Imaging support device, imaging support method, and imaging support program

Abstract

To improve reproducibility of an imaged cross section.SOLUTION: An imaging support device in the present embodiment includes an acquisition part, a calculation part, and a determination part. The acquisition part acquires a first image and a second image older than the first image. The calculation part calculates a correction value for reproducing the second image. The determination part determines reproducibility of the second image on the basis of the correction value.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The embodiments disclosed in this specification and the drawings relate to an imaging support device, an imaging support method, and an imaging support program. [Background technology]

[0002] In a typical diagnostic imaging test, the procedure is assumed to be as follows: after the subject is placed on the bed and the imaging target area is fixed, a protocol is selected, the imaging range is set, and imaging is started. However, depending on the condition of the subject, it is not always possible to set the imaging range appropriately according to the disease and case, so there is a possibility that the imaging cross section desired by the medical staff cannot be reproduced. In addition, in follow-up observations of the same subject, if there is variation in the settings made by operators such as technicians, the quality and speed of image diagnosis is expected to decrease. Therefore, imaging with reproducible settings is required. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Publication No. 2021-40749 [Non-patent literature]

[0004] [Non-Patent Document 1] Hidenori Takeshima, "Deep Learning and Its Application to Function Approximation for MR in Medicine: An Overview", [Online], September 17, 2021, Magnetic Resonance in Medical Sciences, [Retrieved April 24, 2023], Internet<URL:https: / / doi.org / 10.2463 / mrms.rev.2021-0040> Summary of the Invention [Problem to be solved by the invention]

[0005] One of the problems to be solved by the embodiments disclosed in this specification and the drawings is to improve the reproducibility of the imaging cross section. However, the problems to be solved by the embodiments disclosed in this specification and the drawings are not limited to the above problem. Problems corresponding to the effects of each configuration shown in the embodiments described later can also be positioned as other problems. [Means for solving the problem]

[0006] The imaging support device according to the present embodiment includes an acquisition unit, a calculation unit, and a determination unit. The acquisition unit acquires a first image and a second image that is an image older than the first image. The calculation unit calculates a correction value for reproducing the second image. The determination unit determines the reproducibility of the second image based on the correction value. [Brief description of the drawings]

[0007] [Figure 1] FIG. 1 is a block diagram showing an imaging support device according to the present embodiment. [Diagram 2] FIG. 2 is a flowchart showing an example of the operation of the imaging support device according to this embodiment. [Diagram 3] FIG. 3 is a diagram showing an example of a past image and a captured image that needs to be corrected. [Figure 4] FIG. 4 is a conceptual diagram showing an example of the conversion parameters according to the present embodiment. [Diagram 5] FIG. 5 is a diagram illustrating a first application example of the trained model according to this embodiment. [Figure 6] FIG. 6 is a diagram illustrating a second application example of the trained model according to this embodiment. [Figure 7] FIG. 7 is a diagram illustrating a third application example of the trained model according to this embodiment. [Figure 8] FIG. 8 is a diagram showing an example of a display screen relating to the correction value according to the present embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] Hereinafter, an imaging support device, an imaging support method, and an imaging support program according to the present embodiment will be described with reference to the drawings. In the following embodiments, parts with the same reference numerals perform the same operations, and duplicated descriptions will be omitted as appropriate. Hereinafter, one embodiment will be described with reference to the drawings.

[0009] The imaging support device according to this embodiment will be described with reference to the block diagram of Fig. 1. The imaging support device may be included in a console of a medical image diagnostic device (not shown), or may be included in a computer independent of the medical image diagnostic device and connected to the medical image diagnostic device so as to be able to communicate with it. The imaging support device 1 according to this embodiment includes a processing circuit 10, a memory 11, an input interface 12, a communication interface 13, and a display 14.

[0010] The processing circuit 10 includes an acquisition function 101, a calculation function 102, a determination function 103, an instruction function 104, a model execution function 105, and a display control function 106. The processing circuit 10 has a processor (not shown) as a hardware resource.

