JP2021074378A - Medical processing system and program - Google Patents

Abstract

To secure an image quality that does not hinder an image reading while reducing the data size of data stored in a storage unit.SOLUTION: A medical processing system according to an embodiment comprises a storage unit and an image generation unit. The storage unit stores partial data being a portion of first data corresponding to a first resolution collected by executing scanning on a subject, and stores second data corresponding to a second resolution that is lower than the first resolution acquired by executing down sampling on at least a portion of the first data. The image generation unit generates image data by using a learned model generating third data corresponding to the first resolution on the basis of the second data when receiving a request for generating the image data on the basis of the second data, and generates image data without using the learned model when receiving a request for generating the image data on the basis of the partial data.SELECTED DRAWING: Figure 6

Description

Embodiments of the present invention relate to medical processing systems and programs.

The X-ray CT (Computed Tomography) apparatus collects raw data of a subject and stores the raw data in a storage circuit such as a memory. Then, the X-ray CT device generates CT image data from the raw data stored in the storage circuit in response to an instruction from the user, generates image data for display from the CT image data, and image data for display. The image based on is displayed on the display.

Therefore, in order to suppress the decrease in the free space of the storage circuit, it is desired that the data size of the raw data stored in the storage circuit is small. For example, in an X-ray CT apparatus (ultra-high-definition CT apparatus) capable of collecting high-resolution raw data and generating high-resolution CT image data, the data size of the collected raw data tends to increase. is there. Therefore, in particular, in an ultra-high-definition CT apparatus, it is desired to secure free space in a storage circuit. However, simply reducing the data size of the raw data stored in the storage circuit by performing downsampling or the like on the raw data will lower the spatial resolution and resolution of the CT image data.

Japanese Patent Publication No. 2004-518312

An object to be solved by the present invention is to secure image quality that does not hinder image interpretation while reducing the data size of data stored in a storage unit (storage circuit).

The medical processing system of the embodiment includes a storage unit and an image generation unit. The storage unit stores partial data that is a part of the first data corresponding to the first resolution collected by performing a scan on the subject, and for at least a part of the first data. The second data corresponding to the second resolution lower than the first resolution, which is acquired by executing the downsampling, is stored. When the image generator receives a request to generate image data based on the second data, the image generator uses a trained model that generates the third data corresponding to the first resolution based on the second data. When the request to generate the image data based on the partial data is received, the image data is generated without using the trained model.

FIG. 1 is a diagram showing an example of a system configuration according to the first embodiment. FIG. 2 is a diagram for explaining an example of a method of generating a trained model at the time of learning of the first embodiment. FIG. 3 is a diagram showing an example of raw data composed of a plurality of data arranged along the channel direction and the column direction in the first embodiment. FIG. 4 is a diagram showing an example of raw data composed of a plurality of data arranged along the channel direction and the view direction in the first embodiment. FIG. 5 is a diagram showing an example of raw data having a second spatial resolution according to the first embodiment. FIG. 6 is a diagram showing an example of the configuration of the X-ray CT apparatus according to the first embodiment. FIG. 7 is a diagram for explaining an example of processing executed by the image generation function according to the first embodiment. FIG. 8 is a flowchart showing an example of a processing flow executed by the processing circuit according to the first embodiment. FIG. 9 is a diagram showing an example of an image according to the first embodiment. FIG. 10 is a diagram for explaining an example of the process of step S103 of the first embodiment. FIG. 11 is a diagram for explaining an example of processing executed by the trained model generator according to the first modification of the first embodiment. FIG. 12 is a diagram for explaining an example of a method of generating a trained model at the time of learning of the modified example 1 of the first embodiment. FIG. 13 is a flowchart showing an example of a processing flow executed by the processing circuit according to the second modification of the first embodiment. FIG. 14 is a diagram showing an example of the configuration of the X-ray CT apparatus according to the second embodiment. FIG. 15 is a diagram for explaining an example of processing executed by the trained model update function according to the second embodiment. FIG. 16 is a diagram for explaining an example of processing executed by the trained model update function according to the second embodiment. FIG. 17 is a diagram showing an example of the stored contents of the memory according to the third embodiment. FIG. 18 is a flowchart showing an example of a processing flow executed by the processing circuit according to the third embodiment.

Hereinafter, embodiments of the medical processing system and the program will be described in detail with reference to the drawings. The contents described in one embodiment or modification may be similarly applied to other embodiments or other modifications.

(First Embodiment)
FIG. 1 is a diagram showing an example of the configuration of the system 100 according to the first embodiment. The system 100 shown in FIG. 1 includes an X-ray CT (Computed Tomography) device 1 and a trained model generation device 2. In the system 100, the X-ray CT apparatus 1 performs various processes using the trained model generated by the trained model generator 2.

The X-ray CT apparatus 1 and the trained model generator 2 are connected to each other so as to be able to communicate with each other via the network 3. The X-ray CT apparatus 1 collects raw data of a subject and generates CT image data using the collected raw data. Further, the X-ray CT apparatus 1 may transmit raw data to the trained model generator 2 in order to cause the trained model generator 2 to generate the trained model. Details of the X-ray CT apparatus 1 will be described later.

The trained model generation device 2 is connected to a plurality of other X-ray CT devices (not shown) in addition to the X-ray CT device 1 via a network 3. The trained model generator 2 receives raw data of various subjects transmitted from the X-ray CT apparatus 1 and a plurality of other X-ray CT apparatus. The raw data received by the trained model generator 2 is used in generating the trained model. The trained model generator 2 is realized by a server, a PC (Personal Computer), or the like.

The trained model generator 2 includes a processing circuit 2a. The processing circuit 2a is realized by, for example, a processor. The processing circuit 2a includes a trained model generation function 2a_1. The trained model generation function 2a_1 performs a process at the time of learning to generate a trained model. Here, when the processing circuit 2a is realized by the processor, the trained model generation function 2a_1 of the processing circuit 2a is in the form of a program that can be executed by a computer, such as a memory included in the trained model generation device 2. It is stored in a storage circuit (not shown). Then, the processing circuit 2a realizes the trained model generation function 2a_1 corresponding to the program by reading the program from the storage circuit and executing the program. In other words, the processing circuit 2a in the state where the program is read has the trained model generation function 2a_1 shown in the processing circuit 2a of FIG.

FIG. 2 is a diagram for explaining an example of a method of generating the trained model 53 at the time of learning of the first embodiment. For example, the trained model generation function 2a_1 is new by performing downsampling on raw data 51 of various subjects transmitted from the X-ray CT apparatus 1 and a plurality of other X-ray CT apparatus. Generate raw data 52.

The raw data 51 may be composed of, for example, a plurality of data arranged along the channel direction and the column direction, or may be composed of a plurality of data arranged along the channel direction and the view direction. FIG. 3 is a diagram showing an example of raw data 51 composed of a plurality of data arranged along a channel direction and a column direction in the first embodiment. FIG. 4 is a diagram showing an example of raw data 51 composed of a plurality of data arranged along the channel direction and the view direction in the first embodiment.

