Medical image processing device, X-ray CT device, medical image processing method and program

Abstract

To enhance diagnosability of medical image diagnosis.SOLUTION: A medical image processing device includes an acquisition unit and a specification unit. The acquisition unit acquires first medical image data. The specification unit specifies a region of interest in the first medical image data on the basis of a learned model and the first medical image data. The learned model undergoes training on the basis of second medical image data corresponding to the first medical image data, and third medical image data of a kind different from that of the first medical image data and related to the same subject as that of the second medical image data. At least a part of the third medical image data has higher imaging sensitivity on a region of the subject corresponding to the region of interest than the second medical image data.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The embodiments disclosed in this specification and the drawings relate to a medical image processing device, an X-ray CT device, a medical image processing method, and a program. [Background technology]

[0002] Conventionally, in medical image diagnosis using an X-ray computed tomography (CT) device, early CT signs (early ischemic signs) are sometimes used to detect, for example, hyperacute cerebral infarction, with the aim of early detection and appropriate treatment. Generally, from the viewpoint of workflow, an X-ray CT device enables faster image reconstruction than an MRI (Magnetic Resonance Imaging) device, and therefore requires less time for medical image diagnosis.

[0003] However, because detecting early CT signs requires high image interpretation skills, medical image diagnosis using early CT signs has the problem of a high false negative rate.In addition, medical image diagnosis based on diffusion-weighted images (DWI) obtained with MRI devices has a much higher detection ability for hyperacute cerebral infarction than when using X-ray CT devices, but has the problem of the long time required for medical image diagnosis. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2019-010411 Summary of the Invention [Problem to be solved by the invention]

[0005] One of the problems that the embodiments disclosed in this specification and the drawings aim to solve is to improve the diagnostic performance of medical image diagnosis. However, the problems that the embodiments disclosed in this specification and the drawings aim to solve are not limited to the above problem. Problems corresponding to the configurations shown in the embodiments described below can also be considered as other problems. [Means for solving the problem]

[0006] A medical image processing apparatus according to an embodiment includes an acquisition unit and an identification unit. The acquisition unit acquires first medical image data. The identification unit identifies a region of interest in the first medical image data based on a trained model and the first medical image data. The trained model is trained based on second medical image data corresponding to the first medical image data and third medical image data of a different type from the first medical image data but relating to the same subject as the second medical image data. At least a portion of the third medical image data has higher imaging sensitivity for the region of the subject corresponding to the region of interest than the second medical image data. [Brief explanation of the drawings]

[0007] [Figure 1] FIG. 1 is a diagram showing an example of the configuration of an X-ray computed tomography (CT) apparatus equipped with a medical image processing apparatus according to an embodiment. [Figure 2] FIG. 2 is a diagram illustrating learning of a raw database of a model according to the embodiment. [Figure 3] FIG. 3 is a diagram for explaining inference of a raw database using a model according to the embodiment. [Figure 4] FIG. 4 is a flowchart illustrating an example of a learning process of a raw database according to the embodiment. [Figure 5] FIG. 5 is a flowchart illustrating an example of inference processing of a raw database according to the embodiment. [Figure 6]FIG. 6 is a diagram for explaining image-based learning of a model according to the embodiment. [Figure 7] FIG. 7 is a diagram for explaining image-based reasoning using a model according to the embodiment. [Figure 8] FIG. 8 is a flowchart illustrating an example of an image-based learning process according to the embodiment. [Figure 9] FIG. 9 is a flowchart illustrating an example of image-based inference processing according to the embodiment. [Figure 10] FIG. 10 is a diagram for explaining a combination of raw database processing and image-based processing according to the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0008] Hereinafter, a medical image processing apparatus, an X-ray CT apparatus, a medical image processing method, and a program according to each embodiment will be described with reference to the drawings. In the following description, components having the same or substantially the same functions as those described above with reference to the previously mentioned figures will be given the same reference numerals and will be described only if necessary. Even when the same parts are shown, the dimensions and proportions may be different depending on the drawing. Furthermore, for example, in order to ensure the visibility of the drawings, reference numerals may be given to only the main or representative components in the description of each drawing, and reference numerals may not be given to components having the same or substantially the same functions.

[0009] In each of the embodiments described below, a case will be exemplified in which a medical image processing apparatus according to each embodiment is installed in an X-ray computed tomography (CT) apparatus.

[0010] The medical image processing apparatus according to each embodiment is not limited to being mounted on an X-ray CT apparatus, but may be realized as an independent apparatus by a computer having a processor such as a CPU (Central Processing Unit) and memories such as a ROM (Read Only Memory) and a RAM (Random Access Memory) as hardware resources. In this case, the processor mounted on the computer can realize various functions according to each embodiment by executing a program read from the ROM or the like and loaded into the RAM.

[0011] Furthermore, the medical image processing apparatus according to each embodiment may be implemented by being installed in a medical image diagnostic apparatus other than an X-ray CT apparatus. In this case, a processor installed in each medical image diagnostic apparatus can implement the functions according to each embodiment by executing a program read from a ROM or the like and loaded into a RAM. Various functions according to each embodiment can be implemented. Examples of other medical image diagnostic apparatuses include an X-ray diagnostic apparatus, an MRI (Magnetic Resonance Imaging) apparatus, an ultrasound diagnostic apparatus, a SPECT (Single Photon Emission Computed Tomography) apparatus, a PET (Positron Emission Computed Tomography) apparatus, a SPECT-CT apparatus in which a SPECT apparatus and an X-ray CT apparatus are integrated, and a PET-CT apparatus in which a PET apparatus and an X-ray CT apparatus are integrated.

[0012] For example, there are various types of X-ray CT devices, such as third-generation CT and fourth-generation CT, and any of these types can be applied to each embodiment. Here, the third-generation CT is a rotate / rotate-type in which the X-ray tube and detector rotate together around the subject. The fourth-generation CT is a stationary / rotate-type in which a large number of X-ray detection elements arranged in a ring shape are fixed, and only the X-ray tube rotates around the subject.