[0011] The acquisition function 101 acquires a first image and a second image that is an image older than the first image. The first image and the second image are medical images, and are assumed to be pairs of images acquired by the same medical image diagnostic device, such as a pair of MR images acquired by an MRI (Magnetic Resonance Imaging) device, a pair of CT images acquired by an X-ray CT (Computed Tomography) device, a pair of X-ray images acquired by an X-ray diagnostic device, a pair of ultrasound images acquired by an ultrasound diagnostic device, a pair of PET images acquired by a PET (Positron Emission Tomography) device, or a pair of SPECT images acquired by a SPECT (Single Photon Emission Computed Tomography) device, but are not limited thereto. The first image and the second image may be acquired, for example, from a medical image diagnostic device, a PACS server, or various image processing servers. For example, the first image may be a CT image and the second image may be an MR image. That is, the first image and the second image may be a combination of medical images acquired by different medical image diagnostic devices, such as MR images, CT images, ultrasound images, PET images, and SPECT images. Note that the second image does not have to be a medical image captured by the same medical image diagnostic device, and an image captured close to the first image may be selected.

[0012] The calculation function 102 calculates a correction value for reproducing the second image. The correction value is, for example, a movement amount and a rotation angle of the imaging target part depicted in the first image. The correction value may also be a parameter for extracting and reproducing the second image from the three-dimensional model.

[0013] The determination function 103 determines the repeatability of the second image based on the correction value. The repeatability is a criterion for whether or not the setting for imaging based on the correction value is possible for the first image with respect to image registration or image comparison with the second image.

[0014] The instruction function 104 generates an instruction to perform imaging or to correct the settings when the setting based on the correction value is possible for the subject. Alternatively, when the setting based on the correction value is not possible for the imaging target part of the subject, the instruction function 104 generates a message to the effect that the past image cannot be reproduced, or the like. Also, the instruction function 104 issues an instruction to correct the image to be closer to the past image.

[0015] The model execution function 105 executes the trained model and estimates the imaging cross-sectional image, the conversion parameter, or the correction value. The conversion parameter is a parameter when converting the image. In this embodiment, the conversion parameter is assumed to be a parameter related to the moving direction and the rotating direction, but is not limited thereto. Parameters such as a magnification ratio and a reduction ratio may be prepared, and any parameter related to image conversion may be adopted. The trained model will be described later with reference to FIG. 5 to FIG. 7.

[0016] The display control function 106 displays on the display 14, for example, the results of the execution of the trained model by the model execution function 105 (e.g., cross-sectional images, conversion parameters, and correction values), correction instructions by the instruction function 104, information that the past image cannot be reproduced, correction instructions to approximate the past image, and error indications.

[0017] The memory 11 stores various medical data, acquisition conditions, trained models, etc., which will be described later. The memory 11 is a semiconductor memory element such as a random access memory (RAM), a flash memory, a hard disk drive (HDD), a solid state drive (SSD), an optical disk, etc. The memory 11 may also be a drive device that reads and writes various information to and from a portable storage medium such as a CD-ROM drive, a DVD drive, or a flash memory.

[0018] The input interface 12 has a circuit for receiving various instructions and information input from a user. The input interface 12 has a circuit for an input device such as a pointing device like a mouse or a keyboard. The circuits of the input interface 12 are not limited to circuits for physical operation parts like a mouse and a keyboard. For example, the input interface 12 may have an electrical signal processing circuit for receiving an electrical signal corresponding to an input operation from an external input device provided separately from the imaging support device 1 and outputting the received electrical signal to various circuits in the imaging support device 1.

[0019] The communication interface 13 exchanges data with an external device via a wired or wireless connection. Since any common communication method can be used for the communication method and interface structure, a description thereof will be omitted here.

[0020] The display 14 displays a GUI (Graphical User Interface) for accepting various operations from the user. Any display can be used as appropriate, such as a CRT (Cathode Ray Tube) display, a liquid crystal display, an organic EL display, an LED display, a plasma display, or a touch display that allows touch input. Note that the imaging support device 1 may not include a display, and the GUI may be displayed on an external display, or the GUI may be displayed via a projector or the like.

[0021] Next, an example of the operation of the imaging support device 1 according to this embodiment will be described with reference to the flowchart of FIG.

[0022] In step SA1, the processing circuit 10, using the acquisition function 101, acquires a previously captured image (second image).

[0023] In step SA2, the processing circuit 10 acquires a captured image (first image) of the same object as the previous image by the acquisition function 101. The captured image acquired in step SA2 may be a positioning image or may be an captured image collected by a main scan.

