Medical image diagnostic device, medical image processing device, and method

Abstract

To automatically determine an appropriate imaging condition for next imaging according to a previous imaging result, during image diagnostic examination in which two or more times of imaging can be continuously performed.SOLUTION: A medical image diagnostic device according to an embodiment includes an acquisition part and an imaging condition determination part. The acquisition part acquires subject information on subject including a first medical image obtained by imaging a diagnostic target portion of the subject under a first imaging condition. The imaging condition determination part determines a second imaging condition for imaging a second medical image of the subject on the basis of the subject information and a learned model that has learned with evaluation information on a medical image and an imaging condition in association with each other.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The embodiments disclosed in this specification and the drawings relate to a medical image diagnostic apparatus, a medical image processing apparatus, and a method. [Background technology]

[0002] In conventional imaging diagnostic examinations, an operator such as a technician determines imaging conditions such as the imaging range based on an examination order. For example, imaging the entire body of a subject at once, regardless of the region to be diagnosed, allows for a sufficient examination for diagnosis. However, in the case of examinations involving invasiveness due to radiation exposure or in consideration of time and costs, it is undesirable to set the imaging range too wide. For this reason, conventionally, an operator determines an appropriate imaging range. Furthermore, because the imaging conditions appropriate for diagnosis also depend on the condition of the subject at the time of imaging, it has been desirable to be able to dynamically determine the imaging conditions during imaging.

[0003] For example, in the case of blood tests, a model technology is known that dynamically determines the type of blood test to be performed next based on the results of the previous blood test so as to improve the accuracy of disease classification. In imaging diagnostic tests, it is necessary to consider the imaging position according to the subject, so a method that simply determines the type and order, as in blood tests, cannot be applied to automatically determine the imaging conditions. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] Zheng Yu, Yikuan Li, Joseph C. Kim, Kaixuan Huang, Yuan Luo, Mengdi Wang, "DEEP REINFORCEMENT LEARNING FOR COST-EFFECTIVE MEDICAL DIAGNOSIS," published on February 28, 2023, as a presentation for ICLR (International Conference on Learning Representations) 2023. 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 automatically determine appropriate imaging conditions for the next imaging operation during an imaging diagnostic examination in which two or more imaging operations can be performed consecutively, based on the results of the previous imaging operation. 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 the configurations of the embodiments described below can also be considered as other problems. [Means for solving the problem]

[0006] A medical image diagnostic apparatus according to an embodiment includes an acquisition unit and an imaging condition determination unit. The acquisition unit acquires subject information about the subject, including a first medical image in which a diagnostic target region of the subject is imaged under a first imaging condition. The imaging condition determination unit determines second imaging conditions for capturing a second medical image of the subject, based on the subject information and a trained model that has been trained by associating evaluation information about the medical image with the imaging conditions. [Brief explanation of the drawings]

[0007] [Figure 1] FIG. 1 is a diagram showing an example of the configuration of an X-ray CT apparatus according to the first embodiment. [Figure 2] FIG. 2 is a diagram showing an example of object information according to the first embodiment. [Figure 3] FIG. 3 is a diagram illustrating an example of evaluation information according to the first embodiment. [Figure 4] FIG. 4 is a diagram showing an example of imaging positions in an imaging diagnostic examination involving multiple imaging operations according to the first embodiment. [Figure 5] FIG. 5 is a diagram illustrating the formula (2) according to the first embodiment. [Figure 6] FIG. 6 is a flowchart showing an example of the flow of the learning process according to the first embodiment. [Figure 7]FIG. 7 is a flowchart showing an example of the flow of the imaging process according to the first embodiment. [Figure 8] FIG. 8 is a diagram showing an example of the configuration of a magnetic resonance imaging apparatus according to the second embodiment. [Figure 9] FIG. 9 is a flowchart showing an example of the flow of the imaging process according to the second embodiment. [Figure 10] FIG. 10 is a diagram showing an example of the configuration of an ultrasonic diagnostic apparatus according to the third embodiment. [Figure 11] FIG. 11 is a diagram illustrating an example of the configuration of a workstation according to the fourth embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0008] Hereinafter, embodiments of a medical image diagnostic apparatus, a medical image processing apparatus, and a method will be described in detail with reference to the drawings.

[0009] (First embodiment) FIG. 1 is a diagram showing an example of the configuration of an X-ray computed tomography (CT) apparatus 1 (hereinafter referred to as X-ray CT apparatus 1a) according to the first embodiment. The X-ray CT apparatus 1a is an example of a medical image diagnostic apparatus in this embodiment. The X-ray CT apparatus 1a may also be referred to as a radiation image diagnostic apparatus. The X-ray CT apparatus 1a is installed, for example, in a hospital or the like. The configuration of the X-ray CT apparatus 1a shown in FIG. 1 is an example and is not limited thereto. For example, the X-ray CT apparatus 1a may be configured to be able to image a subject in a standing or sitting position.

[0010] The X-ray CT apparatus 1a of this embodiment automatically determines appropriate imaging conditions for the next imaging depending on the results of the previous imaging during an imaging diagnostic examination in which multiple imaging sessions are performed consecutively. The X-ray CT apparatus 1a is assumed to be capable of performing at least two or more imaging sessions consecutively. In this embodiment, the terms "previous imaging" and "next imaging" simply mean the "previous imaging" and "next imaging" included in multiple imaging sessions within a single imaging diagnostic examination.

[0011] 1, the X-ray CT apparatus 1a may be communicably connected to an electronic medical record system 2 via a network N such as an in-hospital LAN (Local Area Network). The X-ray CT apparatus 1a may also be communicably connected to a hospital information system (HIS), a laboratory information system (LIS), a radiology information system (RIS), a medical image storage device, etc.

[0012] 1, the X-ray CT apparatus 1a includes a gantry device 10, a bed device 30, and a console device 40. Note that the X-ray CT apparatus 1a does not include a subject P (for example, a human body).

[0013] In this embodiment, the rotation axis of the rotating frame 13 in a non-tilted state or the longitudinal direction of the tabletop 33 of the bed device 30 is defined as the Z-axis direction, the axis perpendicular to the Z-axis direction and horizontal to the floor surface is defined as the X-axis direction, and the axis perpendicular to the Z-axis direction and perpendicular to the floor surface is defined as the Y-axis direction. For convenience of explanation, multiple gantry devices 10 are depicted in Figure 1, but the actual configuration of the X-ray CT device 1a includes only one gantry device 10.

[0014] The gantry 10 and the bed 30 operate based on a user's operation via the console 40 or an operation unit provided on the gantry 10 or the bed 30. The gantry 10, the bed 30, and the console 40 are connected to each other by wire or wirelessly so as to be able to communicate with each other.

[0015] The gantry device 10 is an apparatus having an imaging system that irradiates an object P with X-rays and collects detection data of the X-rays that have passed through the object P. More specifically, the gantry device 10 has an X-ray tube 11 (X-ray generation unit), a wedge 16, a collimator 17, an X-ray detector 12, an X-ray high voltage device 14, a DAS (Data Acquisition System) 18, a rotating frame 13, and a control device 15.

[0016] The X-ray tube 11 is a vacuum tube that generates X-rays by irradiating thermions from a cathode (filament) toward an anode (target) through the application of high voltage and supply of filament current from the X-ray high voltage device 14. X-rays are generated when thermions collide with the target. X-rays generated at the tube focus in the X-ray tube 11 are shaped into a cone beam via, for example, a collimator 17 and irradiated onto the subject P. For example, the X-ray tube 11 may be a rotating anode type X-ray tube that generates X-rays by irradiating a rotating anode with thermions.

[0017] As shown in Figure 1, X-rays irradiated in a cone beam shape spread out in a fan shape in the X-axis direction. For this reason, the angle indicating the spread of X-rays irradiated in a cone beam shape in the X-axis direction is called the fan angle. Also, the angle indicating the depth of X-rays irradiated in a cone beam shape in the Z-axis direction is called the cone angle. For this reason, the X-axis direction is also called the fan angle direction, and the Z-axis direction is also called the cone angle direction.

[0018] The X-ray detector 12 detects the X-rays emitted from the X-ray tube 11 and passed through the subject P, and outputs an electrical signal corresponding to the X-ray dose to the DAS 18.

[0019] The X-ray detector 12 has, for example, multiple detection element rows in which multiple detection elements are arranged in the channel direction along one arc centered on the focal point of the X-ray tube 11. Each of the multiple detection elements detects the amount of incident X-rays. There are various types of X-ray CT apparatus 1a, such as a rotate / rotate-type (third generation CT) in which the X-ray tube 11 and X-ray detector 12 rotate together around the subject P, and a stationary / rotate-type (fourth generation CT) in which a large number of X-ray detection elements arrayed in a ring shape are fixed and only the X-ray tube 11 rotates around the subject P, and any of these types can be applied to this embodiment.

[0020] More specifically, the X-ray detector 12 is a direct conversion type X-ray detector having a semiconductor element that converts incident X-rays into electric charges. The X-ray detector 12 of this embodiment includes at least one high-voltage electrode, at least one semiconductor element, and multiple readout electrodes. The semiconductor element is also called an X-ray conversion element.

[0021] Furthermore, the X-ray detector 12 of this embodiment may be of an energy integrated type collection method or a photon counting type collection method.

[0022] The rotating frame 13 supports the X-ray tube 11 and the X-ray detector 12 rotatably around a rotation axis. Specifically, the rotating frame 13 is an annular frame that supports the X-ray tube 11 and the X-ray detector 12 facing each other and rotates the X-ray tube 11 and the X-ray detector 12 using a control device 15, which will be described later. The rotating frame 13 is rotatably supported on a fixed frame made of a metal such as aluminum. The rotating frame 13 receives power from a drive mechanism of the control device 15 and rotates around the rotation axis at a constant angular velocity.

[0023] The rotating frame 13 supports not only the X-ray tube 11 and the X-ray detector 12, but also the X-ray high voltage generator 14 and the DAS 18. The rotating frame 13 is housed in a substantially cylindrical housing having an opening (bore) that forms the imaging space. The central axis of the opening coincides with the rotation axis of the rotating frame 13.

