information 512, 513, .

The ideal specific information 511 is obtained by longitudinal imaging or Generated by long-form tomosynthesis imaging. Hereinafter, imaging from a direction different from the reference X-ray irradiation axis of long tomosynthesis imaging performed to obtain a plurality of slice images 411 will be referred to as reference imaging, and an image based on the reference imaging will be referred to as a reference image. The direction of the reference imaging with respect to the subject is, for example, a direction different from the direction of the long tomosynthesis imaging performed to obtain the plurality of slice images 411 by approximately 90 degrees.

In the example shown in FIG. 2, the subject may be placed on the top plate 101c with the side facing the subject for reference imaging. Reference imaging may be performed by rotating 101b by approximately 90 degrees around the body axis of the subject. The range of “approximately 90 degrees” is not limited to a close range (range of 85 to 94 degrees) that is rounded off, and may be a range in which the two directions are significantly different. For example, it may be in the range of 70 to 110 degrees. The reference radiography may be normal long radiography or long tomosynthesis radiography.

When reference imaging is performed from a direction different from the reference X-ray irradiation axis shown in FIG. 2 by approximately 90 degrees, the curvature of the spine in the front-rear direction of the subject can be easily confirmed (see FIG. 4). The ideal specific information 511 can be generated by manually specifying the position of the vertebral bodies by the user who has checked the reference image.

When reference imaging is performed in a direction different from that of long tomosynthesis imaging by approximately 90 degrees, the direction perpendicular to the longitudinal direction of the image plane of the reference image coincides with the slice direction Zs of long tomosynthesis imaging. Therefore, it is possible to easily acquire the position in the slice direction Zs corresponding to the designated position on the reference image and generate the specific information 511 . If it cannot be said that the imaging directions of the reference imaging and the longitudinal tomosynthesis imaging differ by approximately 90 degrees, a trigonometric function based on the angle between the two imaging directions with respect to the subject can be used to obtain the designated image on the reference image. From the position, the position in the slice direction Zs corresponding to this position can be specified.

Further, by increasing the number of designations for a portion having a large curvature among the arrays of a plurality of vertebral bodies, the accuracy of interpolation between designated positions increases. Note that the characteristic site may be specified by a line instead of a point. Designation by line is, for example, a method of designating the position of the vertebral body by drawing a line through an input I/F circuit such as a tablet terminal type. Furthermore, the specification of features is not limited to being performed by points or lines, but when the feature of interest is fixed, such as the spine or femur, for example, image using standard models or databases. It may be performed by recognition or the like.
FIG. 8 is an explanatory diagram showing an example of a data flow during operation of the specific function 22. As shown in FIG.

During operation, the specifying function 22 receives a plurality of slice images 61 based on long tomosynthesis imaging on a predetermined reference X-ray irradiation axis, and outputs specifying information 71 using the learned parameter data 32a.

The neural network 31 and the learned parameter data 32a constitute the learned model 30. FIG. The neural network 31 is stored in the memory circuit 13 in the form of a program. The learned parameter data 32 a may be stored in the storage circuit 13 or in a storage medium connected to the medical image processing apparatus 10 via the network 100 . In this case, the specific function 22 as an example of the processing unit realized by the processor of the processing circuit 15 reads out the learned model from the storage circuit 13 and executes it to perform long tomosynthesis on a predetermined reference X-ray irradiation axis. Specific information is generated based on a plurality of slice images based on photography. Note that the learned model may be constructed by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

Based on a plurality of slice images 61 based on long tomosynthesis imaging in a predetermined reference X-ray irradiation axis, the medical image processing apparatus 10 according to the present embodiment performs the measurement of each of the reference X-ray irradiation axis directions of a plurality of characteristic regions. Specific information 71 that indicates the location can be generated. For this reason, for example, when the characteristic part is a vertebral body, long tomosynthesis imaging is performed with the body axis direction of the subject aligned with the reference X-ray irradiation axis in the longitudinal direction of the subject. Also, based on the specific information, a composite image 72 including all of the plurality of vertebral bodies can be generated.

Therefore, for example, when inspecting the lordosis or scoliosis of the spine, only long tomosynthesis photography from one direction is performed without X-ray photography from multiple directions, and the composite image 72 including all of the plurality of vertebral bodies is obtained. can be displayed. Therefore, compared with the case of performing X-ray imaging from a plurality of directions, it is possible to significantly reduce the exposure of the subject and shorten the examination time significantly.

In addition, even curved or bent bones can be clearly depicted in one composite image 72 . Therefore, scoliosis measurement and posterior curvature measurement by the Cobb method can be facilitated, and diagnosis of bone curvature can be supported. Supplementally, interpretation and measurement can be facilitated by a composite image showing high contrast compared to normal long-length imaging (single image). In contrast to the usual long tomosynthesis imaging in which it is necessary to check a plurality of slice images for image interpretation and measurement, the composite image 72 allows the user to avoid troublesome operations such as changing slice images. can be released.

(Second embodiment)
FIG. 9 is a block diagram showing a configuration example of an X-ray diagnostic apparatus 80 including a medical image processing apparatus according to the second embodiment.

The X-ray diagnostic apparatus 80 includes an imaging apparatus 81 that performs long tomosynthesis imaging on a subject along a predetermined reference X-ray irradiation axis, and a console apparatus 82 as an example of the medical image processing apparatus 10 . The X-ray diagnostic apparatus 80 shown in the second embodiment is the first embodiment in that it can use a plurality of slice images 61 generated by long tomosynthesis imaging of a subject on a predetermined reference X-ray irradiation axis. It is different from the medical image processing apparatus 10 shown in the form. Other configurations and actions are substantially the same as those of the medical image processing apparatus 10 shown in FIG. Also, since the processing related to the generation of the trained model 30 is the same as that of the medical image processing apparatus 10 shown in FIG. 1, the explanation at the time of learning is omitted.