[0024] In step SA3, the processing circuit 10 uses the calculation function 102 to calculate a transformation parameter so that the similarity between the captured image and the previous image is maximized by the transformation process on the captured image. Specifically, for example, the transformation parameter may be estimated when the captured image is linearly or nonlinearly transformed so that the captured image matches the previous image. For example, when an arbitrary point Xa in the captured image is transformed to a point Xr in the previous image, the transformation parameter is calculated based on the formula T(Xa)=Xr. Here, T() is the transformation parameter.

[0025] In step SA4, the processing circuit 10, by means of a calculation function 102, calculates the correction value on the basis of the transformation parameters.

[0026] In step SA5, the processing circuit 10 judges whether the imaging section can be reproduced by the judgment function 103. Specifically, it judges whether the correction value is equal to or less than a threshold. It is assumed that the threshold varies depending on the imaging target part. For example, the bending angle of the neck and the bending angle of the knee have different ranges of motion, so a threshold is set according to each imaging target part. When comparing the correction value with the threshold, it is assumed that the absolute value of the correction value is compared with the threshold, taking into account the positive and negative of the conversion parameter, but it is also possible to set a positive threshold and a negative threshold and compare the correction value with these two thresholds. That is, if the correction value is negative, it is compared with the negative threshold, and if the correction value is positive, it is compared with the positive threshold. If the correction value is equal to or less than the threshold, it is determined that the imaging section of the previous image can be reproduced, and the process proceeds to step SA6. On the other hand, if the correction value is greater than the threshold, it is determined that correction is not possible even if the settings are adjusted, and the imaging section of the previous image cannot be reproduced, and the process proceeds to step SA7.

[0027] In step SA6, the processing circuit 10 judges whether or not the settings need to be modified by the judgment function 103. Specifically, if the correction value is zero, it can be judged that the settings do not need to be modified, and imaging can be performed without changing the current settings. On the other hand, if the correction value is greater than zero, it can be judged that the settings need to be modified. Note that if the correction value is not zero but is a value that does not require modification of the settings, it may be judged that modification is not required even if the correction value is greater than zero. If the settings need to be modified, proceed to step SA8, and if the settings do not need to be modified, proceed to step SA7.

[0028] In step SA7, since the imaging section of the past image can be reproduced using the current settings, for example, the processing circuit 10 generates an imaging instruction to perform imaging using the settings using the instruction function 104, and imaging is performed by the medical image diagnostic device.

[0029] In step SA8, the processing circuit 10 outputs an instruction to correct the settings through the instruction function 104 and the display control function 106. If the settings are corrected, the reproducibility of the imaging cross-section of the past image is improved, and the process proceeds to step SA7, where imaging is performed. Regarding the correction of the settings, the processing circuit 10 displays the correction value on the display 14 or the like by the display control function 106, thereby presenting the correction value to the operator. After that, the subject himself may correct the settings by moving or bending his body, either directly or by speaking to the subject. Alternatively, for example, the processing circuit 10 may generate a control instruction for a control unit that drives a device such as a bed of the medical image diagnostic device by the instruction function 104. The bed of the medical image diagnostic device or the like may be controlled based on the instruction, and the device or the bed may be rotated or moved based on the correction value, thereby automatically correcting the settings.

[0030] In step SA9, the processing circuit 10 generates a message to the effect that the past image cannot be reproduced because the past image cannot be reproduced even if the settings are corrected by the instruction function 104, and displays the message to the user via the display control function 106. Note that, even if the processing circuit 10 cannot reproduce the past image, the instruction function 104 may generate a message including an instruction for modification closest to the past image and an error between the past image and the resultant image when modified based on the instruction for modification, so as to bring the reproduction of the past image as close as possible, and present the message to the user via the display control function 106. Thereafter, the process proceeds to step SA7, and imaging can be performed.

[0031] Next, an example of a past image and a captured image that needs to be corrected will be described with reference to FIG. FIG. 3 shows medical images of the head as a target region, and shows a previous image 31 and a captured image 33 of the same patient. In the example of FIG. 3, the captured image 33 is an image rotated clockwise compared to the previous image 31. Therefore, in order to match the positions of the previous image 31 and the captured image 33, it is necessary to rotate the captured image 33 in the direction of the arrow in FIG. 3, that is, counterclockwise. In this way, values ​​such as the amount of movement and the amount of rotation required to match the positions of the previous image 31 and the captured image 33 are calculated as correction values.