[0024] The X-ray high voltage device 14 includes a high-voltage generator having electrical circuits such as a transformer and a rectifier, and having the function of generating a high voltage to be applied to the X-ray tube 11 and a filament current to be supplied to the X-ray tube 11, and an X-ray control device that controls the output voltage according to the X-rays emitted by the X-ray tube 11. The high-voltage generator may be of a transformer type or an inverter type. The X-ray high-voltage device 14 may be provided on the rotating frame 13, or may be provided on the fixed frame (not shown) side of the gantry device 10. The fixed frame is a frame that rotatably supports the rotating frame 13.

[0025] The control device 15 includes a processing circuit having a central processing unit (CPU) and the like, and a drive mechanism for a motor, an actuator, and the like. The processing circuit includes, as hardware resources, a processor such as a CPU or a microprocessing unit (MPU) and a memory such as a read-only memory (ROM) or a random access memory (RAM). The control device 15 may also be implemented by a processor such as a graphics processing unit (GPU), an application specific integrated circuit (ASIC), or a programmable logic device (e.g., a simple programmable logic device (SPLD), a complex programmable logic device (CPLD), or a field programmable gate array (FPGA)). When the processor is a CPU, for example, the processor realizes a function by reading and executing a program stored in memory. On the other hand, when the processor is an ASIC, instead of storing a program in memory, the function is directly incorporated into the processor circuit as a logic circuit. Each processor in this embodiment is not limited to being configured as a single circuit, but may be configured as a single processor by combining multiple independent circuits, or may be configured to realize the functions of a single processor by integrating multiple components.

[0026] The control device 15 also has a function of receiving input signals from the console device 40 or an input interface 43a attached to the gantry 10 and controlling the operation of the gantry 10 and the bed 30. For example, the control device 15 receives input signals and controls the rotation of the rotating frame 13, the tilt of the gantry 10, and the operation of the bed 30 and the tabletop 33. Note that the control of tilting the gantry 10 may be realized by the control device 15 rotating the rotating frame 13 around an axis parallel to the X-axis direction based on inclination angle (tilt angle) information input via the input interface 43a attached to the gantry 10. The control device 15 may be provided in the gantry 10 or in the console device 40.

[0027] The wedge 16 is a filter for adjusting the amount of X-rays emitted from the X-ray tube 11. The wedge 16 is, for example, a wedge filter or a bow-tie filter, and is a filter made of processed aluminum so as to have a predetermined target angle and a predetermined thickness.

[0028] The collimator 17 is a lead plate or the like for constricting the X-rays transmitted through the wedge 16 to an X-ray irradiation range, and a slit is formed by combining a plurality of lead plates or the like.

[0029] The DAS (Data Acquisition System) 18 has an amplifier that amplifies the electrical signals output from each X-ray detection element of the X-ray detector 12 and an A / D converter that converts the electrical signals into digital signals, and generates detection data. The detection data generated by the DAS 18 is transferred to the console device 40.

[0030] In this embodiment, the term "detection data" encompasses both pure raw data detected by the X-ray detector 12 and before preprocessing, and raw data obtained by preprocessing the pure raw data. The data before preprocessing (detection data) and the data after preprocessing may also be collectively referred to as projection data.

[0031] The bed device 30 is a device on which the subject P to be scanned is placed and moved, and includes a base 31, a bed driving device 32, a top 33, and a top support frame 34. The base 31 is a housing that supports the top support frame 34 so that it can move vertically. The bed driving device 32 is a motor or actuator that moves the top 33, on which the subject P is placed, in the longitudinal direction of the top 33. The bed driving device 32 moves the top 33 under the control of the console device 40 or the control device 15. The top 33, which is provided on the upper surface of the top support frame 34, is a plate on which the subject P is placed. Note that the bed driving device 32 may move the top support frame 34 in addition to the top 33 in the longitudinal direction of the top 33.

[0032] The console device 40 is a device that controls the gantry device 10 and performs operations such as generating CT images based on the scan results obtained by the gantry device 10. The console device 40 has a memory circuitry 41a (memory unit), a display 42a (display unit), an input interface 43a (input unit), an NW (network) interface 44a, and a processing circuitry 45a (processing unit). Data communication between the memory circuitry 41a, the display 42a, the input interface 43a, and the processing circuitry 45a is performed via a bus (BUS).

[0033] The memory circuitry 41a is realized by, for example, a semiconductor memory element such as a random access memory (RAM) or flash memory, a hard disk drive (HDD), a solid state drive (SSD), an optical disk, or the like. The memory circuitry 41a may also be a drive device that reads and writes various information from and to a portable storage medium such as a compact disc (CD), a digital versatile disc (DVD), or flash memory, or a semiconductor memory element such as a random access memory (RAM). The memory circuitry 41a stores, for example, projection data and reconstructed image data. The storage area of ​​the memory circuitry 41a may be located within the X-ray CT apparatus 1a or in an external storage device connected via a network. The memory circuitry 41a stores a control program according to this embodiment. The memory circuitry 41a is an example of a storage unit.

[0034] The display 42a displays various types of information. For example, the display 42a outputs medical images (CT images) generated by the processing circuitry 45a, a GUI (Graphical User Interface) for receiving various operations from the operator, and the like. For example, the display 42a may be a liquid crystal display (LCD), an organic electroluminescence display (OLED), a plasma display, or any other display, as appropriate. The display 42a may also be provided on the gantry device 10. The display 42a may also be a desktop type, or may be configured as a tablet terminal or the like capable of wireless communication with the console device 40 main body.

[0035] The input interface 43a accepts various input operations from the operator, converts the accepted input operations into electrical signals, and outputs the electrical signals to the processing circuitry 45a. For example, the input interface 43a accepts from the operator acquisition conditions for acquiring projection data, reconstruction conditions for reconstructing CT images, image processing conditions for generating post-processed images from CT images, etc. As the input interface 43a, for example, a mouse, keyboard, trackball, switch, button, joystick, touchpad, touch panel display, etc. can be used as appropriate.

[0036] In this embodiment, the input interface 43a is not limited to one having physical operation components such as a mouse, keyboard, trackball, switch, button, joystick, touchpad, and touch panel display. For example, an example of the input interface 43a also includes an electrical signal processing circuit that receives an electrical signal corresponding to an input operation from an external input device provided separately from the device and outputs the electrical signal to the processing circuit 45a. The input interface 43a is also an example of an input unit. The input interface 43a may be provided in the gantry device 10. The input interface 43a may also be configured as a tablet terminal or the like that is capable of wireless communication with the console device 40 main body.

[0037] The NW interface 44a acquires subject information about the subject P from the electronic medical record system 2 or other devices via the network N.

[0038] The processing circuitry 45a controls the overall operation of the X-ray CT apparatus 1a in response to electrical signals of input operations output from the input interface 43a. For example, the processing circuitry 45a includes an imaging function 451a, a transmission / reception function 452a, an estimation function 453a, a learning function 454a, an imaging condition determination function 455a, and a display control function 456a. Here, for example, the processing functions executed by the imaging function 451a, the transmission / reception function 452a, the estimation function 453a, the learning function 454a, the imaging condition determination function 455a, and the display control function 456a, which are components of the processing circuitry 45a shown in FIG. 1, are recorded in the storage circuitry 41a in the form of computer-executable programs. The processing circuitry 45a is, for example, a processor that reads and executes each program from the storage circuitry 41a to realize a function corresponding to the read program. In other words, the processing circuitry 45a in a state in which each program has been read has the functions shown in the processing circuitry 45a of FIG. 1. The photographing function 451a is an example of a photographing unit. The transmission and reception function 452a is an example of a transmission and reception unit and an output unit. The photographing function 451a and the transmission and reception function 452a are examples of an acquisition unit. The estimation function 453a is an example of an estimation unit. The learning function 454a is an example of a learning unit. The photographing condition determination function 455a is an example of a photographing condition determination unit. The display control function 456a is an example of a display control unit and an output unit.

[0039] 1 illustrates a case where the photographing function 451a, the transmitting / receiving function 452a, the estimation function 453a, the learning function 454a, the photographing condition determination function 455a, and the display control function 456a are realized by a single processing circuit 45a, but the embodiment is not limited to this. For example, the processing circuit 45a may be configured by combining multiple independent processors, and each processor may execute a respective program to realize each processing function. Furthermore, each processing function of the processing circuit 45a may be realized by being appropriately distributed or integrated into a single or multiple processing circuits.

[0040] The imaging function 451a controls each component of the X-ray CT device 1a to perform imaging of the subject P. The imaging function 451a obtains CT image data as an imaging result. The CT image represented by the CT image data is an example of the first medical image, the second medical image, and the learning medical image in this embodiment. Note that in this embodiment, the term "CT image" is not limited to tomographic image data of an arbitrary cross section, but may also refer to three-dimensional image data.

[0041] For example, the imaging function 451a performs preprocessing, reconstruction, image processing, etc. Specifically, the imaging function 451a generates projection data by performing preprocessing such as logarithmic conversion processing, offset correction processing, inter-channel sensitivity correction processing, and beam hardening correction on the detection data output from the DAS 18. Furthermore, the imaging function 451a performs reconstruction processing using a filtered back projection method, an iterative reconstruction method, etc. on the projection data generated by the preprocessing to generate a CT image. Furthermore, the imaging function 451a may perform various image processing using known methods on the CT image generated by the reconstruction processing.

[0042] The imaging conditions for imaging the subject P are determined by an imaging condition determination function 455a, which will be described later. The imaging function 451a acquires a CT image by imaging the subject P under the imaging conditions determined by the imaging condition determination function 455a. Note that, in the first imaging, imaging may be performed based on default imaging conditions designated by a user or the like, rather than the imaging conditions determined by the imaging condition determination function 455a.

[0043] The imaging conditions for the CT image may include, for example, the imaging position where X-rays are irradiated, the X-ray dose, image parameters, the operation of the gantry device 10 and the bed driving device 32, the timing of X-ray irradiation, etc. In this embodiment, the imaging conditions include at least the imaging position.

[0044] The transmission / reception function 452a may acquire subject information other than CT images relating to the subject P from the electronic medical record system 2 or other devices via the network N and the NW interface 44a.

[0045] The subject information is information relating to the body of the subject P. More specifically, in this embodiment, the subject information includes at least CT images captured by the X-ray CT apparatus 1a. In addition to the CT image data, the subject information may also include, for example, an examination order for the subject P and medical information on an electronic medical record.