[0013] (First embodiment) 1 is a diagram showing an example of the configuration of an X-ray CT apparatus 1 equipped with a medical image processing apparatus according to an embodiment. The X-ray CT apparatus 1 irradiates an object P with X-rays from an X-ray tube 11 and detects the irradiated X-rays with an X-ray detector 12. The X-ray CT apparatus 1 generates a CT image of the object P based on an output from the X-ray detector 12.

[0014] As shown in FIG. 1, the X-ray CT apparatus 1 has a gantry 10, a bed 30, and a console 40. For convenience of explanation, a plurality of gantry 10 are depicted in FIG. 1. The gantry 10 is a scanning apparatus configured to perform X-ray CT imaging of a subject P. The bed 30 is a transport device on which the subject P to be subjected to X-ray CT imaging is placed and which positions the subject P. The console 40 is a computer that controls the gantry 10. For example, the gantry 10 and the bed 30 are installed in a CT examination room, and the console 40 is installed in a control room adjacent to the CT examination room. The gantry 10, the bed 30, and the console 40 are connected to each other by wire or wirelessly so that they can communicate with each other.

[0015] The console 40 does not necessarily have to be installed in a control room. For example, the console 40 may be installed in the same room as the gantry 10 and the bed 30. Alternatively, the console 40 may be incorporated into the gantry 10.

[0016] In this embodiment, the rotation axis of the rotating frame 13 in the non-tilted state or the longitudinal direction of the tabletop 33 of the bed 30 is defined as the Z-axis direction, the axis direction perpendicular to the Z-axis direction and horizontal to the floor surface is defined as the X-axis direction, and the axis direction perpendicular to the Z-axis direction and perpendicular to the floor surface is defined as the Y-axis direction.

[0017] As shown in FIG. 1, the gantry 10 includes an X-ray tube 11, an X-ray detector 12, a rotating frame 13, an X-ray high voltage device 14, a control device 15, a wedge 16, a collimator 17, and a data acquisition system (DAS) 18.

[0018] The X-ray tube 11 is a vacuum tube having a cathode (filament) that generates thermoelectrons and an anode (target) that generates X-rays upon impact of the thermoelectrons. The X-ray tube 11 irradiates the subject P with X-rays by irradiating the thermoelectrons from the cathode toward the anode using a high voltage supplied from an X-ray high voltage device 14.

[0019] It should be noted that the hardware for generating X-rays is not limited to the X-ray tube 11. For example, X-rays may be generated using a fifth-generation system instead of the X-ray tube 11. The fifth-generation system includes a focus coil that focuses the electron beam generated from the electron gun, a deflection coil that electromagnetically deflects the beam, and a target ring that surrounds half the circumference of the subject P and generates X-rays when the deflected electron beam collides with the target ring.

[0020] The X-ray detector 12 detects X-rays emitted from the X-ray tube 11 and passing through the subject P, and outputs an electrical signal corresponding to the detected X-ray dose to the DAS 18. The X-ray detector 12 has, for example, an X-ray detection element row in which a plurality of X-ray detection elements are arranged in the channel direction along an arc centered on the focal point of the X-ray tube 11. The X-ray detector 12 has, for example, a structure in which a plurality of X-ray detection elements are arranged in the slice direction (column direction, row direction) in the channel direction. The X-ray detector 12 is an indirect conversion type detector having, for example, a grid, a scintillator array, and a photosensor array. The scintillator array has a plurality of scintillators. The scintillator has scintillator crystals that output light with an amount of light corresponding to the dose of incident X-rays. The grid is arranged on the X-ray incident surface side of the scintillator array and has an X-ray shielding plate that has the function of absorbing scattered X-rays. The grid is sometimes called a collimator (one-dimensional collimator or two-dimensional collimator). The photosensor array has a function of converting light from the scintillator into an electrical signal corresponding to the amount of light. For example, a photomultiplier tube (PMT) is used as the photosensor. The X-ray detector 12 may be a direct conversion type detector having a semiconductor element that converts incident X-rays into an electrical signal. Here, the X-ray detector 12 is an example of a detection unit.

[0021] The rotating frame 13 is an annular frame that supports the X-ray tube 11 and the X-ray detector 12 so that they face 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. An image field of view (FOV) is set in an opening 19 of the rotating frame 13. For example, the rotating frame 13 is made of aluminum casting. Note that in addition to the X-ray tube 11 and the X-ray detector 12, the rotating frame 13 can also support an X-ray high voltage device 14, a wedge 16, a collimator 17, a DAS 18, and the like. The rotating frame 13 can also support various components not shown in FIG. 1 .

[0022] The X-ray high voltage device 14 has a high voltage generator and an X-ray control device. The high voltage generator has electrical circuits such as a transformer and a rectifier, and generates 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. The X-ray control device 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 in the gantry 10, or on a fixed frame (not shown) in the gantry 10. The fixed frame is a frame that rotatably supports the rotating frame 13.

[0023] The control device 15 includes a driving mechanism such as a motor and an actuator, and a processing circuit having a processor, memory, etc. that controls the driving mechanism. The control device 15 receives input signals from the input interface 43, an input interface provided on the gantry 10, etc., and controls the operation of the gantry 10 and the bed 30. Examples of operation control by the control device 15 include control to rotate the rotating frame 13, control to tilt the gantry 10, and control to operate the bed 30. Note that the control to tilt the gantry 10 is 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 through an input interface attached to the gantry 10. Note that the control device 15 may be provided on the gantry 10 or on the console 40.

[0024] The wedge 16 is a filter for adjusting the amount of X-rays irradiated from the X-ray tube 11. Specifically, the wedge 16 is a filter that transmits and attenuates the X-rays irradiated from the X-ray tube 11 so that the X-rays irradiated from the X-ray tube 11 to the subject P have a predetermined distribution. For example, the wedge 16 is a wedge filter or a bow-tie filter, and is made by processing aluminum or the like to have a predetermined target angle and a predetermined thickness.