[0032] Next, an example of the conversion parameters according to this embodiment will be described with reference to the conceptual diagram of FIG. 4 shows an example of transformation parameters related to the movement direction and rotation direction, based on a bed 41 inserted into the medical image diagnostic apparatus 40. Here, the movement direction is defined in three directions, with the width direction of the bed 41 being the X direction, the length direction of the bed 41 (the direction in which the medical image diagnostic apparatus 40 is inserted into and removed from the gantry) being the Y direction, and the height direction perpendicular to the bed 41 being the Z direction. Furthermore, the rotation direction is defined in three directions, with the α direction around the X axis, the β direction around the Y axis, and the γ direction around the Z axis.

[0033] For example, the processing circuit 10 calculates the transformation parameters by the calculation function 102 so as to maximize the similarity between the captured image and the past image. The similarity may be calculated using the mean square error, the mutual information, the sum of squared differences, the sum of absolute differences, the normalized cross-correlation, the image uniformity ratio, etc. In the case of linear transformation, the above-mentioned rotation amounts α, β, and γ, and the movement amount in the X direction are t x , the amount of movement in the y direction is t y , the amount of movement in the z direction is t z Then, the correction value T can be calculated using a conversion formula such as formula (1). Note that formula (1) is only an example, and any method may be adopted as long as it can convert the captured image so as to maximize the similarity between the captured image and the past image.

[0034]

number

[0035] In addition, as a method of determining the reference position for conversion (alignment) between the past image and the captured image, for example, when the medical image diagnostic device is an MRI device or an X-ray CT device, the past image and the captured image may be captured with a marker containing a contrast agent attached to the surface of the subject.When the medical image diagnostic device is a PET device or a SPECT device, the past image and the captured image may be captured with a marker containing a radioactive substance attached to the surface of the subject.

[0036] This allows a linear or nonlinear transformation to be performed based on the reference position, making it easier to calculate the correction value. Note that the linear or nonlinear transformation may be performed based on anatomical landmarks such as bones and organs that are common between the past image and the captured image, instead of attaching a marker. In other words, any standard may be adopted as long as the method allows calculation of the correction value.

[0037] Next, a first application example of the trained model according to this embodiment will be described with reference to FIG. In step SA3 shown in FIG. 2, the processing circuit 10 may estimate the transformation parameters 53 by applying the trained model 50 to the past image 51 and the captured image 52 by the model execution function 105 as shown in FIG. 5. In this case, the trained model 50 may be generated by training a machine learning model by supervised learning using training data in which a past captured image (also called a second captured image) and a current captured image (also called a first captured image) related to the same imaging target part are used as input data, and transformation parameters calculated so that the similarity between the first captured image and the second captured image is maximized are used as correct answer data. As the machine learning model, for example, a neural network such as DCNN (Deep Convolutional Neural Network) or BERT (Bidirectional Encoder Representations from Transformers) may be used. A general training method may be used for the training method of the machine learning model, and a detailed description thereof will be omitted.

[0038] In addition, the trained model 50 may be trained to output the correction value 54 directly, not limited to the conversion parameter 53. That is, the trained model 60 may be generated by training a machine learning model using training data in which the past image 51 and the captured image 52 are used as input data and the correction value obtained based on the conversion parameter based on the past image 51 and the captured image 52 is used as correct answer data. In this way, the processing circuit 10 can estimate the correction value 54 of the captured image 52 with respect to the past image 51 by applying the trained model 50 to the past image 51 and the captured image 52 by the model execution function 105. Therefore, the processes of steps SA3 and SA4 shown in FIG. 2 can be executed at once by the trained model 60.

[0039] Next, a second application example of the trained model according to this embodiment will be described with reference to FIG. In the above embodiment, it is assumed that the correction value is calculated based on the positioning image or the actual captured image by the main scan for comparison with the past image, but this is not limiting, and the captured image may be estimated based on the camera image. The camera image may be acquired by capturing an image using a camera device installed on the ceiling of the examination room or on the top of the gantry of the medical image diagnostic device. Note that the user may capture the camera image by holding a device equipped with a camera mechanism such as a tablet terminal, or by fixing the device to a tripod or the like. In other words, any capturing means may be used as long as it is capable of capturing a camera image.