[0046] Fig. 2 is a diagram showing an example of subject information according to this embodiment. As shown in Fig. 2, the subject information according to this embodiment includes a CT image of the subject P and medical information of the subject P acquired from the electronic medical record system 2. The medical information of the subject P includes, for example, the name of a disease.

[0047] In this embodiment, the object information of the object P at time t is S t In an imaging diagnostic examination including multiple imaging sessions, subject information including a CT image captured in the first imaging session is represented as S0. The imaging position 50a shown in FIG. 2 is the imaging position for the first imaging session of an imaging diagnostic examination including multiple imaging sessions performed by the X-ray CT apparatus 1a of this embodiment. In other words, the subject information S0 shown in FIG. 2 is subject information at the time when the first imaging session has been completed.

[0048] In an imaging diagnostic examination including multiple imaging sessions, each time one imaging session is completed by the imaging function 451a, the captured CT image is added to the subject information. For example, after the second imaging session is completed, the CT image of the subject P captured in the second imaging session is included in the subject information. Furthermore, after the third imaging session is completed, the CT image of the subject P captured in the third imaging session is included in the subject information. In other words, the subject information of this embodiment includes the CT images captured from the first imaging session to the previous imaging session during an imaging diagnostic examination including multiple imaging sessions.

[0049] Returning to FIG. 1, the estimation function 453a obtains, based on the subject information, evaluation information about an estimated image that is estimated to be obtained by imaging the diagnostic target region of the subject P under imaging conditions that are different from the imaging conditions when the CT image included in the subject information was captured. Different imaging conditions are, for example, imaging conditions with different imaging positions. Basically, the same position is not imaged multiple times in X-ray CT imaging, so the estimation function 453a obtains evaluation information about an estimated image that is estimated to be obtained by imaging the diagnostic target region of the subject P under imaging positions that are different from the imaging positions when the CT image included in the subject information was captured.

[0050] More specifically, the estimation function 453a estimates an estimated image based on the object information, and generates evaluation information based on the CT image included in the object information and the estimated image.

[0051] The evaluation information in this embodiment includes, for example, evaluation information on the image quality of the estimated image and evaluation information on the uncertainty of the disease classification estimated to be obtained when disease classification is performed based on the CT image and estimated image included in the subject information. The evaluation information on the image quality of the estimated image is an example of first evaluation information in this embodiment. The evaluation information on the uncertainty of the disease classification estimated to be obtained when disease classification is performed based on the CT image and estimated image included in the subject information is an example of second evaluation information in this embodiment.

[0052] FIG. 3 is a diagram showing an example of evaluation information according to this embodiment. As shown in FIG. 3, in this embodiment, the diagnostic target region of the subject P is the whole body. In the example shown in FIG. 3, it is assumed that three imaging sessions have already been performed in an imaging diagnostic examination including multiple imaging sessions. Specifically, of the whole body of the subject P, which is the diagnostic target region, imaging sessions have been performed at three imaging positions 50a to 50c (hereinafter, when there is no need to distinguish between the imaging positions 50a to 50c, simply referred to as imaging position 50). The estimation function 453a estimates an estimated image that is estimated to be obtained by imaging at an imaging position different from the imaging positions used in the past three imaging sessions.

[0053] Specifically, the estimation function 453a estimates estimated images 61a-61d (hereinafter, when the estimated images 61a-61d are not particularly distinguished from one another, they are simply referred to as unphotographed portions 60) that are estimated to be obtained by photographing the unphotographed portions 60a-60d (hereinafter, when the estimated images 61a-61d are not particularly distinguished from one another, they are simply referred to as unphotographed portions 60). The estimated image 61a is an image of the unphotographed portion 60a estimated to have been photographed by the X-ray CT device 1a. The estimated image 61b is an image of the unphotographed portion 60b estimated to have been photographed by the X-ray CT device 1a. The estimated image 61c is an image of the unphotographed portion 60c estimated to have been photographed by the X-ray CT device 1a. The estimated image 61d is an image of the unphotographed portion 60d estimated to have been photographed by the X-ray CT device 1a. That is, the estimation function 453a estimates the state of the entire diagnostic target region of the subject P based on the captured CT image 51 and the estimated image 61 of the uncaptured portion 60.

[0054] Note that one estimated image 61 is not necessarily generated for one unphotographed portion 60, and multiple estimated images 61 may be generated for one unphotographed portion 60. Furthermore, the estimated image 61 may differ from the shooting conditions used when photographing the image at the shooting positions 50a to 50c in terms of conditions other than the shooting positions.

[0055] If the diagnostic target region specified in the test order is not the whole body of the subject P, but a part of the body of the subject P, such as the abdomen or the head, or a specific organ, such as the heart, the estimation function 453a generates an estimated image 61 for an unphotographed portion within the diagnostic target region. In this case, the estimation function 453a does not generate an estimated image 61 for an unphotographed portion outside the diagnostic target region.

[0056] 3 is obtained by imaging at imaging position 50a, CT image 51b is obtained by imaging at imaging position 50b, and CT image 51c is obtained by imaging at imaging position 50c.

[0057] The estimation function 453a evaluates the image quality of CT images 51a to 51c (hereinafter, when there is no need to distinguish between the CT images 51a to 51c) obtained by imaging at imaging positions 50a to 50c, and estimated images 61a to 61c. In this embodiment, the estimation function 453a evaluates the image quality of both the CT image 51 and the estimated image 61, for example, but the estimation function 453a evaluates the image quality of at least the estimated image 61.

[0058] As evaluation information regarding the image quality of the estimated image 61, evaluation values ​​of image quality such as SNR (Signal to Noise Ratio), PSNR (Peak Signal to Noise Ratio), and SSIM (Structural SIMilarity) can be used. The estimation function 453a calculates the SNR, PSNR, or SSIM for each of the CT images 51a to 51c and the estimated images 61a to 61c, and calculates the average value of the calculation results. This average value is the evaluation value of the image quality of the estimated image. For example, the image quality of the estimated image varies depending on the X-ray dose included in the imaging conditions. Note that the estimation function 453a may use any one of SNR, PSNR, and SSIM, or may calculate an evaluation value by combining some of SNR, PSNR, and SSIM. The estimation function 453a assigns a higher evaluation to the estimated image 61 as the image quality of the estimated image 61 increases.

[0059] As evaluation information on the uncertainty of disease classification estimated when disease classification is performed based on the CT image and estimated image included in the subject information (hereinafter referred to as evaluation information on the uncertainty of disease classification), for example, the uncertainty of the model in the disease classification model or the report creation model can be used. The greater the uncertainty, the more insufficient the information (the more likely it is that additional imaging is required and / or imaging is required under different conditions). For example, computer-aided detection (CADe) or computer-aided diagnosis (CADx) may be used to classify diseases. If the disease classification model or the report creation model is a neural network, the uncertainty may be calculated using the Monte Carlo Dropout method. Alternatively, the estimation function 453a may evaluate the uncertainty of disease classification using a rule-based automatic diagnosis method. The lower the uncertainty of disease classification, the higher the evaluation.

[0060] The estimation function 453a calculates an evaluation value r based on evaluation information related to the image quality of the estimated image 61 and evaluation information related to the uncertainty of the disease classification. For example, the evaluation value r increases as the image quality of the estimated image 61 increases, and decreases as the uncertainty of the disease classification decreases. Note that weighting of the image quality and the uncertainty of the disease classification may be set appropriately. The evaluation value r is an example of evaluation information in this embodiment. Note that, although the evaluation value r is minimized as evaluation information in this embodiment, the evaluation information does not necessarily have to be expressed numerically, and may be expressed by a high or low rank, etc.

[0061] The estimation function 453a may use a combination of multiple evaluation methods for evaluating image quality and disease classification. Furthermore, in addition to evaluation information regarding the image quality of the estimated image 61 and evaluation information regarding the uncertainty of disease classification, the estimation function 453a may also use the examination time, dose, and bed movement distance for evaluation. For example, because the time frame of the imaging diagnostic examination for the subject P is predetermined, the estimation function 453a may use a penalty to subtract the evaluation value r if the examination time due to the estimated imaging conditions exceeds a predetermined percentage of the time frame. Furthermore, because there is an upper limit to the X-ray dose that can be irradiated to the subject P, the estimation function 453a may use a penalty to subtract the evaluation value r if the X-ray dose exceeds a threshold. The estimation function 453a may evaluate the X-ray dose threshold not only for the current imaging diagnostic examination but also for the X-ray doses of other examinations within a specified period. Furthermore, if the bed movement distance to move to the next imaging position 50 is long, the movement takes time, which may ultimately affect the length of the examination. For this reason, the estimation function 453a may use a penalty that subtracts from the evaluation value r the longer the movement distance from the bed position under the imaging conditions in the previous imaging.

[0062] Returning to FIG. 1 , the learning function 454a generates a trained model by associating evaluation information about a medical image (e.g., a CT image) with imaging conditions to train the model so that the evaluation value r is high. For example, the learning function 454a trains the model by associating a CT image, an estimated image generated from the CT image, an evaluation value r calculated based on the CT image and the estimated image, and the imaging conditions under which the CT image was captured. More specifically, the learning function 454a trains the model so that the image quality of the medical image is improved and the uncertainty of the disease classification estimated to be obtained when disease classification is performed based on the medical image is reduced. For example, a deep learning model such as a neural network can be adopted as the model.

[0063] Reinforcement learning or machine learning can be used as a method for training the model. As an example of reinforcement learning, a policy-based reinforcement learning algorithm may be used. Policy-based reinforcement learning is an algorithm that improves a policy π to obtain a policy that maximizes value. In this embodiment, the policy π corresponds to the shooting conditions. In this embodiment, the trained model may be referred to as a shooting policy model.

[0064] The learning function 454a of this embodiment may use the Actor-Critic method as an example of policy-based reinforcement learning. The following formula (1) is an example of a mathematical formula representing an algorithm when the learning function 454a performs reinforcement learning using the Actor-Critic method so as to increase the evaluation value r. The learning method using formula (1) is reinforcement learning that progresses as the difference between the target and the state value becomes smaller.