Downsampling is, for example, a plurality of raw data 51 shown in FIG. 3 by performing addition averaging processing on a plurality of data continuously arranged in at least one of the channel direction and the column direction. It is a process of bundling data into one data. Further, downsampling means, for example, in the raw data 51 shown in FIG. 4, by performing addition averaging processing on a plurality of data continuously arranged in at least one of the channel direction and the view direction. This is a process of bundling a plurality of data into one data. Therefore, the data size of the downsampled raw data is smaller than the data size of the raw data before downsampling.

The trained model generation function 2a_1 newly generates raw data 52 having a spatial resolution lower than that of the raw data 51 by executing downsampling. In the following description, the spatial resolution of the raw data 51 is referred to as "first spatial resolution", and the spatial resolution of the raw data 52 is referred to as "second spatial resolution". The first spatial resolution is an example of the first resolution. The second spatial resolution is an example of the second resolution.

FIG. 5 is a diagram showing an example of raw data 52 having a second spatial resolution according to the first embodiment. For example, the trained model generation function 2a_1 generates the raw data 52 shown in FIG. 5 by performing downsampling on the raw data 51 shown in FIG. The data size of the raw data 52 is smaller than the data size of the raw data 51.

Then, as shown in FIG. 2, at the time of learning, the trained model generation function 2a_1 generates the trained model 53 by learning the relationship between the raw data 51 and the raw data 52. The trained model 53 generates the raw data of the first spatial resolution (data corresponding to the raw data of the first spatial resolution) based on the raw data of the second spatial resolution, and the generated first space. Output raw resolution data.

That is, the trained model generation function 2a_1 generates the trained model 53 by learning the raw data 51 and the raw data 52 in association with each other. The trained model 53 receives the input of the raw data of the second spatial resolution, and generates and outputs the data corresponding to the raw data of the first spatial resolution.

For example, the trained model generation function 2a_1 performs machine learning by inputting the raw data 52 as input data and the raw data 51 as teacher data into the machine learning engine.

For example, machine learning engines include Deep Learning, Neural Network, Logistic regression analysis, non-linear discriminant analysis, Support Vector Machine (SVM), and Random Forest. , Naive Bayes, etc., to perform machine learning.

As a result of such machine learning, the trained model generation function 2a_1 generates a trained model 53 that outputs the first spatial resolution raw data in response to the input of the second spatial resolution raw data.

Then, the trained model generation function 2a_1 transmits the generated trained model 53 to the X-ray CT apparatus 1 via the network 3.

An example of the configuration of the X-ray CT apparatus 1 according to the first embodiment will be described with reference to FIG. FIG. 6 is a diagram showing an example of the configuration of the X-ray CT apparatus 1 according to the first embodiment. As shown in FIG. 6, the X-ray CT device 1 includes a gantry device 10, a sleeper device 30, and a console device 40. The X-ray CT apparatus 1 is an example of a medical image diagnostic apparatus, an example of a radiological diagnostic apparatus, and an example of a medical processing system.

In FIG. 6, the rotation axis of the rotation frame 13 in the non-tilt state or the longitudinal direction of the top plate 33 of the sleeper device 30 is the Z-axis direction. Further, the axial direction orthogonal to the Z-axis direction and horizontal to the floor surface is defined as the X-axis direction. Further, the axial direction orthogonal to the Z-axis direction and the X-axis direction and perpendicular to the floor surface is defined as the Y-axis direction. Note that FIG. 6 is a drawing of the gantry device 10 from a plurality of directions for the sake of explanation, and the number of gantry devices 10 included in the X-ray CT device 1 is one.

The gantry device 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 DAS (Data Acquisition System). Has 18 and.

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 collision of thermions. The X-ray tube 11 generates X-rays to irradiate the subject P by irradiating thermions from the cathode toward the anode by applying a high voltage from the X-ray high voltage device 14. For example, the X-ray tube 11 includes a rotating anode type X-ray tube that generates X-rays by irradiating a rotating anode with thermoelectrons. The X-ray tube 11 is an example of an X-ray generating unit.

The X-ray detector 12 detects X-rays irradiated from the X-ray tube 11 and passed through the subject P, and outputs a signal corresponding to the detected X-ray dose to the DAS 18. The X-ray detector 12 has, for example, a plurality of detection element sequences in which a plurality of detection elements are arranged in the channel direction along one 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 detection element sequences in which a plurality of detection elements are arranged in the channel direction are arranged in a row direction (slice direction, row direction). Further, the X-ray detector 12 is an indirect conversion type detector having, for example, a grid, a scintillator array, and an optical sensor array. The scintillator array has a plurality of scintillators. The scintillator has a scintillator crystal that outputs a photon amount of light according to the incident X-ray dose. The grid is arranged on the surface of the scintillator array on the X-ray incident side and has an X-ray shielding plate that absorbs scattered X-rays. The grid may also be called a collimator (one-dimensional collimator or two-dimensional collimator). The optical sensor array has a function of converting into an electric signal according to the amount of light from the scintillator, and has, for example, an optical sensor such as a photodiode. The X-ray detector 12 may be a direct conversion type detector having a semiconductor element that converts incident X-rays into an electric signal. The X-ray detector 12 is an example of an X-ray detector.

The rotating frame 13 is an annular frame that supports the X-ray tube 11 and the X-ray detector 12 so as to face each other and rotates the X-ray tube 11 and the X-ray detector 12 by the control device 15. For example, the rotating frame 13 is a casting made of aluminum. In addition to the X-ray tube 11 and the X-ray detector 12, the rotating frame 13 can further support the X-ray high voltage device 14, the wedge 16, the collimator 17, the DAS 18, and the like. Further, the rotating frame 13 can further support various configurations (not shown in FIG. 6). In the following, in the gantry device 10, the rotating frame 13 and the portion that rotates and moves together with the rotating frame 13 are also referred to as a rotating portion.

The X-ray high-voltage device 14 has an electric circuit such as a transformer and a rectifier, and has a high-voltage generator that generates a high voltage applied to the X-ray tube 11 and an X-ray that is generated by the X-ray tube 11. It has an X-ray control device that controls the output voltage according to the above. The high voltage generator may be a transformer type or an inverter type. The X-ray high voltage device 14 may be provided on the rotating frame 13 or on a fixed frame (not shown).

The control device 15 includes a processing circuit having a CPU (Central Processing Unit) and the like, and a drive mechanism such as a motor and an actuator. The control device 15 receives an input signal from the input interface 43 and controls the operation of the gantry device 10 and the sleeper device 30. For example, the control device 15 controls the rotation of the rotating frame 13, the tilt of the gantry device 10, the operation of the sleeper device 30 and the top plate 33, and the like. As an example, the control device 15 rotates the rotating frame 13 around an axis parallel to the X-axis direction based on the input tilt angle (tilt angle) information as a control for tilting the gantry device 10. The control device 15 may be provided in the gantry device 10 or in the console device 40.