[0025] The collimator 17 limits the irradiation range of the X-rays that have passed through the wedge 16. The collimator 17 slidably supports multiple lead plates that shield the X-rays, and adjusts the shape of the slits formed by the multiple lead plates. The collimator 17 is sometimes called an X-ray aperture.

[0026] The DAS 18 reads out from the X-ray detector 12 an electrical signal corresponding to the X-ray dose detected by the X-ray detector 12. The DAS 18 amplifies the read electrical signal and integrates (adds up) the electrical signal over a view period to collect detection data having a digital value corresponding to the X-ray dose over the view period. The detection data is called projection data. The DAS 18 is realized, for example, by an application specific integrated circuit (ASIC) equipped with circuit elements capable of generating projection data. The projection data is transmitted to the console 40 via a non-contact data transmission device or the like. Here, the DAS 18 is an example of a detection unit.

[0027] The detection data generated by the DAS 18 is transmitted by optical communication from a transmitter having a light emitting diode (LED) provided on the rotating frame 13 to a receiver having a photodiode provided on a non-rotating portion of the gantry 10 (for example, a fixed frame; not shown in FIG. 1), and then transferred to the console 40. The method of transmitting the detection data from the rotating frame 13 of the rotating portion to the non-rotating portion of the gantry 10 is not limited to the optical communication described above, and any method of non-contact data transfer may be used.

[0028] In this embodiment, an X-ray CT device 1 equipped with an integral type X-ray detector 12 will be described as an example, but the technology according to this embodiment can also be realized as an X-ray CT device 1 equipped with a photon counting type X-ray detector.

[0029] The bed 30 is a device on which the subject P to be scanned is placed and moved. The bed 30 includes a base 31, a bed drive device 32, a top plate 33, and a support frame 34. The base 31 is a housing that supports the support frame 34 so that it can move in the vertical direction. The bed drive device 32 is a drive mechanism that moves the top plate 33, on which the subject P is placed, in the longitudinal direction of the top plate 33. The bed drive device 32 includes a motor, an actuator, and the like. The top plate 33 is a plate on which the subject P is placed. The top plate 33 is provided on the upper surface of the support frame 34. The top plate 33 can protrude from the bed 30 toward the gantry 10 so that the entire body of the subject P can be imaged. The top plate 33 is made of, for example, carbon fiber reinforced plastic (CFRP), which has good X-ray transparency and physical properties such as rigidity and strength. Furthermore, the interior of the top plate 33 is hollow, for example. The support frame 34 supports the tabletop 33 so that the tabletop 33 can move in the longitudinal direction of the tabletop 33. The bed driving device 32 may move the support frame 34 in addition to the tabletop 33 in the longitudinal direction of the tabletop 33.

[0030] The console 40 has a memory 41, a display 42, an input interface 43, and a processing circuit 44. Data communication between the memory 41, the display 42, the input interface 43, and the processing circuit 44 is performed via a bus (BUS). Note that although the console 40 will be described as being separate from the gantry 10, the gantry 10 may include the console 40 or some of the components of the console 40.

[0031] The memory 41 is realized by, for example, a semiconductor memory element such as a ROM, a RAM, or a flash memory, a hard disk, an optical disk, or the like. For example, the memory 41 stores projection data and reconstructed image data. Also, for example, the memory 41 stores various programs. Also, for example, the memory 41 stores a model 100, which will be described later. The storage area of the memory 41 may be located within the X-ray CT device 1, or may be located in an external storage device connected via a network. Here, the memory 41 is an example of a storage unit.

[0032] The display 42 displays various types of information. For example, the display 42 displays medical images (CT images) generated by the processing circuitry 44, a GUI (Graphical User Interface) for receiving various operations from the operator, and the like. The information displayed on the display 42 includes images based on various types of medical image data, such as a mask image according to the embodiment and a CT image on which a mask image is superimposed. Any of a variety of displays can be used as the display 42, as appropriate. For example, a liquid crystal display (LCD), a cathode ray tube (CRT) display, an organic electroluminescence display (OLED), or a plasma display can be used as the display 42.

[0033] The display 42 may be installed anywhere in the control room. Alternatively, the display 42 may be installed on the pedestal 10. The display 42 may be a desktop type, or may be configured as a tablet terminal or the like capable of wireless communication with the main body of the console 40. Alternatively, one or more projectors may be used as the display 42. Here, the display 42 is an example of a display unit.

[0034] The input interface 43 accepts various input operations from the operator, converts the accepted input operations into electrical signals, and outputs the electrical signals to the processing circuitry 44. For example, the input interface 43 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. Furthermore, for example, the input interface 43 accepts from the operator operations related to setting a region of interest in DWI (diffusion weighted imaging), etc.

[0035] As the input interface 43, for example, a mouse, keyboard, trackball, switch, button, joystick, touchpad, touch panel display, etc. can be used as appropriate. Note that in this embodiment, the input interface 43 is not limited to one equipped with these physical operation components. For example, an example of the input interface 43 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 this electrical signal to the processing circuit 44. The input interface 43 may also be provided on the pedestal 10. The input interface 43 may also be configured as a tablet terminal or the like that is capable of wireless communication with the console 40 main body. Here, the input interface 43 is an example of an input unit.

[0036] The processing circuitry 44 controls the overall operation of the X-ray CT apparatus 1. The processing circuitry 44 has a processor and memories such as ROM and RAM as hardware resources. The processing circuitry 44 executes a system control function 441, an image generation function 442, an image processing function 443, a learning data generation function 444, a learning function 445, an inference function 446, a display control function 447, etc., by the processor that executes a program loaded into the memory. Here, the processing circuitry 44 is an example of a processing unit.