[0065]

number

[0066] Time t represents the time when the previous image was taken in an imaging diagnostic test that includes multiple images. Time t+1 represents the time when the next image is taken. S t represents the object information at time t, that is, the object information including the CT images taken up until the last time. t includes the CT images taken up to the previous time and the estimated image generated based on the CT images taken up to the previous time. t+1 represents the object information at time t+1, that is, the object information including the next CT image to be taken. t represents the evaluation value r at time t. A represents the imaging conditions including the imaging position. γ is a variable representing the discount rate. The discount rate represents unknown parts in the next CT image to be taken, such as noise. θ is a network parameter. Also, ∇ θ J(θ) represents the gradient in reinforcement learning.

[0067] Note that the method is not limited to the actor-critic method, and other policy-based reinforcement learning methods may be adopted. Furthermore, in this embodiment, a learning process involving actual imaging using the X-ray CT apparatus 1a will be described, but the learning function 454a may also perform learning using simulations or phantoms. Furthermore, the method is not limited to performing imaging and learning in real time, and offline reinforcement learning may be used to learn a policy from imaging data including previously captured CT images and the imaging conditions for those CT images.

[0068] When the trained model generated by the learning function 454a receives input of subject information including a CT image and the imaging conditions of the CT image, it outputs imaging conditions that will increase the evaluation value r as imaging conditions for the next imaging. In other words, the trained model outputs imaging conditions that are likely to improve the image quality of the medical image and reduce the uncertainty of the disease classification that is estimated to be obtained when disease classification is performed based on the medical image.

[0069] Furthermore, the learning function 454a of this embodiment may continue reinforcement learning not only in the learning phase before the start of operation, but also in the operation phase.

[0070] 1, the imaging condition determination function 455a determines imaging conditions for the next imaging of the subject P based on the subject information including the CT image of the previous imaging and the trained model generated by the learning function 454a. More specifically, the imaging condition determination function 455a inputs the subject information including the CT image of the previous imaging and the imaging conditions of the previous imaging into the trained model, thereby obtaining imaging conditions that are most likely to result in a high evaluation value r in the next imaging.

[0071] 4 is a diagram showing an example of imaging positions 50a to 50c in an imaging diagnostic examination including multiple imaging according to this embodiment. In an imaging diagnostic examination including multiple imaging according to this embodiment, the imaging function 451a performs imaging under default imaging conditions for the first time. The default imaging conditions may be automatically determined by an examination order or may be manually set by an operator.

[0072] In the example shown in FIG. 4 , the first imaging is performed at an imaging position 50a determined by the default imaging conditions. Then, the imaging condition determination function 455a determines an imaging position 50b of the diagnostic target region for the second imaging based on the subject information including the CT image 51a obtained by imaging at the imaging position 50a. The imaging conditions for the first imaging are an example of the first imaging conditions in this embodiment, and the CT image 51a is an example of the first medical image in this embodiment. The imaging conditions including the imaging position 50b for the second imaging are an example of the second imaging conditions in this embodiment. The CT image 51b captured under the second imaging conditions is an example of the second medical image in this embodiment. The imaging condition determination function 455a in this embodiment determines the imaging positions included in the imaging conditions for the next imaging from the unimaged portion up to the previous imaging. For example, the imaging position 50b included in the second imaging conditions is included in the unimaged portion at the time the first imaging is completed.

[0073] Furthermore, after the second imaging, the imaging condition determination function 455a determines the imaging position 50c for the third imaging of the diagnostic target region based on the subject information including the CT image 51b obtained by imaging at the imaging position 50b. In this way, the imaging condition determination function 455a dynamically determines the next imaging condition as imaging progresses.

[0074] Furthermore, the imaging conditions determined by the imaging condition determination function 455a include information regarding whether or not the next imaging will be performed, and the imaging position of the subject P when the next imaging will be performed. That is, for the second and subsequent imaging, the imaging condition determination function 455a also determines each time whether or not to perform imaging. That is, in this embodiment, the X-ray CT apparatus 1a does not necessarily image the entire diagnostic target region (for example, the entire body of the subject P), and when the imaging condition determination function 455a determines not to perform the next imaging, the imaging of the subject P is terminated regardless of whether or not there is an unimaged portion in the diagnostic target region.

[0075] The imaging condition determination function 455a uses the trained model (imaging strategy model) trained by the learning function 454a as a method for determining the next imaging condition. The imaging condition determination function 455a inputs subject information including the previously captured CT image to the trained model trained by the learning function 454a, and determines the imaging condition output from the trained model that is most likely to result in a high evaluation value r as the next imaging condition.

[0076] The determination process by the photographing condition determination function 455a is expressed by the following formula (2). The following formula (2) is, for example, the photographing condition a that maximizes the evaluation value at the next photographing time (t+1) by the measure π. t+1 S in equation (2) t represents the object information at time t. More specifically, S in Equation (2) t includes the CT images taken up to the last time and the estimated image estimated from the CT images. In other words, S in Equation (2) t indicates the state of the entire diagnostic target region of the subject P at time t.

[0077]

number

[0078] 5 is a diagram illustrating formula (2) according to this embodiment. As shown in FIG. 5, A represents the entire range of the diagnostic target region of the subject P, that is, the whole body of the subject P in this embodiment. a represents a position included in A, and the imaging position 50a included in the imaging conditions for past imaging, specifically the previous imaging, is denoted as a t , the photographing condition obtained by the learned model, that is, the photographing position 50b included in the photographing condition for the next photographing, is a t+1 It is expressed as:

[0079] Furthermore, when the trained model outputs the result that the next imaging will not be performed, i.e., the end of imaging, there is a state in which the evaluation value r does not improve even if the number of imaging attempts is further increased. A state in which the evaluation value r does not improve even if the number of imaging attempts is further increased is, for example, when the evaluation regarding the uncertainty of the disease classification does not increase. Another example of a state in which the evaluation value r does not improve even if the number of imaging attempts is further increased is when the evaluation value r is reduced due to the above-mentioned penalties, such as an increase in the examination time or an increase in the X-ray dose, caused by further increasing the number of imaging attempts. In such a case, the imaging condition determination function 455a determines to end imaging.

[0080] Returning to FIG. 1 , the display control function 456a controls the display 42a to display various screens and images. For example, the display control function 456a causes the display 42a to display a CT image captured by the imaging function 451a. The display control function 456a may also cause the display 42a to display the next imaging conditions determined by the imaging condition determination function 455a. In this embodiment, the imaging function 451a may automatically start the next imaging based on the next imaging conditions determined by the imaging condition determination function 455a, or the user may confirm the determined next imaging conditions on the display 42a and then start the next imaging by an operation of the user.

[0081] The display control function 456a may also cause the display 42a to display an operation screen on which the user can change the next imaging condition determined by the imaging condition determination function 455a. Also, for example, instead of the imaging function 451a automatically performing imaging based on the imaging conditions, the user may manually set the imaging conditions of the X-ray CT apparatus 1a in accordance with the contents of the imaging conditions displayed on the display 42a.

[0082] Next, the flow of the learning process executed by the X-ray CT apparatus 1a configured as above will be described.

[0083] 6 is a flowchart showing an example of the flow of the learning process according to this embodiment. Before this flowchart is executed, the examination order and medical information of the subject P are acquired by the transmission / reception function 452a.

[0084] First, imaging conditions for the first imaging are determined (S1). In the first imaging, imaging may be performed based on default imaging conditions designated by a user or the like, rather than imaging conditions determined by the imaging condition determination function 455a. In addition, the imaging condition determination function 455a may determine imaging conditions for the first imaging based on an examination order or medical information of the subject P. The imaging conditions in the learning process are an example of learning imaging conditions in this embodiment.

[0085] Then, the imaging function 451a performs a first imaging under the determined imaging conditions to obtain a medical image (CT image 51) (S2). The CT image 51 is an example of a training medical image included in the training subject information.

[0086] Then, the estimation function 453a generates an estimated image 61 of an unimaged portion of the diagnostic target region of the subject P other than the imaging position 50 under the imaging conditions of S1 based on the CT image 51 acquired in S2 (S3). The estimated image 61 is an example of a learning estimated image in this embodiment.

[0087] Next, the estimation function 453a evaluates the image quality of the CT image 51 and the estimated image 61 (S4). Note that the evaluation of image quality may be performed on both the CT image 51 and the estimated image 61, or only on the estimated image 61.

[0088] Next, the estimation function 453a evaluates (S5) the uncertainty of the disease classification estimated to be obtained when the disease classification is performed based on the CT image 51 and the estimated image 61. The estimation function 453a calculates an evaluation value r based on the evaluations of S4 and S5.

[0089] Next, the learning function 454a performs reinforcement learning and updates the model so that the evaluation value r obtained by the estimation function 453a becomes higher (S6).

[0090] Next, the learning function 454a determines whether to end the learning (S7). The condition for ending the learning may be, for example, when the learning has converged or when a predetermined number of trials has been reached. The predetermined number of trials may be determined arbitrarily by the user who executes the learning.

[0091] If the learning is not completed (S7 "No"), the process returns to S1, and the photographing condition determination function 455a determines the photographing conditions for the second photographing. The photographing conditions output by the trained model updated in S6 are used as the photographing conditions for the second photographing. In this way, the processes of S1 to S6 are repeatedly executed while the learning is continuing.

[0092] If the learning is completed (S7 "Yes"), the learning function 454a stores the learned model in, for example, the storage circuitry 41a. At this point, the processing of this flowchart ends.

[0093] Next, the flow of the imaging process during operation after learning will be described. Fig. 7 is a flowchart showing an example of the flow of the imaging process according to this embodiment. This flowchart is executed, for example, in response to an examination order for an imaging diagnostic test of a subject P. It is assumed that the learning process described in Fig. 6 has already been executed before this flowchart is executed.

[0094] First, imaging conditions for the first imaging are determined (S101). The imaging conditions are an example of first imaging conditions. As in S1 of FIG. 6, the first imaging may be performed based on default imaging conditions designated by a user or the like, rather than imaging conditions determined by the imaging condition determination function 455a. Furthermore, the imaging condition determination function 455a may determine the imaging conditions for the first imaging based on an examination order or medical information of the subject P.

[0095] Then, the imaging function 451a executes the first imaging under the determined imaging conditions to acquire a medical image (CT image 51) (S102). The CT image 51 is an example of a first medical image.

[0096] Then, the estimation function 453a generates an estimated image 61 of the unphotographed portion of the diagnostic target region of the subject P other than the photographing position 50 under the photographing conditions of S101, based on the CT image 51 acquired in S102 (S103).