[0040] As shown in FIG. 6, the processing circuit 10 applies the trained model 60 to the camera image 61 by the model execution function 105, thereby generating an imaging section image 62. The trained model 60 may be generated by training a machine learning model by supervised learning, for example, using a camera image captured when the setting of the imaging target part of the subject is completed as input data and a trained section image of the imaging target part captured by the setting as correct answer data. In addition to the above-mentioned neural network, the machine learning model may use a generative neural network such as GAN (Generative Adversarial Networks), cGAN (conditional GAN), or cVAE (conditional Variable Auto-Encoder). When GAN or the like is used, a trained model trained by so-called semi-supervised learning or unsupervised learning may be generated.

[0041] In other words, the trained model has learned what kind of cross-sectional image can be obtained based on the part of the subject depicted in the camera image 61 and the position and angle of that part, and therefore can estimate the cross-sectional image required for image alignment with past images.

[0042] As a result, in step SA2 shown in Fig. 2, instead of actually capturing an image of the same object as the past image, a camera image is captured and the learned model is applied to the camera image to obtain a captured image. Therefore, for example, the past image can be reproduced in a setting in which the subject is placed on a bed, so that the process of actually capturing an image can be omitted, and efficiency can be improved. In addition, if the medical image is an X-ray image, CT image, PET image, or SPECT image, radiation exposure to the subject can be reduced.

[0043] Next, a third application example of the trained model according to this embodiment will be described with reference to FIG. As shown in FIG. 7, a past image 51 of the same subject, a camera image at the setting when the past image 51 was captured (hereinafter also referred to as a past camera image 71 or a fourth camera image), and a camera image at the current setting (hereinafter also referred to as a current camera image 72 or a third camera image) are input to a trained model 70 to generate an imaging section image 73. The trained model 70 for estimating the imaging section image 73 is generated by using the past image, a camera image (also referred to as a second camera image) related to the setting before the past image was captured, and another camera image (also referred to as a first camera image) different from the second camera image of the same imaging target part as input data, and training a machine learning model using a medical image showing an imaging section captured at a setting captured with the other camera image as correct answer data. It is assumed that, compared to the trained model 50 shown in FIG. 5, an imaging section image that better reflects the correspondence between the past image and the past camera image is generated by further using the past image and the past camera image for training.

[0044] 5, the processing circuit 10 may estimate the conversion parameter 53 or the correction value 54 by using the past image 51, the past camera image 71, and the current camera image 72 by the model execution function 105. Specifically, the past camera image captured when the setting of the imaging target part of the subject is completed, the past image captured with the setting, and the camera image related to the setting when the other medical image is captured are used as input data, and the conversion parameters related to the other medical image and the past image are used as correct answer data to train the machine learning model by supervised learning, thereby generating the learned model 70. This makes it possible to estimate the conversion parameter 53 indicating the difference between the medical image captured with the setting of the current camera image 72 and the past image 51.

[0045] Similarly, a past camera image captured when the setting of the subject's imaging target part is completed, a past image captured with that setting, and a camera image related to the setting when another medical image is captured are used as input data, and a machine learning model is trained with learning data in which correction values ​​based on conversion parameters related to the other medical images and the past images are used as correct answer data, to obtain a learned model 70. This makes it possible to estimate a correction value 54 for a medical image when captured with the setting of a current camera image 72 in order to reproduce a past image 51.

[0046] Next, an example of a display screen relating to correction values ​​will be described with reference to FIG. 8 is an example of a user interface of correction values ​​presented to a user. A display screen 80 displays past images 81 showing three different cross sections, and a current image 82 showing three cross sections corresponding to the past images 81.

[0047] By looking at the display screen 80, the user can see from the message 83 "Settings need to be modified to ensure reproducibility" that modifications are necessary to reproduce the past image. Furthermore, from the message 84 regarding the modification values, the user can see that while the two current images 82 in the ZX plane and the YZ plane need modification of the current settings to reproduce the past image, the current image 82 in the XY plane can reproduce the past image with the current settings.

[0048] More specifically, it is clear that the current image 82 in the ZX plane requires modification of the transformation parameter β, and that the angle at which the knee is internally rotated is 40 degrees. Similarly, it is clear that the current image 82 in the YZ plane requires modification of the transformation parameters α and Y, and that the angle at which the knee is flexed is 15 degrees, and that the movement toward the head is 5 cm. The user can perform imaging that ensures the reproducibility of past images by moving the subject 5 cm toward the head (insertion direction), bending the knee at an angle of 15 degrees, and internally rotating it by 40 degrees.