The wedge 16 is a filter for adjusting the X-ray dose emitted from the X-ray tube 11. Specifically, the wedge 16 transmits the X-rays emitted from the X-ray tube 11 so that the distribution of the X-rays emitted from the X-ray tube 11 to the subject P becomes a predetermined distribution. It is a filter that attenuates. For example, the wedge 16 is a wedge filter or a bow-tie filter, which is a filter obtained by processing aluminum or the like so as to have a predetermined target angle and a predetermined thickness.

The collimator 17 is a lead plate or the like for narrowing the irradiation range of X-rays transmitted through the wedge 16, and a slit is formed by a combination of a plurality of lead plates or the like. The collimator 17 may be called an X-ray diaphragm. Further, in FIG. 6, the case where the wedge 16 is arranged between the X-ray tube 11 and the collimator 17 is shown, but it is the case where the collimator 17 is arranged between the X-ray tube 11 and the wedge 16. May be good. In this case, the wedge 16 is irradiated from the X-ray tube 11, and the X-ray whose irradiation range is limited by the collimator 17 is transmitted and attenuated.

The DAS 18 collects X-ray signals detected by each detection element included in the X-ray detector 12. For example, the DAS 18 has an amplifier that amplifies an electric signal output from each detection element and an A / D converter that converts an electric signal that is an analog signal into a digital signal, and generates detection data. To do.

The detection data generated by the DAS 18 is transmitted from a transmitter having a light emitting diode (LED) provided on the rotating frame 13 by optical communication to a non-rotating portion (for example, a fixed frame, etc.) of the gantry device 10. FIG. It is transmitted to a receiver having a light diode provided in (not shown in) and transferred to the console device 40. Here, the non-rotating portion is, for example, a fixed frame that rotatably supports the rotating frame 13. The method of transmitting data from the rotating frame 13 to the non-rotating portion of the gantry device 10 is not limited to optical communication, and any non-contact data transmission method may be adopted, and the contact-type data transmission method may be used. You may adopt it.

The sleeper device 30 is a device for placing and moving the subject P to be scanned, and has a base 31, a sleeper 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 as to be movable in the vertical direction. The sleeper 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, and includes a motor, an actuator, and the like. The top plate 33 provided on the upper surface of the support frame 34 is a plate on which the subject P is placed. In addition to the top plate 33, the sleeper drive device 32 may move the support frame 34 in the longitudinal direction of the top plate 33.

The console device 40 includes a memory 41, a display 42, an input interface 43, and a processing circuit 44. Although the console device 40 will be described as a separate body from the gantry device 10, the gantry device 10 may include a part of each component of the console device 40 or the console device 40.

The memory 41 is realized by, for example, a RAM (Random Access Memory), a semiconductor memory element such as a flash memory, a hard disk, an optical disk, or the like. The memory 41 is an example of a storage unit. The memory 41 stores, for example, raw data, projection data, and CT image data. Further, for example, the memory 41 stores a program for the circuit included in the X-ray CT apparatus 1 to realize its function. The memory 41 may be realized by a server group (cloud) connected to the X-ray CT apparatus 1 via a network. In this case, the system composed of the server group and the X-ray CT apparatus 1 is an example of a medical processing system.

The display 42 displays various information. For example, the display 42 displays various images indicated by various image data generated by the processing circuit 44, and is a GUI (Graphical User Interface) or the like for receiving various operations from users such as doctors and radiological technologists. Or display. For example, the display 42 is a liquid crystal display or a CRT (Cathode Ray Tube) display. The display 42 may be a desktop type, or may be composed of a tablet terminal or the like capable of wireless communication with the console device 40 main body.

The input interface 43 receives various input operations from the user, converts the received input operations into electric signals, and outputs the received input operations to the processing circuit 44. For example, the input interface 43 receives from the user a collection condition when collecting raw data, a reconstruction condition when reconstructing CT image data, and the like. For example, the input interface 43 includes a mouse, a keyboard, a trackball, a switch, a button, a joystick, a touch pad for performing input operations by touching an operation surface, a touch screen in which a display screen and a touch pad are integrated, and an optical sensor. It is realized by the non-contact input circuit, voice input circuit, etc. used. The input interface 43 may be provided in the gantry device 10. Further, the input interface 43 may be composed of a tablet terminal or the like capable of wireless communication with the console device 40 main body. Further, the input interface 43 is not limited to the one provided with physical operating parts such as a mouse and a keyboard. For example, an electric signal processing circuit that receives an electric signal corresponding to an input operation from an external input device provided separately from the console device 40 and outputs the electric signal to the processing circuit 44 is also an example of the input interface 43. included.

The processing circuit 44 controls the operation of the entire X-ray CT apparatus 1. The processing circuit 44 is not limited to the case where it is included in the console device 40. For example, the processing circuit 44 may be included in an integrated server that collectively processes detection data acquired by a plurality of X-ray CT devices. For example, an integrated server connected to the X-ray CT apparatus 1 via a network may have a processing circuit 44. In this case, the X-ray CT apparatus 1 transmits the collected raw data to the integrated server. The integrated server then receives the raw data. Then, the processing circuit 44 of the integrated server executes various processes described below with respect to the received raw data. In this case, the system composed of the integrated server and the X-ray CT apparatus 1 is an example of a medical processing system.

For example, the processing circuit 44 executes the system control function 441, the preprocessing function 442, the image generation function 443, the image processing function 444, and the output function 445. For example, the processing circuit 44 reads a program corresponding to the system control function 441 from the memory 41, executes the read program, and is based on an input operation received from the user via the input interface 43. It controls various functions of 44.

For example, when the system control function 441 receives the trained model 53 transmitted from the trained model generation device 2, the system control function 441 stores the received trained model 53 in the memory 41. The trained model 53 stored in the memory 41 is used by the image generation function 443.

Further, for example, the system control function 441 controls the X-ray CT apparatus 1 to execute a scan. The scan includes a positioning scan (positioning photography) and a main scan (main photography). For example, the system control function 441 selects a scan protocol according to the diagnosis from a plurality of types of scan protocols stored in the memory 41. The system control function 441 then performs a scan based on the selected scan protocol. The system control function 441 can execute scanning by various methods such as a conventional scan, a helical scan, a step-and-shoot method and a dynamic scan.

For example, the system control function 441 controls the X-ray CT apparatus 1 to perform a positioning scan. For example, the system control function 441 fixes the position of the X-ray tube 11 at a predetermined rotation angle, and irradiates the subject P with X-rays from the X-ray tube 11 while moving the top plate 33 in the Z direction. , Perform a positioning scan. Further, the processing circuit 44 reads a program corresponding to the image generation function 443 from the memory 41 and executes the read program, so that the two-dimensional positioning image data is based on the raw data collected by the positioning scan. Alternatively, three-dimensional positioning image data is generated. The positioning image data may be referred to as a scanno image data or a scout image data.

In addition, the system control function 441 controls the X-ray CT apparatus 1 to execute the main scan for collecting projection data. For example, the system control function 441 sets the scan conditions (collection conditions) for the main scan based on the positioning image data. As a collection condition, the system control function 441 generates, for example, timing data indicating the timing of changing the tube voltage supplied to the X-ray tube 11, and stores the generated timing data in the memory 41. Next, the system control function 441 moves the subject P into the photographing port of the gantry device 10 by controlling the sleeper drive device 32. Further, the system control function 441 adjusts the opening degree and the position of the collimator 17. Further, the system control function 441 rotates the rotating portion by controlling the control device 15.