[0097] Then, the estimation function 453a evaluates the image quality of the CT image 51 acquired in S102 and the estimated image 61 estimated in S103 (S104). As in S4, the evaluation of image quality may be performed on both the CT image 51 and the estimated image 61, or only on the estimated image 61.

[0098] Next, the estimation function 453a evaluates (S105) the uncertainty of the disease classification estimated to be obtained when the disease classification is performed based on the CT image 51 and the estimated image 61. The estimation function 453a calculates an evaluation value r based on the evaluations of S104 and S105.

[0099] Then, the imaging condition determination function 455a determines the imaging conditions for the second imaging by inputting the imaging conditions determined in S101, the CT image 51 acquired in S102, and the estimated image 61 estimated in S103 into the trained model (S106). The imaging condition determination function 455a may update the trained model using the evaluation value r obtained by the estimation function 453a. Note that learning during operation is not essential. If learning is not performed during operation, the process of evaluating image quality in S104, the process of evaluating the uncertainty of disease classification in S105, and the process of calculating the evaluation value r may not be executed.

[0100] If the photographing conditions determined in S106 do not indicate the end of photographing (S107 "No"), the process returns to S101, and the photographing condition determination function 455a determines the photographing conditions for the second photographing.

[0101] If the imaging condition determined in S106 is the end of imaging (S107 "Yes"), the imaging diagnostic examination of the subject P ends. Here, the processing of this flowchart ends.

[0102] In this manner, the X-ray CT apparatus 1a of this embodiment acquires subject information about the subject P, including the CT image 51 of the subject P's diagnostic target region, which was previously captured, and determines the imaging conditions for the next imaging of the subject P based on the subject information and the trained model. Furthermore, the X-ray CT apparatus 1a of this embodiment determines the imaging conditions for the next imaging using a trained model that correlates evaluation information about medical images with imaging conditions. Therefore, the X-ray CT apparatus 1a of this embodiment can automatically determine appropriate imaging conditions for the next imaging based on the results of the previous imaging during an imaging diagnostic examination in which two or more imaging sessions can be performed consecutively. This can reduce the time and effort required for an operator to consider and determine imaging conditions, such as the imaging range, based on an examination order, as in the past.

[0103] Furthermore, the imaging conditions determined by the X-ray CT apparatus 1a of this embodiment include information regarding whether or not the next imaging will be performed, and the imaging position of the subject if the next imaging will be performed. That is, when the trained model outputs that the next imaging will not be performed, the X-ray CT apparatus 1a of this embodiment determines to end imaging even if there are unimaging portions within the diagnostic target region. Therefore, according to the X-ray CT apparatus 1a of this embodiment, it is possible to reduce the examination time and the X-ray dose received by the subject P after capturing the CT images 51 necessary for diagnosis, compared to, for example, capturing the entire diagnostic target region of the subject P.

[0104] Furthermore, the X-ray CT apparatus 1a of this embodiment obtains evaluation information regarding the estimated image 61 based on the subject information obtained in the previous imaging. The evaluation information includes at least one of first evaluation information regarding the image quality of the estimated image 61 and second evaluation information regarding the uncertainty of the disease classification estimated to be obtained when disease classification is performed based on the CT image 51 and the estimated image 61. Therefore, the X-ray CT apparatus 1a of this embodiment can determine imaging conditions for the next imaging in consideration of both high image quality and diagnostic accuracy.

[0105] The estimated image of this embodiment is a medical image that is estimated to be captured when an unphotographed portion of the diagnostic target region of the subject P is photographed. The X-ray CT apparatus 1a of this embodiment obtains evaluation information based on the previously photographed CT image 51 and the estimated image 61. Therefore, the X-ray CT apparatus 1a of this embodiment can estimate and evaluate the state of the entire diagnostic target region of the subject P based on the photographed CT image 51 and the estimated image 61 of the unphotographed portion.

[0106] Furthermore, the X-ray CT apparatus 1a of this embodiment trains a model based on the learning imaging conditions, learning object information, and learning evaluation information so as to improve the evaluation information regarding the learning estimated image 61, and determines the imaging conditions for the next imaging by inputting the object information including the CT image 51 captured in the previous imaging and the previous imaging conditions into the trained model. Therefore, the X-ray CT apparatus 1a of this embodiment can appropriately determine the imaging conditions for the unimaged portion.

[0107] In this embodiment, the CT image 51 is an example of the first medical image, the second medical image, and the learning medical image, but the first medical image, the second medical image, and the learning medical image may also be data that serves as the source of the image, such as a signal detected by the X-ray detector 12 and detection data generated by the DAS 18.

[0108] In the X-ray CT apparatus 1a of this embodiment, in an imaging diagnostic examination including multiple imaging, the first imaging is performed under predetermined imaging conditions, and imaging conditions for the second and subsequent imaging are determined by estimation based on the subject information. However, the X-ray CT apparatus 1a may automatically determine imaging conditions for the first imaging in an imaging diagnostic examination including multiple imaging. For example, the transmission / reception function 452a may acquire, before the first imaging, a CT image captured before the current imaging diagnostic examination including multiple imaging, for example, a CT image captured on the previous day or earlier, as subject information.

[0109] Furthermore, medical image data captured by another modality prior to the current imaging diagnostic examination including multiple imaging may be acquired as subject information. Furthermore, the subject information used to determine the imaging conditions for the first imaging may not include medical images. For example, the transmission / reception function 452a may acquire the subject information used to determine the imaging conditions for the first imaging, such as an examination order for subject P or medical information in an electronic medical record. In this case, the estimation function 453a generates an estimated image based on the subject information and obtains evaluation information regarding the generated estimated image.

[0110] In this embodiment, the estimation function 453a estimates evaluation information regarding the image quality of the estimated image and evaluation information regarding the uncertainty of the disease classification estimated to be obtained when disease classification is performed based on the CT image data and the estimated image included in the subject information, but the content of the evaluation information is not limited to this example. For example, the estimation function 453a may obtain at least one of evaluation information regarding the image quality of the estimated image and evaluation information regarding the uncertainty of the disease classification estimated to be obtained when disease classification is performed based on the CT image data and the estimated image included in the subject information. Alternatively, the estimation function 453a may further obtain other evaluation information.

[0111] In this embodiment, the trained model is stored in the memory circuitry 41a, for example, but the trained model may be incorporated into the imaging condition determination function 455a. Alternatively, the trained model may be stored in a storage device outside the X-ray CT apparatus 1a and read during processing.

[0112] (Second embodiment) In the first embodiment described above, the X-ray CT apparatus 1a has been described as an example of a medical image diagnostic apparatus. In this second embodiment, a magnetic resonance imaging (MRI) apparatus will be described as an example of a medical image diagnostic apparatus.

[0113] Fig. 8 is a diagram showing an example of the configuration of a magnetic resonance imaging apparatus 1b according to this embodiment. As shown in Fig. 8, the magnetic resonance imaging apparatus 1b includes a static magnetic field magnet 101, a static magnetic field power supply (not shown), a gradient magnetic field coil 103, a gradient magnetic field power supply 104, a bed 105, a bed control circuit 106, a transmission coil 107, a transmission circuit 108, a reception coil 109, a reception circuit 110, a sequence control circuit 120, and a computer system 130. Note that the configuration of the magnetic resonance imaging apparatus 1b shown in Fig. 8 is an example and is not limited to this.

[0114] 8 is merely an example. For example, the sequence control circuit 120 and the components in the computer system 130 may be integrated or separated as appropriate. The magnetic resonance imaging apparatus 1b does not include a subject P.

[0115] The X-axis, Y-axis, and Z-axis shown in Fig. 8 form an apparatus coordinate system specific to the magnetic resonance imaging apparatus 1b. For example, the Z-axis direction coincides with the axial direction of the cylinder of the gradient magnetic field coil 103 and is set along the magnetic flux of the static magnetic field generated by the static magnetic field magnet 101. The Z-axis direction is the same as the longitudinal direction of the bed 105 and also the same as the craniocaudal direction of the subject P placed on the bed 105. The X-axis direction is set along the horizontal direction perpendicular to the Z-axis direction. The Y-axis direction is set along the vertical direction perpendicular to the Z-axis direction.

[0116] The static magnetic field magnet 101 is a magnet formed in a hollow, approximately cylindrical shape, and generates a static magnetic field in the internal space. The static magnetic field magnet 101 is, for example, a superconducting magnet, and is excited by receiving a current from a static magnetic field power supply. The static magnetic field power supply supplies a current to the static magnetic field magnet 101. As another example, the static magnetic field magnet 101 may be a permanent magnet, in which case the magnetic resonance imaging apparatus 1b may not be equipped with a static magnetic field power supply. Furthermore, the static magnetic field power supply may be provided separately from the magnetic resonance imaging apparatus 1b.

[0117] The gradient magnetic field coil 103 is a hollow coil formed in a substantially cylindrical shape, and is disposed inside the static magnetic field magnet 101. The gradient magnetic field coil 103 is formed by combining three coils corresponding to the mutually orthogonal X, Y, and Z axes, and these three coils are individually supplied with current from a gradient magnetic field power supply 104 to generate a gradient magnetic field whose magnetic field strength changes along each of the X, Y, and Z axes. Moreover, the gradient magnetic field power supply 104 supplies current to the gradient magnetic field coil 103 under the control of a sequence control circuit 120.

[0118] The bed 105 includes a top plate 105a on which the subject P is placed, and the top plate 105a is inserted into the imaging cavity with the subject P, such as a patient, placed on it under the control of a bed control circuit 106. Under the control of a computer system 130, the bed control circuit 106 drives the bed 105 to move the top plate 105a in the longitudinal direction and the up-down direction.

[0119] The transmitting coil 107 excites a given region of the subject P by applying a high-frequency magnetic field. The transmitting coil 107 is, for example, a whole-body coil that surrounds the entire body of the subject P. The transmitting coil 107 receives RF pulses from the transmitting circuit 108, generates a high-frequency magnetic field, and applies the high-frequency magnetic field to the subject P. The transmitting circuit 108 supplies RF pulses to the transmitting coil 107 under the control of the sequence control circuit 120.