[0049] In addition, since the coordinate system of the image (called the image coordinate system) differs from the coordinate system for correcting the position of the subject in the examination room (called the measurement coordinate system), it is necessary to align the coordinate systems. Since the transformation parameters are calculated in the image coordinate system, for example, a transformation process to transform the transformation parameters of the image coordinate system into the measurement coordinate system is executed to calculate a correction value. In the transformation process, for example, a parameter for aligning coordinates obtained from data such as the header information of the image is multiplied by the transformation parameters to align the coordinates and calculate a correction value. However, the present invention is not limited to this, and any process may be used as long as it is possible to transform from the image coordinate system to the measurement coordinate system.

[0050] The instruction function 104 may also determine the order in which the subject is rotated or moved depending on the type and correction value of the transformation parameter. For example, in the case of correcting the setting on the YZ plane shown in Fig. 8, it is possible that a deviation also occurs in the X or Y direction due to bending of the knee, so if the transformation parameter X or Y relating to the amount of movement is corrected first, an error may occur when correcting the transformation parameter α. Therefore, the order of correction is to first correct the transformation parameter α, that is, to correct the bending angle of the knee, and then correct the movement Y.

[0051] Regarding the order of such correction, for example, a table is prepared in advance that specifies the correction order according to the part to be imaged, the type of conversion parameter that needs to be corrected, and the range of the correction value. The processing circuit 10 may determine the correction order of rotation and movement based on the conversion parameter by referring to the conversion parameter calculated by the calculation function 102 and the table by the instruction function 104. In addition, a learned model may be generated by training a machine learning model based on the relationship between the conversion parameters and correction values ​​that have already been performed and the correction order. Thereby, the correction order can be estimated by inputting the conversion parameters and correction values ​​obtained by the calculation function 102 into the learned model. Note that the machine learning model may be trained so that the correction order is also output as the output of the learned model shown in FIG. 5 or FIG. 7.

[0052] Furthermore, the display control function 106 may cause the processing circuit 10 to display the conversion parameters and modification values ​​in descending order of modification from the top to the bottom of the display screen 90. This allows the user to efficiently modify the settings.

[0053] According to the present embodiment described above, a current captured image or camera image and a past image captured in the past for the same subject are acquired, a correction value for reproducing the past image is calculated, and the reproducibility of whether the past image can be reproduced is determined based on the correction value. If the correction value is within a range in which the past image can be reproduced, the subject setting can be corrected based on the correction value. This improves the reproducibility of the imaging section. Therefore, for example, in follow-up observation, images with high reproducibility of the imaging section can be collected regardless of the operator, thereby improving the quality and speed of image diagnosis.

[0054] The various functions of the processing circuit 10 may be stored in the memory 11 in the form of a program executable by a computer. In this case, the processing circuit 10 can be said to be a processor that realizes the functions corresponding to each program by reading the programs corresponding to these various functions from the memory 11 and executing them. In other words, the processing circuit 10 in a state in which each program has been read out has the multiple functions and the like shown in the processing circuit 10 in FIG. 1.

[0055] 1, the various functions are realized by a single processing circuit 10, but the processing circuit 10 may be configured by combining multiple independent processors, and each processor may execute a program to realize the functions. In other words, each of the above-mentioned functions may be configured as a program, and one processing circuit may execute each program, or a dedicated independent program execution circuit may execute a specific function.

[0056] According to at least one of the embodiments described above, the reproducibility of the imaging cross section can be improved.

[0057] In addition, each function according to the embodiment can be realized by installing a program for executing the above-mentioned processes in a computer such as a workstation and expanding the program in memory. In this case, the program for causing the computer to execute the above-mentioned method can be stored in a storage medium such as a magnetic disk (such as a hard disk), an optical disk (such as a CD-ROM or DVD), or a semiconductor memory and distributed.

[0058] In addition, the term "processor" used in the above description refers to circuits such as a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), or an Application Specific Integrated Circuit (ASIC), a programmable logic device (e.g., a Simple Programmable Logic Device (SPLD), a Complex Programmable Logic Device (CPLD), and a Field Programmable Gate Array (FPGA)).