Further, the system control function 441 controls the X-ray high voltage device 14 to supply a high voltage to the X-ray tube 11. As a result, the X-ray tube 11 generates X-rays to irradiate the subject P.

While the scan is performed by the system control function 441, the plurality of DAS18s collect a plurality of X-ray signals detected by the plurality of detecting elements and generate detection data. Further, the processing circuit 44 reads a program corresponding to the pre-processing function 442 from the memory 41 and executes the read program to perform pre-processing on the detection data output from the DAS 18. For example, the preprocessing function 442 performs preprocessing such as logarithmic conversion processing, offset correction processing, sensitivity correction processing between channels, and beam hardening correction on the detection data output from DAS18. The data after preprocessing is called raw data. Further, the detection data before the preprocessing and the raw data after the preprocessing are referred to as projection data. The X-ray CT apparatus 1 collects raw data in this way. For example, the preprocessing function 442 collects the raw data 51 by generating the raw data 51 as shown in FIGS. 3 and 4 above. The gantry device 10 and the preprocessing function 442 are examples of a collection unit that collects raw data. Raw data is an example of data.

Further, the processing circuit 44 reads a program corresponding to the image generation function 443 from the memory 41 and executes the read program to reconstruct the CT image data based on the raw data. Specifically, the image generation function 443 reconstructs the CT image data by performing reconstruction processing using a filter correction back projection method, a successive approximation reconstruction method, or the like on the raw data. In this way, the image generation function 443 generates CT image data based on the raw data. The image generation function 443 is an example of an image generation unit. The CT image data is an example of image data.

Further, the processing circuit 44 reads a program corresponding to the image processing function 444 from the memory 41 and executes the read program to perform various image processing on the CT image data. For example, the image processing function 444 uses a known method to obtain tomographic image data of an arbitrary cross section (for example, an axial cross section) of CT image data reconstructed based on an input operation received from a user via the input interface 43. Converts to three-dimensional image data or MIP (Maximum Intensity Projection) image data. That is, the image processing function 444 generates tomographic image data, three-dimensional image data, or MIP image data based on the CT image data. Further, the image processing function 444 stores tomographic image data, three-dimensional image data, or MIP image data in the memory 41. The tomographic image data, three-dimensional image data, or MIP image data is an example of image data for display. Further, the tomographic image shown by the tomographic image data, the three-dimensional image shown by the three-dimensional image data, and the MIP image shown by the MIP image data are examples of images.

Further, the processing circuit 44 reads a program corresponding to the output function 445 from the memory 41 and executes the read program to output tomographic image data, three-dimensional image data, CT image data, and the like. For example, the output function 445 displays various images such as a tomographic image indicated by the tomographic image data, a three-dimensional image indicated by the three-dimensional image data, and a MIP image indicated by the MIP image data on the display 42. The output function 445 is an example of a display control unit.

Further, for example, the output function 445 stores CT image data, tomographic image data, three-dimensional image data, and MIP image data in an external device (for example, image data) connected to the X-ray CT device 1 via the network 3. Output to the server device, etc. Further, for example, the output function 445 transmits the raw data generated by the preprocessing function 442 to the trained model generator 2 via the network 3 in order to cause the trained model generator 2 to generate the trained model. You may.

In the X-ray CT apparatus 1 shown in FIG. 6, each processing function is stored in the memory 41 in the form of a program that can be executed by a computer. The processing circuit 44 is a processor that realizes a function corresponding to each program by reading each program from the memory 41 and executing the program. In other words, the processing circuit 44 in the state where each program is read has a function corresponding to the read program. Note that FIG. 6 shows a case where each processing function of the system control function 441, the preprocessing function 442, the image generation function 443, the image processing function 444, and the output function 445 is realized by a single processing circuit 44. , The embodiment is not limited to this. For example, the processing circuit 44 may be configured by combining a plurality of independent processors, and each processor may execute each program to realize each processing function. Further, each processing function of the processing circuit 44 may be appropriately distributed or integrated into a single or a plurality of processing circuits.

The term "processor" used in the above description refers to, for example, a CPU, a GPU (Graphics Processing Unit), an application specific integrated circuit (ASIC), or a programmable logic device (for example, a simple programmable logic device (for example, a simple programmable logic device). It means a circuit such as a Simple Programmable Logic Device (SPLD), a Complex Programmable Logic Device (CPLD), or a Field Programmable Gate Array (FPGA). The processor realizes the function by reading and executing the program stored in the memory 41.

In FIG. 6, it has been described that a single memory 41 stores each program corresponding to each processing function. However, a plurality of memories 41 may be distributed and arranged, and the processing circuit 44 may be configured to read a corresponding program from the individual memories 41. Further, instead of storing the program in the memory 41, the program may be directly incorporated in the circuit of the processor. In this case, the processor realizes the function by reading and executing the program embedded in the circuit.

Further, the processing circuit 44 may realize various functions by using a processor of an external device connected via a network. For example, the processing circuit 44 reads a program corresponding to each function from the memory 41 and executes it, and uses an external workstation or server group (cloud) connected to the X-ray CT device 1 via a network as a computational resource. By using it, each function shown in FIG. 6 may be realized. In this case, the system composed of the X-ray CT apparatus 1 and an external workstation or server group is an example of a medical processing system.

The example of the configuration of the X-ray CT apparatus 1 has been described above. Under such a configuration, the X-ray CT apparatus 1 can suppress the increase in the data size of the raw data stored in the memory 41 and the decrease in the spatial resolution of the image data based on the raw data. Various processes described below are executed so as to be possible. That is, the X-ray CT apparatus 1 executes various processes described below so as to ensure image quality that does not interfere with image interpretation while reducing the data size of the raw data stored in the memory 41. ..

FIG. 7 is a diagram for explaining an example of processing executed by the image generation function 443 according to the first embodiment. As shown in FIG. 7, for example, the image generation function 443 identifies a part of the raw data 56 of the first spatial resolution raw data 55 collected by the main scan, and stores the specified raw data 56 in the memory. Store in 41. The raw data 55 is composed of, for example, a plurality of raw data. In addition, the image generation function 443 identifies at least a part of the raw data 57 out of the raw data 55. Then, the image generation function 443 generates the raw data 58 having the second spatial resolution by performing downsampling on the specified raw data 57. Then, the image generation function 443 stores the raw data 58 in the memory 41.

Therefore, the memory 41 stores the raw data 56, which is a part of the raw data 55 corresponding to the first spatial resolution collected by executing the scan on the subject P. Further, the memory 41 stores the raw data 58 corresponding to the second spatial resolution lower than the first spatial resolution, which is acquired by performing downsampling on at least a part of the raw data 55. The raw data 55 is an example of the first data. The raw data 56 is an example of partial data. The raw data 58 is an example of the second data. The raw data 55, the raw data 56, and the raw data 58 are data corresponding to the intensity of the X-ray transmitted through the subject P.