[0120] The receiving coil 109 is disposed inside the gradient magnetic field coil 103, and receives magnetic resonance signals (hereinafter referred to as MR (Magnetic Resonance) signals) emitted from the subject P due to the influence of a high frequency magnetic field. Upon receiving the MR signals, the receiving coil 109 outputs the received MR signals to the receiving circuit 110.

[0121] 8, the receiving coil 109 is configured to be provided separately from the transmitting coil 107, but this is just an example and the present invention is not limited to this configuration. For example, a configuration in which the receiving coil 109 also serves as the transmitting coil 107 may be adopted.

[0122] The receiving circuit 110 performs analog-to-digital (AD) conversion on the analog MR signal output from the receiving coil 109 to generate MR data. The receiving circuit 110 also transmits the generated MR data to the sequence control circuit 120. Note that the AD conversion may be performed within the receiving coil 109. The receiving circuit 110 is also capable of performing any signal processing other than AD conversion.

[0123] The sequence control circuit 120 drives the gradient magnetic field power supply 104, the transmission circuitry 108, and the reception circuitry 110 based on sequence information transmitted from the computer system 130, thereby imaging the subject P.

[0124] Here, the sequence information is information that defines a procedure for performing imaging. The sequence information defines the strength of the current that the gradient magnetic field power supply 104 supplies to the gradient magnetic field coil 103 and the timing of supplying the current, the strength of the RF pulse that the transmission circuit 108 supplies to the transmission coil 107 and the timing of applying the RF pulse, the timing of detecting the MR signal by the reception circuit 110, etc. The sequence information differs depending on the range of the region of the body of the subject P that is to be imaged.

[0125] The sequence control circuit 120 may be realized by a processor, or may be realized by a combination of software and hardware.

[0126] Furthermore, when the sequence control circuit 120 receives MR data from the receiving circuit 110 as a result of driving the gradient magnetic field power supply 104, the transmitting circuit 108, and the receiving circuit 110 to image the subject P, the sequence control circuit 120 transfers the received MR data to the computer system 130.

[0127] The computer system 130 performs overall control of the magnetic resonance imaging apparatus 1b, generates MR images, etc. The computer system 130 also performs overall control of the magnetic resonance imaging apparatus 1b, generates MR images, etc. As shown in Fig. 8, the computer system 130 includes a NW interface 44b, a memory circuitry 41b, a processing circuitry 45b, an input interface 43b, and a display 42b.

[0128] Furthermore, the magnetic resonance imaging apparatus 1b may be communicably connected to the electronic medical record system 2 and the like, similar to the X-ray CT apparatus 1a of the first embodiment described with reference to FIG.

[0129] The processing circuitry 45b controls the overall operation of the magnetic resonance imaging apparatus 1b. For example, the processing circuitry 45b includes an imaging function 451b, a transmission / reception function 452b, an estimation function 453b, a learning function 454b, an imaging condition determination function 455b, and a display control function 456b. Here, for example, the processing functions executed by the imaging function 451b, the transmission / reception function 452b, the estimation function 453b, the learning function 454b, the imaging condition determination function 455b, and the display control function 456b, which are components of the processing circuitry 45b shown in FIG. 8, are recorded in the storage circuitry 41b in the form of a computer-executable program. The processing circuitry 45b is, for example, a processor that reads and executes each program from the storage circuitry 41b to realize a function corresponding to the read program. In other words, the processing circuitry 45b in a state in which each program has been read has the functions shown in the processing circuitry 45b in FIG. 8. The imaging function 451b is an example of an imaging unit. The transmission and reception function 452b is an example of a transmission and reception unit and an output unit. The photographing function 451b and the transmission and reception function 452b are an example of an acquisition unit. The estimation function 453b is an example of an estimation unit. The learning function 454b is an example of a learning unit. The photographing condition determination function 455b is an example of a photographing condition determination unit. The display control function 456b is an example of a display control unit and an output unit.

[0130] 8 illustrates a case in which the photographing function 451b, the transmitting / receiving function 452b, the estimation function 453b, the learning function 454b, the photographing condition determination function 455b, and the display control function 456b are realized by a single processing circuit 45b, but the embodiment is not limited to this. For example, the processing circuit 45b may be configured by combining multiple independent processors, and each processor may execute a respective program to realize each processing function. Furthermore, each processing function of the processing circuit 45b may be realized by being appropriately distributed or integrated into a single or multiple processing circuits.

[0131] The imaging function 451b controls each part of the magnetic resonance imaging apparatus 1b based on the imaging conditions to capture a magnetic resonance image. For example, the imaging function 451b generates sequence information, collects MR data, generates k-space data, and generates a magnetic resonance image.

[0132] More specifically, the imaging function 451b generates sequence information that defines imaging procedures according to the type of magnetic resonance image to be captured and the imaging target region, and transmits the generated sequence information to the sequence control circuit 120 via the NW interface 44b. The sequence control circuit 120 executes various pulse sequences based on the sequence information generated by the imaging function 451b.

[0133] The imaging function 451b also collects MR data converted from MR signals emitted from the subject P by executing various pulse sequences from the sequence control circuit 120 via the NW interface 44b. The imaging function 451b also arranges the collected MR data in accordance with the phase encoding amount and frequency encoding amount imparted by the gradient magnetic field. The MR data arranged in the k-space is called k-space data. The k-space data is stored in the memory circuit 41b.

[0134] The imaging function 451b also generates a magnetic resonance image based on the k-space data stored in the memory circuitry 41b. For example, the imaging function 451b generates the magnetic resonance image by performing reconstruction processing such as Fourier transform on the k-space data. The imaging function 451b stores the generated magnetic resonance image in, for example, the memory circuitry 41b. The magnetic resonance image captured by the imaging function 451b is an example of the first medical image, the second medical image, and the learning medical image in this embodiment.

[0135] The imaging conditions in this embodiment include at least an imaging position and a type of magnetic resonance image. The imaging conditions may include an imaging region, TR (Repetition Time), TE (Echo Time), the number of slices, imaging direction, slice thickness, etc. The types of magnetic resonance images include, for example, a T1 weighted image (T1WI), a T2 weighted image (T2WI), and a FLAIR (Fluid Attenuated Inversion Recovery) image.

[0136] Furthermore, the subject information in this embodiment includes at least the magnetic resonance image captured by the magnetic resonance imaging apparatus 1b.

[0137] The transmitting / receiving function 452b may acquire subject information other than magnetic resonance images relating to the subject P from the electronic medical record system 2 or other devices via the NW interface 44b, similar to the transmitting / receiving function 452a of the first embodiment.

[0138] Based on the subject information, the estimation function 453b calculates evaluation information regarding an estimated image that is estimated to be obtained by imaging the diagnostic target area of ​​the subject P under imaging conditions different from the imaging conditions when the magnetic resonance image included in the subject information was taken.

[0139] In magnetic resonance imaging, the same imaging position may be imaged using different pulse sequences, and therefore the imaging position included in the next imaging condition may correspond to not only an unimaged portion of the diagnostic target region of the subject P, but also a portion that has already been imaged previously. Different imaging conditions differ in at least one of the imaging position, imaging direction, and image type. The imaging condition does not need to directly specify the image type, and may indirectly specify the image type using various parameters of the pulse sequence. The imaging direction in magnetic resonance imaging includes, for example, coronal, sagittal, and horizontal sections.

[0140] The evaluation information can be obtained by the method described in the first embodiment.

[0141] The learning function 454b generates a trained model by learning a model by associating evaluation information regarding medical images (e.g., magnetic resonance images) with shooting conditions so that the evaluation information (e.g., evaluation value r) is high, using the method described in the first embodiment.

[0142] The imaging condition determination function 455b determines imaging conditions for the next imaging of the subject P based on the subject information including the magnetic resonance image of the previous imaging and the trained model generated by the learning function 454b. As a method for determining imaging conditions, the method described in the first embodiment can be adopted.

[0143] The display control function 456b controls the display 42b to display various screens and images.

[0144] Next, the flow of imaging processing by the magnetic resonance imaging apparatus 1b of this embodiment configured as above will be described.

[0145] 9 is a flowchart showing an example of the flow of imaging processing according to this embodiment. The imaging processing in this embodiment is executed as an MRI sequence.

[0146] First, imaging conditions for the first imaging are determined, and the imaging function 451a executes the first imaging under the determined imaging conditions to acquire a magnetic resonance image (S201). This process corresponds to the processes of S101 and S102 in the imaging process flow of the first embodiment described in FIG. 7. In the example shown in FIG. 9, a T1-weighted image is designated as the image type under the first imaging conditions. As in the first embodiment, the first imaging may be performed based on default imaging conditions designated by a user or the like, rather than on imaging conditions determined by the imaging condition determination function 455b. In addition, in FIG. 9, the magnetic resonance image acquired in the first imaging is represented as object information S0.

[0147] Then, the estimation function 453b generates an estimated image based on the magnetic resonance image (subject information S0) acquired in S101 (S202). This process corresponds to the process of S103 in the flow of the imaging process of the first embodiment described in FIG. 7. In FIG. 9, the estimated image based on the magnetic resonance image captured in the first imaging is represented as estimated image Img0. The estimated image Img0 differs in imaging position or image type from the magnetic resonance image (subject information S0). Note that when the image types are different, the imaging positions of the estimated image Img0 and the subject information S0 may be the same. That is, the imaging position of the estimated image in this embodiment is not limited to an unimaged portion.

[0148] Then, the estimation function 453b evaluates the image quality of the estimated image Img0 and the magnetic resonance image (subject information S0) and the uncertainty of the disease classification estimated to be obtained when disease classification is performed based on the estimated image Img0 and the magnetic resonance image (subject information S0) (S203). This processing corresponds to the processing of S104 and S105 in the flow of the imaging processing of the first embodiment described in Fig. 7. In Fig. 9, the evaluation value of the estimated image Img0 is represented as r0.

[0149] Then, the imaging condition determination function 455b determines the imaging conditions for the second imaging by inputting the imaging conditions of the imaging in S201 and the magnetic resonance image (subject information S0) acquired in S201 into the trained model (S204). This processing corresponds to the processing of S106 in the flow of the imaging processing of the first embodiment described in FIG. 7. The imaging condition determination function 455b may update the trained model using the evaluation value r obtained by the estimation function 453b. The execution conditions for the second imaging are represented as a0.