[0037] In the system control function 441, the processing circuitry 44 controls various functions of the processing circuitry 44 based on input operations received from an operator via the input interface 43. For example, the processing circuitry 44 controls a CT scan performed on the gantry 10. The processing circuitry 44 acquires detection data obtained by the CT scan. Note that the processing circuitry 44 may acquire detection data related to the subject P from outside the X-ray CT apparatus 1.

[0038] In the image generation function 442, the processing circuitry 44 generates 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. The processing circuitry 44 stores the generated data in the memory 41. Note that the data before preprocessing (detection data) and the data after preprocessing may also be collectively referred to as projection data. The processing circuitry 44 performs reconstruction processing using a filtered back projection method, an iterative reconstruction method, machine learning, or the like on the generated projection data (preprocessed projection data) to generate CT image data. The processing circuitry 44 stores the generated CT image data in the memory 41.

[0039] In this way, the system control function 441 and the image generation function 442 acquire projection data relating to the subject P. Here, the processing circuitry 44 that realizes the system control function 441 and the image generation function 442 is an example of an acquisition unit.

[0040] In the image processing function 443, the processing circuitry 44 converts the CT image data generated by the image generation function 442 into tomographic image data of an arbitrary cross section or three-dimensional image data using a known method, based on an input operation received from the operator via the input interface 43. For example, the processing circuitry 44 performs three-dimensional image processing such as volume rendering, surface rendering, image value projection processing, MPR (Multi-Planar Reconstruction) processing, and CPR (Curved MPR) processing on the CT image data to generate rendered image data of an arbitrary viewpoint direction. Note that the generation of three-dimensional image data such as rendered image data of an arbitrary viewpoint direction, i.e., volume data, may be performed directly by the image generation function 442. The processing circuitry 44 stores the tomographic image data and three-dimensional image data in the memory 41.

[0041] In the training data generation function 444, the processing circuitry 44 performs a generation process to generate training data for the model 100. The generation process of the training data for the model 100 will be described later. Here, the processing circuitry 44 that realizes the training data generation function 444 is an example of a training data generation unit.

[0042] In the learning function 445, the processing circuitry 44 performs a learning process to train the model 100 using the learning data set generated by the learning data generation function 444. The learning process of the model 100 will be described later. Here, the processing circuitry 44 that realizes the learning function 445 is an example of a learning unit.

[0043] In the inference function 446, the processing circuitry 44 performs an identification process to identify a region of interest in the projection data based on the trained model 100 and the projection data. The process of identifying a region of interest will be described later. Here, the processing circuitry 44 that realizes the inference function 446 is an example of an identification unit.

[0044] In the display control function 447, the processing circuitry 44 causes the display 42 to display images based on various image data generated by the image processing function 443. The images displayed on the display 42 include reconstructed images as display images in which a specified region of interest is highlighted on the image. The display may be performed by superimposing a mask image indicating the region of interest on the reconstructed image, or may be performed based on reconstructed image data for display on which the mask image data is combined. The images displayed on the display 42 include CT images based on CT image data, cross-sectional images based on cross-sectional image data of an arbitrary cross section, and rendered images of an arbitrary viewpoint based on rendered image data of an arbitrary viewpoint. The images displayed on the display 42 include images for displaying an operation screen and images for displaying notifications and warnings to the operator. The processing circuitry 44, which realizes the display control function 447, is an example of a display control unit.

[0045] Note that image data showing a display image in which the identified region of interest is highlighted on the image may be generated by either the image processing function 443 or the display control function 447.

[0046] Note that the functions 441 to 447 are not limited to being realized by a single processing circuit. The processing circuit 44 may be configured by combining multiple independent processors, and each processor may execute a program to realize each of the functions 441 to 447. Here, the functions 441 to 447 may be realized by being distributed or integrated as appropriate across a single or multiple processing circuits.

[0047] Although the console 40 has been described as a single console that executes multiple functions, multiple functions may be executed by separate consoles. For example, the functions of the processing circuit 44, such as the image generation function 442, image processing function 443, learning data generation function 444, learning function 445, and inference function 446, may be distributed.

[0048] Note that part or all of the processing circuitry 44 is not limited to being included in the console 40, but may be included in an integrated server that collectively processes detection data acquired by a plurality of medical image diagnostic devices.

[0049] At least one of the post-processing, generation processing, learning processing, identification processing, and display processing may be performed by either the console 40 or an external workstation. Furthermore, the processing may be performed simultaneously by both the console 40 and the workstation. As the workstation, for example, a computer having a processor that realizes the functions corresponding to each process and memories such as ROM and RAM as hardware resources can be appropriately used.

[0050] In addition, in the reconstruction of X-ray CT image data, either a full scan reconstruction method or a half scan reconstruction method may be applied. For example, in the image generation function 442, the processing circuitry 44 uses projection data for 360 degrees, that is, one circumference around the subject P, in the full scan reconstruction method. In addition, in the half scan reconstruction method, the processing circuitry 44 uses projection data for 180 degrees + fan angle. In the following, for simplicity of explanation, it is assumed that the processing circuitry 44 uses the full scan reconstruction method, which performs reconstruction using projection data for 360 degrees, that is, one circumference around the subject P.

[0051] The technology according to this embodiment can be applied to both single-tube X-ray computed tomography apparatuses and so-called multi-tube X-ray computed tomography apparatuses in which multiple pairs of X-ray tubes and detectors are mounted on a rotating ring.

[0052] The technology according to this embodiment can also be applied to an X-ray CT apparatus 1 configured to be capable of imaging using a dual energy method. In this case, the X-ray high voltage device 14 can alternately switch the energy spectrum of the X-rays emitted from the X-ray tube 11, for example, by high-speed switching between two voltage values. In other words, the X-ray CT apparatus 1 is configured to acquire projection data for each acquisition view while modulating the tube voltage at a timing according to a control signal for tube voltage modulation. By imaging the subject with different tube voltages, it is possible to improve the contrast of shading in the CT image based on the energy transmittance of materials for each X-ray energy spectrum.