The imaging device 81 is composed of, for example, an imaging system of an X-ray TV device, and includes an X-ray tube 101a for long tomosynthesis imaging of a subject placed on a tabletop 101c along a predetermined reference X-ray irradiation axis, an X-ray It has an imaging system such as the detector 101b, and provides the console device 82 with projection data of the subject obtained by imaging. The imaging device 81 is an example of an imaging unit.

The reconstruction function 20 of the processing circuit 15x of the console device 82 as an example of the medical image processing apparatus 10 reconstructs projection data of the subject to generate a plurality of slice images 61 of the subject.

Acquisition function 21 x of processing circuit 15 x acquires a plurality of slice images 61 from reconstruction function 20 . The specifying function 22 x outputs specifying information 71 based on a plurality of slice images 61 of the subject using the trained model 30 . The image synthesizing function 23 x displays the synthetic image 72 on the display 12 of the console device 82 .

Similarly to the medical image processing apparatus 10 according to the first embodiment, the X-ray diagnostic apparatus 80 according to the second embodiment also generates a plurality of slice images 61 based on long tomosynthesis imaging on a predetermined reference X-ray irradiation axis. Based on this, it is possible to generate specific information 71 indicating the position of each of the plurality of characteristic regions in the reference X-ray irradiation axis direction.

According to at least one embodiment described above, a single image including a plurality of parts having different positions in the direction of the reference X-ray irradiation axis is generated based on long tomosynthesis imaging in a predetermined reference X-ray irradiation axis. can support

In the above embodiment, the word "processor" is, for example, a dedicated or general-purpose CPU (Central Processing Unit), a GPU (Graphics Processing Unit), or an application specific integrated circuit (ASIC), Circuits such as programmable logic devices (eg, Simple Programmable Logic Devices (SPLDs), Complex Programmable Logic Devices (CPLDs), and FPGAs) shall be meant. The processor implements various functions by reading and executing programs stored in the storage medium.

Further, in the above embodiments, an example of a case where a single processor of the processing circuit realizes each function is shown, but a processing circuit is configured by combining a plurality of independent processors, and each processor realizes each function. good too. Further, when a plurality of processors are provided, a storage medium for storing programs may be provided individually for each processor, or a single storage medium may collectively store programs corresponding to the functions of all processors. good too.

It should be noted that although several embodiments of the invention have been described, these embodiments are provided by way of example and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, replacements, and modifications can be made without departing from the scope of the invention. These embodiments and their modifications are included in the scope and spirit of the invention, as well as the scope of the invention described in the claims and equivalents thereof.

10 Medical image processing apparatus 21, 21x Acquisition function 22, 22x Specific function 23, 23x Image composition function 30 Trained model 61 Slice image 71 Specific information 72 Synthetic image 80 X-ray diagnostic device 81 Imaging device 82 Console device

Claims (7)

A plurality of slice images with the direction of the reference X-ray irradiation axis as the slice direction, obtained based on the long tomosynthesis imaging of the subject performed with the direction orthogonal to the reference X-ray irradiation axis as the long direction. an acquisition unit that acquires
Based on the plurality of slice images, for each divided range divided into a plurality along the longitudinal direction, the characteristic part of the subject included in the divided range is orthogonal to the reference X-ray irradiation axis. A processing unit that generates the specific information by inputting the plurality of slice images having the direction of the reference X-ray irradiation axis as the slice direction to a trained model that generates the specific information indicating the position in the direction. and,
A medical image processing device with The trained model is
Based on the plurality of slice images, for each divided range, generating the specific information that identifies a slice image that includes the characteristic part of the subject in the divided range.
The medical image processing apparatus according to claim 1. An image for generating a composite image by extracting an image corresponding to the divided range of the slice image containing the characteristic part of the subject based on the specific information and synthesizing the extracted images for each divided range. synthesis department,
The medical image processing apparatus according to claim 1 or 2, further comprising: the longitudinal direction is the head-to-foot direction of the subject;
The characteristic site is the vertebral body of the subject,
The image synthesizing unit
at least a portion of each of a plurality of vertebral bodies contained in the subject is visualized based on the specific information, and the composite image is formed such that all of the plurality of vertebral bodies are included in the single composite image; to generate
The medical image processing apparatus according to claim 3. The direction of the reference X-ray irradiation axis is
front-back direction or left-right direction of the subject,
The medical image processing apparatus according to any one of claims 1 to 4. An imaging unit that generates a plurality of slice images with the direction of the reference X-ray irradiation axis as the slice direction by executing long tomosynthesis imaging of the subject with the direction orthogonal to the reference X-ray irradiation axis as the longitudinal direction. and,
an acquisition unit that acquires the plurality of slice images;
Based on the plurality of slice images, for each divided range divided into a plurality along the longitudinal direction, the characteristic part of the subject included in the divided range is orthogonal to the reference X-ray irradiation axis. A processing unit that generates the specific information by inputting the plurality of slice images to a trained model that generates the specific information indicating the position in the direction;
An X-ray diagnostic device with to the computer,
A plurality of slice images with the direction of the reference X-ray irradiation axis as the slice direction, obtained based on the long tomosynthesis imaging of the subject performed with the direction orthogonal to the reference X-ray irradiation axis as the long direction. a step of obtaining;
Based on the plurality of slice images, for each divided range divided into a plurality along the longitudinal direction, the characteristic part of the subject included in the divided range is orthogonal to the reference X-ray irradiation axis. generating the specific information by inputting the plurality of slice images having the direction of the reference X-ray irradiation axis as the slice direction to a trained model that generates the specific information indicating the position in the direction; ,
program to run the

Images