[0150] In the example shown in Fig. 9, a T2 weighted image is specified as the type of image in the imaging conditions for the second imaging. In this case, the imaging function 451a captures a T2 weighted image (S205). In Fig. 9, object information including the magnetic resonance image captured in the second imaging is represented as object information S1.

[0151] Then, the estimation function 453b generates an estimated image Img1 based on the magnetic resonance image (subject information S1) acquired in S205 (S206).

[0152] Then, the estimation function 453b calculates an evaluation value r1 that evaluates the image quality of the estimated image Img1 and the magnetic resonance image (subject information S1) and the uncertainty of the disease classification that is estimated to be obtained when disease classification is performed based on the estimated image Img1 and the magnetic resonance image (subject information S1) (S207).

[0153] Then, the imaging condition determination function 455b determines the imaging conditions a1 for the third imaging by inputting the imaging conditions for the imaging in S205 and the magnetic resonance image (subject information S1) acquired in S205 into the trained model (S208).

[0154] In the example shown in Fig. 9, a FLAIR image is specified as the type of image in the imaging conditions for the third imaging. From the FLAIR imaging in S209 to the evaluation process in S211, the processes in S205 to S207 are executed in the same flow as those in S205 to S207. In the example shown in Fig. 9, in S212, the imaging condition determination function 455b determines the end of imaging as the next imaging condition. Here, the processing of this flowchart ends. Note that if the imaging condition determination function 455b further determines to capture another type of image or to capture at another imaging position, the imaging process continues.

[0155] Thus, in this embodiment, similar to the X-ray CT apparatus 1a in the first embodiment, the magnetic resonance imaging apparatus 1b can automatically determine appropriate imaging conditions for the next imaging depending on the results of the previous imaging during an imaging diagnostic examination in which two or more imaging sessions can be performed consecutively.

[0156] (Third embodiment) In this third embodiment, an ultrasonic diagnostic apparatus will be described as an example of a medical image diagnostic apparatus.

[0157] Fig. 10 is a diagram showing an example of the configuration of an ultrasound diagnostic device 1c according to this embodiment. As shown in Fig. 10, the ultrasound diagnostic device 1c according to this embodiment has an ultrasound probe 100, a display 42c, an input interface 43c, and a device main body 400, and the ultrasound probe 100, the display 42c, and the input interface 43c are communicably connected to the device main body 400. Note that the configuration of the ultrasound diagnostic device 1c shown in Fig. 10 is an example and is not limited to this.

[0158] The ultrasonic probe 100 is operated by a user and transmits ultrasonic waves to the subject P. Specifically, the ultrasonic probe 100 has a plurality of piezoelectric vibrators, which generate ultrasonic waves based on drive signals supplied from a transmission / reception circuit 401. The ultrasonic probe 100 also receives reflected waves from the subject P and converts them into electrical signals. The ultrasonic probe 100 also has a matching layer provided on the piezoelectric vibrators, a backing material that prevents ultrasonic waves from propagating backward from the piezoelectric vibrators, and the like. The ultrasonic probe 100 is detachably connected to the device main body 400.

[0159] When ultrasonic waves are transmitted from the ultrasonic probe 100 to the subject P, the transmitted ultrasonic waves are reflected successively by discontinuous surfaces of acoustic impedance in the tissues of the subject P, and are received as reflected wave signals by the multiple piezoelectric transducers of the ultrasonic probe 100. The amplitude of the received reflected wave signals depends on the difference in acoustic impedance at the discontinuous surfaces from which the ultrasonic waves are reflected. When the transmitted ultrasonic pulses are reflected by the surface of a moving blood flow, heart wall, or the like, the reflected wave signals undergo a frequency shift due to the Doppler effect, depending on the velocity component of the moving object in the direction of ultrasonic transmission.

[0160] The ultrasonic probe 100 may be a one-dimensional ultrasonic probe in which a plurality of piezoelectric vibrators are arranged in a row, or may be an ultrasonic probe in which a plurality of piezoelectric vibrators of a one-dimensional ultrasonic probe are mechanically oscillated, or may be a two-dimensional ultrasonic probe in which a plurality of piezoelectric vibrators are arranged two-dimensionally in a lattice pattern.

[0161] The display 42c displays a GUI (Graphical User Interface) that allows the operator of the ultrasound diagnostic apparatus 1c to input various setting requests using the input interface 43c, ultrasound images generated in the apparatus main body 400, etc. The display 42c also displays various messages and display information to notify the operator of the processing status and processing results of the apparatus main body 400. The display 42c also has a speaker and can output sound.

[0162] The input interface 43c is operated to set a predetermined position (e.g., the position of a region of interest (ROI)), and is realized by, for example, a trackball, switch buttons, a mouse, a keyboard, a touchpad that performs input operations by touching the operation surface, a touch monitor in which a display screen and a touchpad are integrated, a non-contact input circuit using an optical sensor, and a voice input circuit. The input interface 43c is connected to a processing circuit 45c (described later) and converts input operations received from an operator into electrical signals and outputs the electrical signals to the processing circuit 45c. Note that, in this specification, the input interface 43c is not limited to those having physical operation components such as a mouse and a keyboard. For example, an electrical signal processing circuit that receives electrical signals corresponding to input operations from an external input device provided separately from the device and outputs the electrical signals to the processing circuit 45c is also included as an example of an input interface.

[0163] The device main body 400 is a device that generates an ultrasound image based on the reflected wave signal received by the ultrasound probe 100, and as shown in FIG. 10, has a transmission / reception circuit 401, a signal processing circuit 402, an image memory 403, a memory circuit 41c, a NW interface 44c, and a processing circuit 45c.

[0164] Furthermore, the ultrasound diagnostic apparatus 1c may be communicably connected to the electronic medical record system 2 and the like, similar to the X-ray CT apparatus 1a of the first embodiment described with reference to FIG.

[0165] The processing circuitry 45c controls the overall operation of the ultrasound diagnostic apparatus 1c. For example, the processing circuitry 45c includes an imaging function 451c, a transmission / reception function 452c, an estimation function 453c, a learning function 454c, an imaging condition determination function 455c, and a display control function 456c. Here, for example, the processing functions executed by the imaging function 451c, the transmission / reception function 452c, the estimation function 453c, the learning function 454c, the imaging condition determination function 455c, and the display control function 456c, which are components of the processing circuitry 45c shown in FIG. 10, are recorded in the storage circuitry 41c in the form of computer-executable programs. The processing circuitry 45c is, for example, a processor that reads and executes each program from the storage circuitry 41c to realize a function corresponding to the read program. In other words, the processing circuitry 45c in a state in which each program has been read has the functions shown in the processing circuitry 45c in FIG. 10. The imaging function 451c is an example of an imaging unit. The transmission and reception function 452c is an example of a transmission and reception unit and an output unit. The photographing function 451c and the transmission and reception function 452c are an example of an acquisition unit. The estimation function 453c is an example of an estimation unit. The learning function 454c is an example of a learning unit. The photographing condition determination function 455c is an example of a photographing condition determination unit. The display control function 456c is an example of a display control unit and an output unit.

[0166] 10 illustrates a case where the photographing function 451c, the transmitting / receiving function 452c, the estimation function 453c, the learning function 454c, the photographing condition determination function 455c, and the display control function 456c are realized by a single processing circuit 45c, but the embodiment is not limited to this. For example, the processing circuit 45c may be configured by combining multiple independent processors, and each processor may execute a respective program to realize each processing function. Furthermore, each processing function of the processing circuit 45c may be realized by being appropriately distributed or integrated into a single or multiple processing circuits.

[0167] The imaging conditions in this embodiment include probe information related to the operation of the ultrasonic probe 100, and a dynamic range, gain, focus, and depth related to capturing an ultrasonic image. The probe information includes the angle and position of the ultrasonic probe 100. The angle and position of the ultrasonic probe 100 are an example of an imaging position in this embodiment.

[0168] The imaging function 451c controls various components of the ultrasound diagnostic apparatus 1c to capture ultrasound images. The ultrasound images captured by the imaging function 451c are examples of the first medical image, the second medical image, and the learning medical image in this embodiment.

[0169] The transmission / reception function 452c may acquire object information other than ultrasound images relating to the object P from the electronic medical record system 2 or other devices via the NW interface 44c, similar to the transmission / reception function 452a of the first embodiment.

[0170] Based on the subject information, the estimation function 453c obtains evaluation information regarding an estimated image that is estimated to be obtained by photographing the diagnostic target area of ​​the subject P under imaging conditions different from the imaging conditions when the ultrasound image included in the subject information was photographed.

[0171] In ultrasound image diagnosis, the same imaging position may be imaged with different types of images, and therefore the imaging position included in the next imaging condition may correspond to not only an unimaged portion of the diagnostic target region of the subject P, but also a portion that has been imaged previously. Different imaging conditions differ in at least one of the imaging position or the type of image. Note that the imaging condition does not need to directly specify the type of image, and may indirectly specify the type of image using various parameters.

[0172] The evaluation information can be obtained by the method described in the first embodiment.

[0173] The learning function 454c generates a trained model by learning a model by associating evaluation information regarding medical images (e.g., ultrasound images) with shooting conditions so that the evaluation information (e.g., evaluation value r) is high, using the method described in the first embodiment.

[0174] The imaging condition determination function 455c determines imaging conditions for the next imaging of the subject P based on the subject information including the ultrasound image of the previous imaging and the trained model generated by the learning function 454c. As a method for determining the imaging conditions, the method described in the first embodiment can be adopted.

[0175] The display control function 456c controls the display 42c to display various screens and images. The display control function 456b may also cause the display 42c to display probe information for the next imaging determined by the imaging condition determination function 455c. The user presses the ultrasound probe 100 against the subject P based on the probe information displayed on the display 42c.

[0176] If the ultrasonic diagnostic device 1c is a fully automatic ultrasonic diagnostic device, the ultrasonic probe 100 may operate automatically. In this case, the imaging function 451c controls the angle and position of the ultrasonic probe 100 based on the probe information determined by the imaging condition determination function 455c.

[0177] As described above, in this embodiment, similar to the X-ray CT apparatus 1a in the first embodiment, the ultrasound diagnostic apparatus 1c can automatically determine appropriate imaging conditions for the next imaging depending on the results of the previous imaging during an imaging diagnostic examination in which two or more imaging sessions can be performed consecutively.