[0053] It is assumed that the X-ray CT apparatus 1 according to this embodiment is configured to read out electrical signals from the X-ray detector 12 by a sequential readout method.

[0054] The X-ray CT apparatus 1 according to this embodiment may be configured as a standing CT. In this case, instead of moving the tabletop 33, a patient support mechanism may be provided that supports the subject P in a standing position and is configured to be movable along the rotation axis of the rotating part of the gantry 10. The X-ray CT apparatus 1 according to this embodiment may also be configured as a mobile CT in which the gantry 10 and the bed 30 are movable.

[0055] Here, the process of generating learning data for the model 100 according to the embodiment and the learning process of the model 100 will be described in more detail with reference to the drawings. Fig. 2 is a diagram for explaining learning of the raw database of the model 100 according to the embodiment.

[0056] The processing circuitry 44 acquires, for example, from the memory 41, the projection data RCT acquired by the X-ray CT device 1 and the data of the CT image ICT acquired by image reconstruction based on the projection data RCT. The processing circuitry 44 also acquires, for example, from the memory 41, the data of the diffusion-weighted image IDWI acquired by the MRI device. The processing circuitry 44 may acquire the data of the diffusion-weighted image IDWI from the MRI device or the PACS via a network such as an in-hospital network. The projection data RCT and the data of the diffusion-weighted image IDWI are assumed to be data relating to the same region of the same subject. Here, the projection data RCT is an example of second medical image data. The data of the diffusion-weighted image IDWI is an example of third medical image data. The diffusion-weighted image IDWI is an image with significantly higher detectability for hyperacute cerebral infarction than the CT image ICT.

[0057] The processing circuitry 44 identifies an ischemic region R0 in the diffusion-weighted image IDWI based on, for example, an operator's input received via the input interface 43. The processing circuitry 44 aligns the diffusion-weighted image IDWI with the CT image ICT. The processing circuitry 44 also generates a mask image IM by extracting a region R1 corresponding to the ischemic region R0 on the CT image ICT. Here, the region R1 is an example of a region of interest.

[0058] The processing circuitry 44 retains the CT values of the CT image ICT for the region R1 in the mask image IM. On the other hand, the processing circuitry 44 discards the CT values of the CT image ICT for the region R2 other than the region R1 in the mask image IM. Therefore, in the mask image IM, the CT values exist only in the extracted region R1.

[0059] In the mask image IM, the processing circuitry 44 may normalize the CT value of the CT image ICT for the region R1 or may assign the value "1." In either case, the CT value of the CT image ICT for the region R2 is discarded and assigned the value "0."

[0060] The processing circuitry 44 generates a sinogram RM by forward projection based on a mask image IM that takes into account the geometry of the X-ray CT apparatus 1 and the physical characteristics of X-rays. As shown in Fig. 2, in the sinogram RM, values exist in the region of interest corresponding to the region R1.

[0061] The processing circuitry 44 performs a learning process to train the model 100 using the projection data RCT and the generated sinogram RM. Specifically, the parameters of the model 100 are determined so that a sinogram having values present in the ischemic region R1, i.e., the region of interest, is output in response to the input of the projection data. In other words, the trained model 100 is a parameterized composite function obtained by combining multiple functions, which receives the projection data as input and outputs a sinogram showing the region of interest in the projection data.

[0062] The parameterized composite function is defined by a combination of multiple adjustable functions and parameters. This model 100 may be any parameterized composite function that satisfies the above requirements, but is assumed to be a multi-layer network model. One example of the model 100 is a deep neural network (DNN). This DNN may have any structure, and examples of such DNNs include Residual Network (ResNet), Dense Convolutional Network (DenseNet), and U-Net. The configuration of the model 100 and the initial parameter values may be predetermined and stored in the memory 41, for example.

[0063] The model 100 can be prepared for each lesion, each type of medical image diagnostic device, and each part of the subject, for example.

[0064] Next, the process of identifying a region of interest according to the embodiment will be described in more detail with reference to the drawings. Fig. 3 is a diagram for explaining inference of a raw database by the model 100 according to the embodiment.

[0065] The processing circuitry 44 acquires the projection data RCT obtained by the X-ray CT apparatus 1 and the data of the CT image ICT obtained by image reconstruction based on the projection data RCT, for example, from the memory 41. Here, the projection data RCT is an example of first medical image data.

[0066] The processing circuitry 44 inputs the projection data RCT into the trained model 100. The processing circuitry 44 acquires a sinogram RM from the trained model 100 in response to the input of the projection data RCT. The processing circuitry 44 generates a mask image IM by image reconstruction based on the acquired sinogram RM.

[0067] The processing circuitry 44 superimposes the mask image IM on the CT image ICT and displays the CT image ICT' in which the ischemic region R0' is highlighted on the display 42. The processing circuitry 44 highlights the ischemic region R0' in the CT image ICT' by changing the color of the ischemic region R0' from that of the other regions.

[0068] FIG. 4 is a flowchart illustrating an example of a learning process of a raw database according to the embodiment.

[0069] First, the learning data generation function 444 acquires CT image data and DWI image data relating to hyperacute cerebral infarction of the same subject (S101).

[0070] The training data generation function 444 generates a mask image (S102). Specifically, the training data generation function 444 identifies an ischemic region in the DWI image based on, for example, an input result from the operator received via the input interface 43. The training data generation function 444 also extracts a region of the CT image corresponding to the identified ischemic region. The training data generation function 444 also generates a mask image by deleting CT values corresponding to regions other than the extracted region of the CT image data. Note that deleting CT values means padding with the value "0". Note that the training data generation function 444 may normalize the CT values of the extracted region of the CT image data or may pad with the value "1".