The raw data 57 to which downsampling is executed may be, for example, raw data other than the raw data 56 in the raw data 55, or may be the raw data 55 itself. In any case, downsampling is performed under the condition that the sum of the data size of the raw data 56 and the data size of the raw data 58 is smaller than the data size of the raw data 55.

Therefore, the free space of the memory 41 when the memory 41 stores the raw data 56 and the raw data 58 is larger than the free space of the memory 41 when the memory 41 stores the raw data 55. Therefore, according to the first embodiment, it is possible to suppress an increase in the data size of the raw data stored in the memory 41. That is, according to the first embodiment, the data size of the raw data stored in the memory 41 can be reduced.

Further, the transmission time required when the output function 445 transmits the raw data 56 and the raw data 58 to an external device (not shown) connected to the network 3 for storing the raw data transmits the raw data 55. It will be shorter than the transmission time required in some cases. Therefore, according to the first embodiment, it is possible to suppress a long transmission time when transmitting the raw data to the external device.

Here, an example of a method of specifying the raw data 56 from the raw data 55 and an example of a method of specifying the raw data 57 to be downsampled from the raw data 55 will be described.

First, the case where the X-ray CT apparatus 1 collects the raw data of the first spatial resolution by the helical scan or the step-and-shoot method will be described. In this case, for example, the image generation function 443 generates CT image data based on the raw data for each raw data corresponding to the CT image data for one frame of the raw data 55. Then, the image processing function 444 converts the CT image data into tomographic image data, three-dimensional image data, or MIP image data. Then, the image processing function 444 is a diagnostic target such as a tumor to be diagnosed or an organ to be diagnosed in the image data with respect to image data such as CT image data, tomographic image data, three-dimensional image data or MIP image data. A CAD (Computer Aided Detection) process for automatically detecting a site is executed. As a result, for example, it is divided into image data having a part to be diagnosed and image data not having a part to be diagnosed.

Then, the image generation function 443 specifies the raw data used when generating the image data having the site to be diagnosed among the plurality of raw data constituting the raw data 55 as the raw data 56. In this case, the raw data 56 is important data including the site to be diagnosed. In this way, the raw data 56 is identified from the raw data 55 by analyzing the image data based on the raw data 55. The image data based on the raw data 55 is an example of the first image data.

Further, the image generation function 443 specifies raw data other than the raw data 56 as the raw data 57 among the plurality of raw data constituting the raw data 55. In this case, the raw data 57 is less important than the raw data 56 because it does not include the site to be diagnosed.

The image generation function 443 may specify the raw data 56 by another method. For example, the image generation function 443 executes CAD processing or the like on the positioning image data of the subject P instead of the image data such as CT image data, tomographic image data, three-dimensional image data, and MIP image data. , The raw data 56 may be specified by the same method as described above. The positioning image data is image data different from image data such as CT image data, tomographic image data, three-dimensional image data, and MIP image data, and is image data of the subject P. In this case, the raw data 56 is identified from the raw data 55 by analysis of the positioning image data. The positioning image data is an example of the second image data.

Further, the image generation function 443 is past image data (CT) which is image data based on raw data collected by scanning the subject P in the past diagnosis prior to the present diagnosis by the X-ray CT apparatus 1. By executing CAD processing or the like on image data, tomographic image data, three-dimensional image data, MIP image data, etc.), the raw data 56 may be specified by the same method as described above. The past image data is image data different from the image data based on the raw data collected in this diagnosis, and is the image data of the subject P. That is, the raw data 56 is identified from the raw data 55 by analyzing the past image data. The past image data is an example of the second image data.

Further, the image generation function 443 performs CAD processing or the like on the past image data obtained by imaging or photographing the subject P in the past diagnosis by another medical image diagnostic device different from the X-ray CT device 1. May identify the raw data 56 in a manner similar to that described above. For example, the image generation function 443 acquires past image data from another medical image diagnostic device (not shown) or an external device (not shown) connected to the network 3. Then, the image generation function 443 executes a process for identifying the raw data 56 with respect to the acquired past image data. Examples of the acquired past image data include ultrasonic image data, MR (Magnetic Resonance) image data, PET (Positron Emission Tomography) -CT image data, PET-MR image data, X-ray image data, and optical image data. That is, the raw data 56 is identified from the raw data 55 by such analysis of the past image data. Such past image data is also an example of the second image data.

For example, the image generation function 443 executes CAD processing on the acquired past image data. As a result, for example, it is divided into past image data having a part to be diagnosed and past image data having no part to be diagnosed. Then, the image generation function 443 specifies, among the raw data 56, the raw data corresponding to the past image data having the site to be diagnosed as the raw data 56.

Further, the image generation function 443 displays an image represented by the image data given as an example of the first image data described above, or an image represented by the image data cited as an example of the second image data described above. It may be displayed in. Then, the image generation function 443 may accept the range of the raw data 56 from the user who has confirmed the image displayed on the display 42 via the input interface 43. Then, the image generation function 443 may specify a part of the raw data 56 of the raw data 55 based on the range received from the user. That is, the raw data 56 is specified from the raw data 55 by the user's operation on the image data given as an example of the first image data or the image data given as an example of the second image data described above. ..

The image generation function 443 may also identify the raw data 56 using the scan protocol selected when performing the main scan. For example, each of the plurality of scanning protocols is a protocol for scanning a specific part. Therefore, the selected scan protocol is a protocol for scanning a specific site to be diagnosed. Therefore, the image generation function 443 identifies a site to be diagnosed based on the selected scan protocol. Then, the image generation function 443 specifies the raw data including the specified portion of the raw data 55 as the raw data 56. That is, the raw data 56 is identified from the raw data 55 by the scan protocol selected when performing the scan.

In addition, the user may know the position of the part to be diagnosed in advance from the past diagnosis or the like. In this case, the user may input the range of the position of the part to be diagnosed in advance via the input interface 43. In this case, the image generation function 443 stores the input position information indicating the range of the position of the diagnosis target portion in the memory 41 in advance. The range of the part indicated by the position information may be the range of the position relative to the entire collection range of the raw data 55, or the absolute range in the coordinate system defined for the X-ray CT apparatus 1. It may be in the range of positions. Then, the image generation function 443 acquires the position information stored in the memory 41 when specifying the raw data 56. Then, the image generation function 443 specifies the raw data within the range indicated by the position information among the raw data 55 as the raw data 56. That is, the raw data 56 is specified from the raw data 55 by the position information. The position information is an example of information regarding a predetermined data collection position.

Next, a case where the X-ray CT apparatus 1 collects the raw data 55 having the first spatial resolution by a dynamic scan in which the same position is photographed over a plurality of time phases will be described. In this case, the raw data 55 is composed of raw data of a plurality of time phases. The image generation function 443 specifies the raw data of a specific time phase among the raw data 55 as the raw data 56. For example, the specific time phase may include a specific cardiac time phase or a specific heartbeat. Then, the image generation function 443 specifies the raw data other than the raw data 56 among the raw data of the plurality of time phases constituting the raw data 55 as the raw data 57.