[0178] In the above-described first to third embodiments, the X-ray CT apparatus 1a, the magnetic resonance imaging apparatus 1b, and the ultrasound diagnostic apparatus 1c are given as examples of medical image diagnostic apparatuses, but the medical image diagnostic apparatus is not limited to these and may be of other modalities.

[0179] (Fourth embodiment) In the first to third embodiments described above, the processing is executed by a medical image diagnostic apparatus, but in this fourth embodiment, an example in which the processing is executed by a medical image processing apparatus will be described.

[0180] Fig. 11 is a diagram showing an example of the configuration of a workstation 1d according to the fourth embodiment. The workstation 1d is an example of a medical image processing device in this embodiment. The workstation 1d is communicably connected to, for example, an electronic medical record system 2 and a modality 9 via a network N. In Fig. 11, the workstation 1d is connected to one modality 9, but the workstation 1d may be connected to multiple modalities 9.

[0181] As shown in Fig. 11, the workstation 1d includes a storage circuit 41d, a display 42d, an input interface 43d, a NW interface 44d, and a processing circuit 45d. Note that the configuration of the workstation 1d shown in Fig. 11 is an example and is not limited to this.

[0182] The processing circuitry 45d controls the overall operation of the workstation 1d. For example, the processing circuitry 45d includes an acquisition function 457, an estimation function 453d, a learning function 454d, a shooting condition determination function 455d, a display control function 456d, and an output function 458. Here, for example, the processing functions executed by the acquisition function 457, the estimation function 453d, the learning function 454d, the shooting condition determination function 455d, the display control function 456d, and the output function 458, which are components of the processing circuitry 45d shown in FIG. 11, are recorded in the storage circuitry 41d in the form of computer-executable programs. The processing circuitry 45d is, for example, a processor that reads and executes each program from the storage circuitry 41d to realize the function corresponding to the read program. In other words, the processing circuitry 45d in a state in which each program has been read has the functions shown in the processing circuitry 45d of FIG. 11. The acquisition function 457 is an example of an acquisition unit. The estimation function 453d is an example of an estimation unit. The learning function 454d is an example of a learning unit. The photographing condition determination function 455d is an example of a photographing condition determination unit. The display control function 456d is an example of a display control unit. The output function 458 is an example of an output unit.

[0183] 11 illustrates a case where the acquisition function 457, estimation function 453d, learning function 454d, shooting condition determination function 455d, display control function 456d, and output function 458 are implemented by a single processing circuit 45d, but the embodiment is not limited to this. For example, the processing circuit 45d may be configured by combining multiple independent processors, and each processor may implement each processing function by executing a respective program. Furthermore, each processing function of the processing circuit 45d may be implemented by being appropriately distributed or integrated into a single or multiple processing circuits.

[0184] The imaging conditions in this embodiment may be changed depending on the type of the target modality 9.

[0185] The ultrasound images captured by the modality 9 are examples of the first medical image, the second medical image, and the training medical image in this embodiment.

[0186] The acquisition function 457 acquires subject information related to the subject P, including medical images of a diagnostic target region of the subject P, from the modality 9. Similarly to the transmission / reception function 452a of the first embodiment, the acquisition function 457 may acquire subject information related to the subject P other than medical images from the electronic medical record system 2 or other devices.

[0187] The processing by the estimation function 453d, the learning function 454d, the imaging condition determination function 455d, and the display control function 456d can be, for example, the same processing as in the first embodiment. Furthermore, depending on the type of modality 9, the same processing as in the second and third embodiments may be applied.

[0188] The output function 458 transmits the imaging conditions determined by the imaging condition determination function 455d to the modality 9. As a result, imaging based on the determined imaging conditions is performed by the modality 9. Note that the method of outputting the imaging conditions is not limited to this, and the display control function 456d may display the imaging conditions on the display 42d. In this case, the user may manually set the imaging conditions of the modality 9 according to the contents of the imaging conditions displayed on the display 42d.

[0189] As described above, in this embodiment, during an imaging diagnostic examination in which the modality 9 can perform two or more consecutive imaging operations, the workstation 1d, rather than the modality 9, can automatically determine appropriate imaging conditions for the next imaging operation according to the results of the previous imaging operation. According to the workstation 1d of this embodiment, the modality 9 itself does not need to have a function for determining imaging conditions, so that the modality 9 having an existing configuration can be utilized.

[0190] The medical image processing device is not limited to the workstation 1d, but may be any information processing device capable of communicating with the modality 9. For example, the medical image processing device may be various types of servers, a PC (Personal Computer), etc. Furthermore, the functions of the workstation 1d in this embodiment may be shared and executed by a plurality of PCs, etc.

[0191] In the first to fourth embodiments described above, the learning phase process and the operation phase process are executed by one device, but the learning phase process and the operation phase process may be executed by different devices. For example, the learning phase process may be executed by a server or a PC, and the operation phase process may be executed by various modalities or workstations.

[0192] Instead of or in addition to the deep learning or machine learning model in each of the above embodiments, a mathematical model, a lookup table, a database, or the like may be adopted.

[0193] The various data handled in this specification are typically digital data.

[0194] According to at least one of the embodiments described above, during an imaging diagnostic examination in which two or more shots can be taken consecutively, appropriate shooting conditions for the next shot can be automatically determined based on the results of the previous shot.

[0195] Although several 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 within the scope and spirit of the invention, as well as within the scope of the invention and its equivalents as defined in the claims. [Explanation of symbols]

[0196] 1a X-ray CT device 1b Magnetic resonance imaging device 1c Ultrasound diagnostic equipment 1d workstation 2. Electronic medical record system 9 Modalities 10 Mounting device 11 X-ray tube 12 X-ray detector 13 Rotating Frame 14 X-ray high voltage device 15 Control device 16 Wedge 17 Collimator 30 Bed Device 31 Foundation 32 Bed drive unit 33 Top plate 34 Top plate support frame 40 Console device 41a, 41b, 41c, 41d memory circuits 42a, 42b, 42c, 42d Display 43a, 43b, 43c, 43d Input interface 44a, 44b, 44c, 44d NW interface 45a, 45b, 45c, 45d Processing circuit 50, 50a~50c Shooting position 51,51a~51c CT images 60, 60a~60d Unfilmed parts 61, 61a~61d Estimated images 100 Ultrasound Probe 101 Static Magnetic Field Magnet 103 Gradient magnetic field coil 104 Gradient magnetic field power supply 105 berths 105a Top plate 106 Bed control circuit 107 Transmitting Coil 108 Transmitting circuit 109 Receiving Coil 110 Receiving circuit 120 Sequence control circuit 130 Computer Systems 400 Device body 401 Transmitting and receiving circuit 402 Signal Processing Circuit 403 Image Memory 451a, 451b, 451c Shooting function 452a, 452b, 452c Sending and receiving functions 453a, 453b, 453c, 453d Estimated Function 454a, 454b, 454c, 454d Learning Function 455a, 455b, 455c, 455d Shooting condition determination function 456a, 456b, 456c, 456d Display control function 457 Acquisition Function 458 Output Function N Network P Subject

Claims

1. an acquisition unit that acquires subject information regarding the subject, the subject information including a first medical image of a diagnostic target region of the subject captured under a first imaging condition; an imaging condition determination unit that determines second imaging conditions for capturing a second medical image of the subject based on the subject information and a trained model that has been trained by associating evaluation information related to medical images with imaging conditions; and A medical image diagnostic device comprising:

2. the second imaging condition includes information on whether or not the next imaging will be performed, and an imaging position of the subject when the next imaging will be performed. The medical image diagnostic apparatus according to claim 1 .

3. an estimation unit that calculates, based on the subject information, evaluation information regarding an estimated image that is estimated to be obtained by imaging the diagnostic target region under imaging conditions different from the first imaging condition; The evaluation information includes at least one of first evaluation information regarding image quality of the estimated image and second evaluation information regarding uncertainty of a disease classification estimated to be obtained when disease classification is performed based on the first medical image and the estimated image. The medical image diagnostic apparatus according to claim 1 .

4. the estimated image is a medical image that is estimated to be captured when an unphotographed portion of the diagnostic target region of the subject is photographed, the estimation unit determines the evaluation information based on the first medical image and the estimated image. The medical image diagnostic apparatus according to claim 3 .

5. Further, a learning unit is provided to generate the trained model by training a model, the estimation unit calculates, based on training subject information including a training medical image, training evaluation information related to a training estimated image that is estimated to be obtained by imaging a diagnostic target region under imaging conditions different from the training imaging conditions under which the training medical image was captured; the learning unit learns the model based on the learning imaging conditions, the learning object information, and the learning evaluation information so that the learning evaluation information is improved; the imaging condition determination unit determines the second imaging condition by inputting the object information and the first imaging condition to the trained model; The medical image diagnostic apparatus according to claim 3 .

6. an X-ray generating unit that generates X-rays; an X-ray detector for detecting the X-rays, the first imaging condition and the second imaging condition include imaging positions to which the X-rays are irradiated, the imaging position included in the second imaging condition is included in an unimaged portion of the diagnostic target region excluding the imaging position of the first medical image; The medical image diagnostic apparatus according to any one of claims 1 to 5.

7. the first medical image and the second medical image are magnetic resonance images; the first imaging condition and the second imaging condition include an imaging position and a type of image in magnetic resonance imaging; The medical image diagnostic apparatus according to any one of claims 1 to 5.

8. an ultrasound probe for transmitting ultrasound to the subject; the first imaging condition and the second imaging condition include information on a position and an angle of the ultrasound probe; The medical image diagnostic apparatus according to any one of claims 1 to 5.

9. an acquisition unit that acquires subject information regarding the subject, the subject information including a first medical image of a diagnostic target region of the subject captured under a first imaging condition; an imaging condition determination unit that determines second imaging conditions for capturing a second medical image of the subject based on the subject information and a trained model that has been trained by associating evaluation information related to medical images with imaging conditions; and an output unit that outputs the second photographing condition; A medical image processing device comprising:

10. an acquiring step of acquiring subject information regarding the subject, the subject information including a first medical image of a diagnostic target region of the subject captured under a first imaging condition; an imaging condition determination step of determining second imaging conditions for capturing a second medical image of the subject based on the subject information and a trained model that has been trained by associating evaluation information related to medical images with imaging conditions; A method comprising:

Images