[0071] The training data generation function 444 generates a sinogram by forward projection based on the mask image (S103). This sinogram can also be expressed as an infarction weight that indicates the distribution of ischemic regions, i.e., infarctions.

[0072] The learning data generation function 444 generates a set of projection data (raw data) corresponding to the CT image and the generated sinogram as learning data (S104).

[0073] Thereafter, the learning function 445 performs a learning process to train the model 100 using the generated learning data set (S105).

[0074] FIG. 5 is a flowchart illustrating an example of inference processing of a raw database according to the embodiment.

[0075] The inference function 446 inputs projection data (raw data) to the trained model 100 (S201). In addition, the inference function 446 acquires a sinogram output from the trained model 100 in response to the input of the projection data (S202).

[0076] The image generation function 442 generates mask image data by image reconstruction based on the acquired sinogram (S203). The image generation function 442 also generates CT image data by image reconstruction based on the projection data (S204). The display control function 447 superimposes the mask image on the CT image and displays it on the display 42 (S205).

[0077] As described above, in the X-ray CT apparatus 1 equipped with the medical image processing apparatus according to the embodiment, the processing circuitry 44 is configured to be able to realize a system control function 441, an image generation function 442, a learning function 445, and an inference function 446. In the system control function 441 and the image generation function 442, the processing circuitry 44 acquires projection data of the subject P. The learning function 445 trains the model 100 based on the projection data and diffusion-weighted image data related to the same subject. The inference function 446 identifies a region of interest in the projection data based on the trained model 100 and the projection data of the subject P.

[0078] According to this configuration, it is possible to detect hyperacute cerebral infarction with accuracy comparable to that of MRI images, which have higher imaging sensitivity than CT images, while maintaining the speed of CT imaging in the X-ray CT device 1. In other words, the technology according to the embodiment can improve the diagnostic performance of medical image diagnosis.

[0079] Furthermore, in the X-ray CT apparatus 1 equipped with the medical image processing apparatus according to the embodiment, the processing circuitry 44 is configured to further realize a display control function 447. The display control function 447 superimposes a mask image on the CT image and displays it on the display 42. With this configuration, the X-ray CT apparatus 1 can display a CT image in which the ischemic region is highlighted by changing the color, for example. In other words, the technology according to the embodiment can improve the diagnostic performance of medical image diagnosis.

[0080] (Second embodiment) In this embodiment, differences from the first embodiment will be mainly described.

[0081] FIG. 6 is a diagram for explaining image-based learning of the model 100 according to the embodiment.

[0082] The processing circuitry 44 acquires data of a CT image ICT obtained by the X-ray CT device 1, for example, from the memory 41. The processing circuitry 44 also acquires data of a diffusion-weighted image IDWI obtained by an MRI device, for example, from the memory 41. Note that the data of the CT image ICT and the data of the diffusion-weighted image IDWI are assumed to be data relating to the same region of the same subject. Here, the data of the CT image ICT is an example of second medical image data. The data of the diffusion-weighted image IDWI is an example of third medical image data.

[0083] The processing circuitry 44 generates data for a mask image IM by extracting a region R1 on the CT image ICT that corresponds to the region R0 of ischemia in the diffusion-weighted image IDWI.

[0084] The processing circuitry 44 performs a learning process using the data of the CT image ICT and the data of the generated mask image IM to train the model 100. Specifically, the parameters of the model 100 are determined in response to the input of CT image data so as to output a mask image in which values exist in the ischemic region R1, i.e., the region of interest.

[0085] FIG. 7 is a diagram for explaining image-based inference using the model 100 according to the embodiment.

[0086] The processing circuitry 44 acquires data of the CT image ICT obtained by the X-ray CT apparatus 1, for example, from the memory 41. Here, the data of the CT image ICT is an example of first medical image data.

[0087] The processing circuitry 44 inputs the data of the CT image ICT to the trained model 100. The processing circuitry 44 acquires data of the mask image IM from the trained model 100 in response to the input of the data of the CT image ICT.

[0088] The processing circuitry 44 superimposes the mask image IM on the CT image ICT, and displays the CT image ICT' on the display 42, in which the ischemic region R0' is highlighted.

[0089] FIG. 8 is a flowchart illustrating an example of an image-based learning process according to the embodiment.

[0090] The training data generation function 444 acquires CT image data and DWI image data related to hyperacute cerebral infarction of the same subject (S301), similar to S101 in Fig. 4. The training data generation function 444 also generates a mask image (S302), similar to S102 in Fig. 4.

[0091] The training data generation function 444 generates a set of CT image data and mask image data as training data (S303). Thereafter, the training function 445 performs a training process to train the model 100 using the generated training data set (S304).

[0092] FIG. 9 is a flowchart illustrating an example of image-based inference processing according to the embodiment.

[0093] The inference function 446 inputs CT image data to the trained model 100 (S401). The inference function 446 also acquires mask image data output from the trained model 100 in response to the input of the CT image data (S402). The display control function 447 superimposes a mask image on the CT image and displays it on the display 42, similar to S205 in Fig. 5 (S403).

[0094] In this way, in the X-ray CT apparatus 1 equipped with the medical image processing apparatus according to the embodiment, for example, learning processing and inference processing are performed on an image basis, whereas in the first embodiment, they are performed on a raw database. Even with this configuration, the same effects as in the first embodiment can be obtained.

[0095] (Third embodiment) The raw database processing according to the first embodiment and the image-based processing according to the second embodiment can be combined. Fig. 10 is a diagram for explaining the combination of the raw database processing and the image-based processing according to the embodiments.

[0096] Assume that an ischemic region R11 is estimated based on the raw database inference process according to the first embodiment, as shown in Fig. 10. On the other hand, assume that an ischemic region R12 is estimated based on the image-based inference process according to the second embodiment, as shown in Fig. 10.

[0097] For example, in the case of hyperacute cerebral infarction, early detection and appropriate treatment are required, so it is preferable to prevent false negatives such as overlooking early CT signs (early ischemic signs).