[0059] Although some embodiments have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, substitutions, modifications, and combinations of embodiments can be made without departing from the spirit of the invention. These embodiments and their modifications are included in the scope of the invention and its equivalents as described in the claims, as well as in the scope and spirit of the invention. [Explanation of symbols]

[0060] 1 Imaging support device 10 Processing circuit 11. Memory 12 Input Interface 13 Communication Interface 14 Display 31 Past Images 33 Captured images 40 Medical imaging diagnostic equipment 41 Bed 50,60,70 Trained models 51,81 Past images 52 Captured images 53 Conversion parameters 54 Modification 61 Camera Images 62,73 Cross-sectional imaging 71 Past camera images 72 Current camera image 80 display screen 82 Current Images 83,84 Messages 101 Acquisition Function 102 Calculation Function 103 Judgment function 104 Instruction function 105 Model execution function 106 Display control function

Claims

1. an acquisition unit that acquires a first image and a second image that is an image older than the first image; a calculation unit for calculating a correction value for reproducing the second image; a determination unit that determines reproducibility of the second image based on the correction value; An imaging support device comprising:

2. The imaging support device according to claim 1 , wherein the reproducibility indicates a criterion as to whether or not settings for imaging based on the correction value are possible for an imaging target part.

3. The imaging support device according to claim 2 , further comprising an instruction unit that outputs an instruction to correct the setting when a setting based on the correction value is possible for the imaging target part.

4. The imaging support device according to claim 2 , further comprising an instruction unit that outputs a message indicating that the imaging target part cannot be reproduced when the setting based on the correction value is impossible for the imaging target part.

5. The imaging support device of claim 2, further comprising a display control unit that, when setting based on the correction value is not possible for the imaging target area, presents correction instructions that most closely resemble the second image and an error between the second image and the image when corrected based on the correction instructions.

6. The imaging support device according to claim 3 , wherein the instruction unit executes at least one of presenting the correction value to an operator and issuing a control instruction to move or rotate the imaging target portion by driving a device.

7. The imaging support device according to claim 1 , wherein the calculation unit calculates a transformation parameter that maximizes a similarity between the first image and the second image through a transformation process on the first image, and calculates the correction value from the transformation parameter.

8. The imaging support device according to claim 7 , wherein the similarity is at least one of a mean square error, a mutual information amount, a sum of squared differences, a sum of absolute differences, a normalized cross-correlation, and an image uniformity ratio.

9. The imaging support device according to claim 7 , wherein the correction value is a movement amount and a rotation angle of an imaging target part depicted in the first image, the movement amount and the rotation angle being calculated from the transformation parameter.

10. The present invention further includes a model execution unit that generates a conversion parameter by inputting a first captured image of an imaging target site and a second captured image captured earlier than the first captured image to a learned model that is trained using conversion parameters calculated so as to maximize a similarity between the first captured image and the second captured image as correct answer data, by inputting the first image and the second image to the learned model, The imaging support device according to claim 1 , wherein the calculation unit calculates the correction value from the transformation parameters generated by the model execution unit.

11. 2. The imaging support device according to claim 1, further comprising a model execution unit that generates, as the first image, a cross-sectional image of the imaging target part by inputting a camera image of the imaging target part taken by a camera device into a learned model that is trained using a cross-sectional image of the imaging target part taken by a medical image diagnostic device as correct answer data.

12. a model execution unit that generates the conversion parameters by inputting a third camera image of the imaging target part taken by the camera device, the second image, and a fourth camera image corresponding to the second image to a learned model that has been trained using as input data a first camera image of the imaging target part taken by a camera device, a second camera image of the imaging target part different from the first camera image, and a previous image of the imaging target part taken by a medical image diagnostic device corresponding to the second camera image, and conversion parameters calculated so as to maximize a similarity to the previous image when an image corresponding to the first camera image is assumed, as correct answer data; The imaging support device according to claim 1 , wherein the calculation unit calculates the correction value from the transformation parameters generated by the model execution unit.

13. The imaging support device according to claim 1 , wherein the first image is a medical image acquired of the same imaging target site as the second image by using a medical image diagnostic device or a camera device.

14. The imaging support device according to claim 1 , wherein the first image and the second image are medical images acquired of the same imaging target site using the same medical image diagnostic device or different medical image diagnostic devices.

15. acquiring a first image and a second image that is an image older than the first image; calculating a correction value for reproducing the second image; determining a repeatability of the second image based on the correction value; Imaging assistance method.

16. On the computer, An acquisition function for acquiring a first image and a second image that is an image older than the first image; a calculation function for calculating a correction value for reproducing the second image; a determination function for determining reproducibility of the second image based on the correction value; An imaging support program to achieve this.

Images