Next, with reference to FIG. 8, an example of a processing flow executed by the processing circuit 44 when the trained model 53, the raw data 56, and the raw data 58 are stored in the memory 41 of the X-ray CT apparatus 1. explain. FIG. 8 is a flowchart showing an example of the flow of processing executed by the processing circuit 44 according to the first embodiment. In the process shown in FIG. 8, when the trained model 53, the raw data 56, and the raw data 58 are stored in the memory 41 of the X-ray CT apparatus 1, the input interface 43 requests the user to generate CT image data. Executed when accepted.

In the first embodiment, there are two requirements for generating CT image data. The first request is a request to generate CT image data based on the raw data 56. The second request is a request to generate CT image data based on the raw data 58. In the following description, the first requirement is referred to as the first requirement. Further, in the following description, the second requirement is referred to as a second requirement.

The first request is a request that the user inputs to the X-ray CT apparatus 1 when, for example, the user wants to diagnose the state of an important part to be diagnosed included in the raw data 56. The second request is, for example, a request that the user inputs to the X-ray CT apparatus 1 when a part that the user needs to diagnose is newly discovered and this part is included in the raw data 58. ..

As shown in FIG. 8, the image generation function 443 determines whether or not the received request is the first request (step S100). When the received request is the first request (step S100: Yes (first request)), the image generation function 443 generates CT image data corresponding to the first spatial resolution based on the raw data 56. (Step S101). That is, when the image generation function 443 receives the request to generate the CT image data based on the raw data 56, the image generation function 443 generates the CT image data without using the trained model 53.

Then, the image processing function 444 generates image data for display such as tomographic image data, three-dimensional image data, or MIP image data from the generated CT image data, and the output function 445 is indicated by the image data for display. The image to be displayed is displayed on the display 42 (step S102), and the process ends.

FIG. 9 is a diagram showing an example of an image according to the first embodiment. For example, the output function 445 causes the tomographic image 59 shown in FIG. 9 to be displayed on the display 42 in step S102.

On the other hand, when the received request is not the first request (step S100: No (second request)), the received request is the second request. Therefore, the image generation function 443 uses the trained model 53 to generate the first spatial resolution raw data from the second spatial resolution raw data 58 (step S103).

FIG. 10 is a diagram for explaining an example of the process of step S103 of the first embodiment. As shown in FIG. 10, during operation, the image generation function 443 inputs the raw data 58 of the second spatial resolution into the trained model 53, so that the raw data 60 of the first spatial resolution is input to the trained model 53. Is output. That is, the image generation function 443 causes the trained model 53 to generate the raw data 60 having the first spatial resolution. Then, the image generation function 443 acquires the generated (output) raw data 60. In this way, the image generation function 443 generates the raw data 60.

Then, the image generation function 443 generates CT image data based on the raw data 60 (step S104). That is, when the image generation function 443 receives the second request for generating CT image data based on the raw data 58, the image generation function 443 learns to generate the raw data 60 corresponding to the first spatial resolution based on the raw data 58. CT image data is generated using the finished model 53. Raw data 60 is an example of a third piece of data. The raw data 60 is data corresponding to the intensity of X-rays transmitted through the subject P.

Then, the image processing function 444 generates image data for display from the generated CT image data, and the output function 445 causes the image indicated by the image data for display to be displayed on the display 42 (step S105), and the process is performed. To finish.

The CT image data generated in steps S101 and S104 is generated based on the raw data 60 having a first spatial resolution higher than the second spatial resolution, not the raw data 58 having a second spatial resolution. Therefore, the spatial resolution of the CT image data generated in steps S101 and S104 corresponds to the first spatial resolution, not the second spatial resolution. Therefore, according to the first embodiment, it is possible to suppress a decrease in the spatial resolution of the CT image data. Therefore, according to the first embodiment, it is possible to secure the image quality of the image data for display without hindering the interpretation.

Further, as described above, according to the first embodiment, the data size of the raw data stored in the memory 41 can be reduced. Therefore, according to the first embodiment, it is possible to secure the image quality that does not hinder the interpretation while reducing the data size of the raw data stored in the memory 41.

Further, in the first embodiment, the raw data 56 including the part to be diagnosed is stored in the memory 41 in a state having the first spatial resolution. Then, CT image data is generated from the raw data 56 without using the trained model 53. Here, the accuracy of the first spatial resolution raw data 60 generated by the trained model 53 may not be better than the accuracy of the first spatial resolution raw data 57 before being downsampled. In this case, the accuracy of the CT image data and the image data for display based on the raw data 60 generated by the trained model 53 may not be good. However, according to the first embodiment, the trained model 53 is not used, and the CT image data and the image data for display are generated from the raw data 56 having the first spatial resolution. Therefore, according to the first embodiment, the accuracy of the CT image data based on the raw data including the part to be diagnosed and the image data for display can be improved.

The first embodiment has been described above. According to the first embodiment, as described above, it is possible to secure the image quality that does not hinder the interpretation while reducing the data size of the raw data stored in the memory 41.

(Modification 1 of the first embodiment)
In the first embodiment described above, the case where the trained model generation device 2 generates the trained model 53 by using the raw data 52 as the input data and the raw data 51 as the teacher data has been described. However, the trained model generator 2 may generate the trained model by using other data as input data instead of the raw data 52. Therefore, such a modification will be described as a modification 1 of the first embodiment. In the description of the first embodiment, the points different from the first embodiment will be mainly described, and the description of the configuration similar to that of the first embodiment may be omitted.

FIG. 11 is a diagram for explaining an example of processing executed by the trained model generation device 2 according to the first modification of the first embodiment. As shown in FIG. 11, the trained model generation function 2a_1 of the trained model generation device 2 reconstructs the CT image data 61 based on the raw data 51. Specifically, the trained model generation function 2a_1 reconstructs the CT image data 61 by performing reconstruction processing on the raw data 51 using a filter correction back projection method, a successive approximation reconstruction method, or the like. In this way, the trained model generation function 2a_1 generates CT image data 61 based on the raw data 51.

Then, the trained model generation function 2a_1 performs forward projection on the CT image data 61 to generate raw data 62 having a second spatial resolution.

FIG. 12 is a diagram for explaining an example of a method of generating the trained model 53a at the time of learning of the modified example 1 of the first embodiment. As shown in FIG. 12, at the time of learning, the trained model generation function 2a_1 generates the trained model 53a by learning the relationship between the raw data 51 and the raw data 62. The trained model 53a generates the raw data of the first spatial resolution (data corresponding to the raw data of the first spatial resolution) based on the raw data of the second spatial resolution, and the generated second space. Output raw resolution data.

That is, the trained model generation function 2a_1 generates the trained model 53a by learning the raw data 51 and the raw data 62 in association with each other. The trained model 53a receives the input of the raw data of the second spatial resolution, and generates and outputs the raw data of the first spatial resolution.

For example, the trained model generation function 2a_1 performs machine learning by inputting raw data 62 as input data and raw data 51 as teacher data into a machine learning engine.

As a result of such machine learning, the trained model generation function 2a_1 generates a trained model 53a that outputs the first spatial resolution raw data in response to the input of the second spatial resolution raw data.