[0098] Therefore, in the X-ray CT apparatus 1 equipped with the medical image processing apparatus according to this embodiment, the region R1 of ischemia is identified as the logical sum of the region R11 of ischemia estimated based on the inference processing of the raw database and the region R12 of ischemia estimated based on the image-based inference processing. In other words, the mask image IM according to this embodiment includes the region R1 identified to include at least one of the region R11 of ischemia estimated based on the inference processing of the raw database and the region R12 of ischemia estimated based on the image-based inference processing.

[0099] This configuration has the effect of further reducing the occurrence of false negatives compared to the above-described embodiments.

[0100] In the above-described embodiments, the learning process and the inference process are performed based on DWI image data obtained by an MRI device, using cerebral infarction as an example. However, the present invention is not limited to this. For example, instead of the DWI image data, image data obtained by various medical image diagnostic devices can be used as appropriate. The image data used instead of the DWI image data may be any image data that can be mapped to a CT image. As an example, instead of the DWI image data, image data obtained by a medical image diagnostic device such as a PET device, a SPECT device, an X-ray device, an angiography device, an ultrasound diagnostic device, or a thermography device can also be used.

[0101] Furthermore, the learning process and inference process according to each of the above-described embodiments may be performed based on image data that is different from the image data in at least one of the scan conditions used for imaging, the applied reconstruction conditions, and the applied image processing. That is, instead of DWI image data, CT image data that is different from the image data in at least one of the scan conditions used for imaging, the applied reconstruction conditions, and the applied image processing may be used.

[0102] It is preferable that the projection data RCT as an example of the second medical image data be the same type of image data as the projection data RCT as an example of the first medical image data. However, the projection data RCT as an example of the first medical image data and the projection data RCT as an example of the second medical image data may be image data that differ from each other in at least one type of, for example, the scan conditions at the time of imaging, the applied reconstruction conditions, and the applied image processing. Similarly, it is preferable that the data of the CT image ICT as an example of the second medical image data be the same type of image data as the data of the CT image ICT as an example of the first medical image data. However, the data of the CT image ICT as an example of the first medical image data and the data of the CT image ICT as an example of the second medical image data may be image data that differ from each other in at least one type of, for example, the scan conditions at the time of imaging, the applied reconstruction conditions, and the applied image processing.

[0103] The term "processor" used in the above description refers to circuits such as a CPU, a graphics processing unit (GPU), an ASIC, and a programmable logic device (PLD). PLDs include simple programmable logic devices (SPLDs), complex programmable logic devices (CPLDs), and field programmable gate arrays (FPGAs). A processor realizes its functions by reading and executing a program stored in a memory circuit. The memory circuit storing the program is a computer-readable non-transitory recording medium. Note that instead of storing a program in a memory circuit, the program may be directly embedded in the processor circuit. In this case, the processor realizes its functions by reading and executing the program embedded in the circuit. Alternatively, instead of executing a program, the function corresponding to the program may be realized by combining logic circuits. Note that 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 to realize its functions. Furthermore, the functions of the multiple components in FIG. 1 may be realized by integrating them into a single processor.

[0104] According to at least one of the embodiments described above, the diagnostic performance of medical image diagnosis can be improved.

[0105] Although several embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be embodied in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are intended to be included within the scope and spirit of the invention, as well as within the scope of the inventions and their equivalents as defined in the claims. [Explanation of symbols]

[0106] 1 X-ray CT device 10 Mounting stand 11 X-ray tube 12 X-ray detector 13 Rotating Frame 14 X-ray high voltage device 15 Control device 16 Wedge 17 Collimator 18 Data Collection Circuit 19 Opening 30 berths 31 Foundation 32 Bed drive unit 33 Top plate 34 Support frame 40 Console 41 memory 42 Display (display unit) 43 Input Interface 44 Processing circuit 441 System Control Function (Acquisition Section) 442 Image generation function (acquisition part) 443 Image Processing Function 444 Learning data generation function (learning data generation part) 445 Learning Function 446 Inference function (specific part) 447 Display control function (display control unit)

Claims

1. an acquisition unit that acquires first medical image data; an identification unit that inputs the first medical image data into a trained model corresponding to a data type of the first medical image data, and identifies the region of interest in the first medical image data based on data indicating the region of interest output from the trained model in response to the input of the first medical image data; Equipped with the first medical image data is one of data types of projection data and reconstructed image data generated by image reconstruction based on the projection data, The trained model is prepared for any one of the data types of the first medical image data, second medical image data of a common data type with the first medical image data, the second medical image data including a region of hyperacute cerebral infarction; data generated based on third medical image data, which is a different data type from the first medical image data and includes a region of hyperacute cerebral infarction relating to the same part of the same subject as the second medical image data; is trained to output data indicating the region of interest in response to input of the first medical image data using the set of the third medical image data is diffusion-weighted image data; the data generated based on the third medical image data is mask image data in which CT values of regions other than the region of interest identified on the image of the second medical image data, which correspond to the region of hyperacute cerebral infarction on the image of the third medical image data, have been deleted, or projection data generated by forward projection based on the mask image; The data indicating the region of interest output from the trained model is data of a common data type as the data generated based on the third medical image data. Medical imaging equipment.

2. the first medical image data is the reconstructed image data, a generating unit that generates a display image in which the identified region of interest is highlighted on a reconstructed image represented by the first medical image data; The medical image processing device according to claim 1 .

3. the data generated based on the third medical image data is data of the mask image in which CT values of regions other than the region of interest identified on the image of the second medical image data are deleted; The data indicating the region of interest output from the trained model is data of a mask image in which CT values of regions other than the region of interest identified on the image of the reconstructed image indicated by the first medical image data are deleted, The display image is an image in which the mask image output from the trained model is superimposed on the reconstructed image indicated by the first medical image data. The medical image processing device according to claim 2 .