Then, the trained model generation function 2a_1 transmits the generated trained model 53a to the X-ray CT apparatus 1 via the network 3. The X-ray CT apparatus 1 uses the trained model 53a to perform the same processing as the various processing performed using the trained model 53 described above. Therefore, according to the first modification of the first embodiment, the same effect as that of the first embodiment described above can be obtained.

(Modification 2 of the first embodiment)
In the first embodiment described above, the case where the X-ray CT apparatus 1 accepts the first request and the second request has been described. However, the X-ray CT apparatus 1 may accept other requests instead of the second request. Therefore, such a modification will be described as a modification 2 of the first embodiment. In the description of the modified example 2, the points different from the first embodiment and the modified example 1 will be mainly described, and the description of the same configuration as the first embodiment and the modified example 1 may be omitted.

In the second modification, the X-ray CT apparatus 1 accepts three requests, a first request, a third request, and a fourth request. The third requirement is to generate CT image data corresponding to the first spatial resolution based on the raw data 58. The fourth requirement is to generate CT image data corresponding to the second spatial resolution based on the raw data 58. The three requirements, the first requirement, the third requirement, and the fourth requirement, are requests for generating CT image data.

With reference to FIG. 13, an example of a processing flow executed by the processing circuit 44 when the trained model 53, the raw data 56, and the raw data 58 are stored in the memory 41 of the X-ray CT apparatus 1 will be described. FIG. 13 is a flowchart showing an example of the flow of processing executed by the processing circuit 44 according to the second modification of the first embodiment. In the process shown in FIG. 13, when the trained model 53, the raw data 56, and the raw data 58 are stored in the memory 41 of the X-ray CT apparatus 1, the input interface 43 requests the user to generate CT image data. Executed when accepted.

As shown in FIG. 13, the image generation function 443 determines whether or not the received request is the first request, as in the first embodiment (step S100). When the received request is the first request (step S100: Yes (first request)), the image generation function 443 generates CT image data based on the raw data 56, as in the first embodiment. (Step S101). Then, as in the first embodiment, the image processing function 444 generates image data for display from the generated CT image data, and the output function 445 displays the image indicated by the image data for display 42. Is displayed (step S102), and the process ends.

On the other hand, when the received request is not the first request (step S100: No), it is determined whether or not the received request is the third request (step S200). If the received request is a third request (step S200: Yes (third request)), the image generation function 443 uses the trained model 53 to take from the raw data 58 of the second spatial resolution. The raw data 60 of the first spatial resolution is generated (step S201).

Then, the image generation function 443 generates CT image data corresponding to the first spatial resolution based on the raw data 60 (step S202). That is, when the image generation function 443 receives a request to generate CT image data corresponding to the first spatial resolution based on the raw data 58, the trained model 53 is used to correspond to the first spatial resolution. Generate CT image data.

Then, the image processing function 444 generates image data for display from the generated CT image data, and the output function 445 causes the image indicated by the image data for display to be displayed on the display 42 (step S203), and the process is performed. To finish.

On the other hand, when the received request is not the third request (step S200: No (fourth request)), the received request is the fourth request. Therefore, the image generation function 443 generates CT image data corresponding to the second spatial resolution based on the raw data 58 (step S204). That is, when the image generation function 443 receives a request to generate CT image data corresponding to the second spatial resolution based on the raw data 58, the image generation function 443 corresponds to the second spatial resolution without using the trained model 53. Generate CT image data to be used.

Then, the image processing function 444 generates image data for display from the generated CT image data, and the output function 445 causes the image indicated by the image data for display to be displayed on the display 42 (step S205), and the process is performed. To finish.

As described above, in the second modification, the CT image data corresponding to the first spatial resolution based on the raw data 58 and the CT image data corresponding to the second spatial resolution based on the raw data 58 are generated. Therefore, according to the second modification, CT image data corresponding to the spatial resolution desired by the user can be generated from the same raw data 58.

The modified example 2 of the first embodiment has been described above. As described above, according to the second modification, CT image data corresponding to the spatial resolution desired by the user can be generated from the same raw data 58. Further, according to the second modification, the same effect as that of the first embodiment or the like can be obtained.

(Second embodiment)
In the first embodiment and the first and second modifications 1 and 2 described above, the case where the learned model 53 stored in the memory 41 of the X-ray CT apparatus 1 is not updated has been described. However, the trained model 53 stored in the memory 41 may be updated. Therefore, such an embodiment will be described as a second embodiment. In the description of the second embodiment, the points different from those of the first embodiment and the modified examples 1 and 2 will be mainly described. Further, the same reference numerals may be given to the same configurations as those of the first embodiment and the first and second modifications, and the description thereof may be omitted.

FIG. 14 is a diagram showing an example of the configuration of the X-ray CT apparatus 1a according to the second embodiment. The X-ray CT apparatus 1a shown in FIG. 14 is different from the X-ray CT apparatus 1 according to the first embodiment shown in FIG. 6 in that the console apparatus 40a is provided in place of the console apparatus 40. Further, the console device 40a according to the second embodiment is different from the console device 40 shown in FIG. 6 in that the processing circuit 44a is provided in place of the processing circuit 44 shown in FIG. Further, the processing circuit 44a according to the second embodiment is different from the processing circuit 44 shown in FIG. 6 in that it includes the trained model update function 446.

15 and 16 are diagrams for explaining an example of the process executed by the trained model update function 446 according to the second embodiment. The trained model update function 446 additionally trains (additionally trains) the trained model 53 to generate the trained model 53b. That is, the trained model update function 446 generates the trained model 53b by updating the trained model 53.

For example, as shown in FIG. 15, the trained model update function 446 uses raw data 58 (see FIG. 7) as input data and raw data 57 (see FIG. 7) as teacher data, and raw data 58 and raw data 57. The trained model 53b is generated by further training the trained model 53 in association with the above. That is, the trained model update function 446 has been trained by using the raw data 57 which is at least a part of the raw data 55 and the raw data 58 acquired by performing downsampling on the raw data 57. By learning the model 53, the trained model 53b is acquired. The raw data 57 is an example of partial data. Further, the trained model 53 is an example of the first trained model. Further, the trained model 53b is an example of the second trained model.

For example, when the raw data 58 is input to the trained model 53a, the trained model 53a generates and outputs data corresponding to the raw data 57. Therefore, the trained model 53a can output accurate data in response to the input of the raw data 58.

Here, for example, the trained model update function 446 generates a copy of the trained model 53, stores the copy of the trained model 53 in the memory 41, and then updates the copy of the trained model 53. , The trained model 53a may be generated. In this case, the trained model 53 and the trained model 53b are stored in the memory 41. Therefore, as it is, when the image generation function 443 receives a request to generate CT image data based on the raw data 58, the trained model to be used is selected from the trained model 53 and the trained model 53b. Difficult to identify.