4. the first medical image data is the projection data, a generating unit that generates a display image in which the identified region of interest is highlighted on a reconstructed image obtained by image reconstruction based on the first medical image data; The medical image processing device according to claim 1 .

5. the data generated based on the third medical image data is the projection data generated by forward projection based on the mask image, from which CT values of regions other than the region of interest identified on the image of a reconstructed image generated by image reconstruction based on the second medical image data have been deleted; The data indicating the region of interest output from the trained model is projection data that generates, by image reconstruction, a mask image in which CT values of regions other than the region of interest identified on an image of a reconstructed image generated by image reconstruction based on the first medical image data are deleted, The display image is an image in which the mask image based on the projection data output from the trained model is superimposed on the reconstructed image based on the first medical image data. The medical image processing device according to claim 4 .

6. An acquisition unit that acquires first medical image data; an identification unit that inputs the first medical image data into a trained model corresponding to a data type of the first medical image data, and identifies the region of interest in the first medical image data based on data indicating the region of interest output from the trained model in response to the input of the first medical image data; Equipped with the first medical image data is data including projection data and reconstructed image data generated by image reconstruction based on the projection data, The trained model includes a first trained model corresponding to the input of the projection data and a second trained model corresponding to the input of the reconstructed image data, second medical image data of a common data type with the first medical image data, the second medical image data including a region of hyperacute cerebral infarction; data generated based on third medical image data, which is a different data type from the first medical image data and includes a region of hyperacute cerebral infarction relating to the same part of the same subject as the second medical image data; is trained to output data indicating the region of interest in response to input of the first medical image data using the set of the third medical image data is diffusion-weighted image data; the data generated based on the third medical image data is mask image data in which CT values of regions other than the region of interest identified on the image of the second medical image data, which correspond to the region of hyperacute cerebral infarction on the image of the third medical image data, have been deleted, or projection data generated by forward projection based on the mask image; the data indicating the region of interest output from the trained model is data of a common data type as data generated based on the third medical image data, The identification unit Identifying a first region of interest in a reconstructed image generated by image reconstruction based on the projection data based on the first trained model and the projection data; Identifying a second region of interest in a reconstructed image represented by the reconstructed image data based on the second trained model and the reconstructed image data; identifying a region including at least one of the first region of interest and the second region of interest as the region of interest; Medical imaging equipment.

7. an X-ray tube that generates X-rays; a detector that detects X-rays from the X-ray tube that have passed through a subject and generates projection data; an identification unit that inputs first medical image data, which is the projection data or reconstructed image data obtained by image reconstruction based on the projection data, into a trained model corresponding to a data type of the first medical image data, and identifies the region of interest in the first medical image data based on data indicating the region of interest output from the trained model in response to the input of the first medical image data; Equipped with the first medical image data is one of data types of projection data and reconstructed image data generated by image reconstruction based on the projection data, The trained model is prepared for any one of the data types of the first medical image data, second medical image data of a common data type with the first medical image data, the second medical image data including a region of hyperacute cerebral infarction; data generated based on third medical image data, which is a different data type from the first medical image data and includes a region of hyperacute cerebral infarction relating to the same part of the same subject as the second medical image data; is trained to output data indicating the region of interest in response to input of the first medical image data using the set of the third medical image data is diffusion-weighted image data; the data generated based on the third medical image data is mask image data in which CT values of regions other than the region of interest identified on the image of the second medical image data, which correspond to the region of hyperacute cerebral infarction on the image of the third medical image data, have been deleted, or projection data generated by forward projection based on the mask image; The data indicating the region of interest output from the trained model is data of a common data type as the data generated based on the third medical image data. X-ray CT device.

8. acquiring first medical image data; inputting the first medical image data into a trained model corresponding to a data type of the first medical image data, and identifying the region of interest in the first medical image data based on data indicating the region of interest output from the trained model in response to the input of the first medical image data; Including, the first medical image data is one of data types of projection data and reconstructed image data generated by image reconstruction based on the projection data, The trained model is prepared for any one of the data types of the first medical image data, second medical image data of a common data type with the first medical image data, the second medical image data including a region of hyperacute cerebral infarction; data generated based on third medical image data, which is a different data type from the first medical image data and includes a region of hyperacute cerebral infarction relating to the same part of the same subject as the second medical image data; is trained to output data indicating the region of interest in response to input of the first medical image data using the set of the third medical image data is diffusion-weighted image data; the data generated based on the third medical image data is mask image data in which CT values of regions other than the region of interest identified on the image of the second medical image data, which correspond to the region of hyperacute cerebral infarction on the image of the third medical image data, have been deleted, or projection data generated by forward projection based on the mask image; The data indicating the region of interest output from the trained model is data of a common data type as the data generated based on the third medical image data. Medical image processing methods.

9. acquiring first medical image data; inputting the first medical image data into a trained model corresponding to a data type of the first medical image data, and identifying the region of interest in the first medical image data based on data indicating the region of interest output from the trained model in response to the input of the first medical image data; A program for causing a computer to execute the above, the first medical image data is one of data types of projection data and reconstructed image data generated by image reconstruction based on the projection data, The trained model is prepared for any one of the data types of the first medical image data, second medical image data of a common data type with the first medical image data, the second medical image data including a region of hyperacute cerebral infarction; data generated based on third medical image data, which is a different data type from the first medical image data and includes a region of hyperacute cerebral infarction relating to the same part of the same subject as the second medical image data; is trained to output data indicating the region of interest in response to input of the first medical image data using the set of the third medical image data is diffusion-weighted image data; the data generated based on the third medical image data is mask image data in which CT values of regions other than the region of interest identified on the image of the second medical image data, which correspond to the region of hyperacute cerebral infarction on the image of the third medical image data, have been deleted, or projection data generated by forward projection based on the mask image; The data indicating the region of interest output from the trained model is data of a common data type as the data generated based on the third medical image data. program.

Images