Further, the X-ray CT apparatus 1a generates raw data 55 (see FIG. 7) one after another. Therefore, the image generation function 443 specifies the raw data 56 (see FIG. 7) and the raw data 57 (see FIG. 7) each time the raw data 55 is generated, and the raw data 57 to the raw data 58 (see FIG. 7). ) Is generated. Therefore, the trained model update function 446 updates the trained model 53a by updating the copy of the trained model 53 using the newly specified raw data 57 each time the raw data 57 is newly specified. It may be generated. In this case, the trained model update function 446 generates the trained model 53a one after another. Therefore, the trained model 53 and the plurality of trained models 53a generated one after another are stored in the memory 41. Therefore, when the image generation function 443 receives a request to generate CT image data based on the newly generated raw data 58, the learning to be used from the trained model 53 and the plurality of trained models 53b is used. It is difficult to identify the finished model.

Therefore, it is necessary to associate the newly generated raw data 58 with the trained model 53b generated by using the raw data 58.

Therefore, as shown in FIG. 16, the trained model update function 446 stores the trained model 53b and the raw data 58 in the memory 41 in association with each other. For example, the trained model update function 446 may associate the trained model 53b with the raw data 58 by adding an identifier that identifies the raw data 58 to the trained model 53b. In this way, the memory 41 stores the raw data 58 in association with the trained model 53b.

Then, when the image generation function 443 receives the request to generate the CT image data based on the raw data 58, the image generation function 443 executes the process described below in the previous step S103 or step S201. For example, the image generation function 443 refers to the stored contents of the memory 41 and identifies the trained model 53b associated with the raw data 58. The image generation function 443 then uses the identified trained model 53b to generate the raw data 60 with the first spatial resolution. That is, when the image generation function 443 receives a request to generate CT image data based on the raw data 58, the image generation function 443 applies the raw data 58 to the trained model 53b associated with the raw data 58, thereby performing CT. Generate image data.

Therefore, according to the second embodiment, even when a plurality of trained models are stored in the memory 41, CT image data can be generated using an appropriate trained model. Therefore, according to the second embodiment, CT image data can be generated with high accuracy.

The second embodiment has been described above. According to the second embodiment, as described above, CT image data can be generated with high accuracy. Further, in the second embodiment, the same effects as those of the first embodiment and the modified examples 1 and 2 described above can be obtained.

(Third Embodiment)
According to the first embodiment, the second embodiment and the modified examples 1 and 2 described above, the case where the X-ray CT apparatus 1, 1a collects the raw data 55 of the first spatial resolution has been described. However, the X-ray CT devices 1 and 1a may be able to collect the raw data 55 having the first spatial resolution and may be able to collect the raw data having the second spatial resolution. Therefore, such an embodiment will be described as a third embodiment. In the description of the third embodiment, the points different from those of the first embodiment will be mainly described. However, the contents of the third embodiment described below can be applied to the second embodiment and the first and second modifications. Further, in the description of the third embodiment, the same reference numerals may be given to the same configurations as those of the first embodiment, and the description may be omitted.

For example, in the third embodiment, the X-ray CT apparatus 1 collects the first spatial resolution raw data 55 and the second spatial resolution raw data. This scan is performed in the mode selected by the user from the collection modes.

For example, when performing the main scan in the first collection mode, the X-ray CT apparatus 1 collects raw data 55 based on electrical signals output from all detection elements within a predetermined range of the X-ray detector 12. To do.

Further, when the X-ray CT apparatus 1 performs the main scan in the second collection mode, the electricity output from a predetermined number of detection elements among all the detection elements within the predetermined range of the X-ray detector 12. Collect raw signal-based data. For example, the X-ray CT apparatus 1 has four detection elements of "2 (channel direction) x 2 (column direction)" as one block, and a predetermined range so that a plurality of blocks are continuous in the channel direction and the column direction. Set multiple blocks for the detection element inside. Then, when the main scan is performed in the second collection mode, the X-ray CT apparatus 1 collects raw data based on the electric signal output from one detection element for each block. In this case, since the electric signal output from one detection element is used for each block composed of four detection elements, the data size of the raw data collected in the second collection mode is the first collection mode. It is a quarter of the data size of the raw data collected in.

Here, the subject to be main-scanned in the first collection mode and the subject to be main-scanned in the second collection mode may be the same or different. The main scan performed in the first collection mode is an example of the first scan. The main scan executed in the second collection mode is an example of the second scan.

FIG. 17 is a diagram showing an example of the stored contents of the memory 41 according to the third embodiment. For example, in the third embodiment, as shown in FIG. 17, the image generation function 443 stores the raw data 56 and the raw data 58 in the memory 41 as in the first embodiment. Further, in the third embodiment, the image generation function 443 stores the raw data 65 collected by the main scan in the second collection mode in the memory 41.

That is, the memory 41 stores the raw data 56, which is a part of the raw data 55 corresponding to the first spatial resolution collected by executing the main scan on the subject in the first collection mode. Further, the memory 41 stores the raw data 58 acquired by performing downsampling on at least a part of the raw data 55. In addition, the memory 41 stores the raw data 65 corresponding to the second spatial resolution collected by executing the main scan on the subject in the second collection mode. The raw data 65 is an example of the third data.

Next, with reference to FIG. 18, an example of a processing flow executed by the processing circuit 44 when the trained model 53, the raw data 56, the raw data 58, and the raw data 65 are stored in the memory 41 will be described. .. FIG. 18 is a flowchart showing an example of the flow of processing executed by the processing circuit 44 according to the third embodiment. The process shown in FIG. 18 is when the input interface 43 receives a request to generate CT image data when the trained model 53, the raw data 56, the raw data 58, and the raw data 65 are stored in the memory 41. Will be executed.

In the third embodiment, there are three requirements for generating CT image data. The first requirement is the first requirement described above, and the second requirement is the second requirement described above. The third request is a request to generate CT image data based on the raw data 65. In the following description, the third requirement is referred to as the fifth requirement.

As shown in FIG. 18, the image generation function 443 determines whether or not the received request is the first request, as in the first embodiment (step S100). When the received request is the first request (step S100: Yes (first request)), the image generation function 443 has the first spatial resolution based on the raw data 56, as in the first embodiment. The CT image data corresponding to is generated (step S101). Then, the image processing function 444 generates image data for display from the generated CT image data, and the output function 445 causes the image indicated by the image data for display to be displayed on the display 42 (step S102), and the process is performed. To finish.

On the other hand, when the received request is not the first request (step S100: No), the image generation function 443 determines whether or not the received request is the second request (step S300). When the received request is the second request (step S300: Yes (second request)), the image generation function 443, the image processing function 444, and the output function 445 execute steps S301 to S303. Steps S301 to S303 are the same processes as steps S201 to S203 in the first embodiment.

Specifically, the image generation function 443 uses the trained model 53 to generate the first spatial resolution raw data 60 from the second spatial resolution raw data 58 obtained by downsampling (). Step S301). Then, the image generation function 443 generates CT image data based on the raw data 60 (step S302). Then, the image processing function 444 generates image data for display from the generated CT image data, and the output function 445 causes the image indicated by the image data for display to be displayed on the display 42 (step S303), and the process is performed. To finish.

On the other hand, when the received request is not the second request (step S300: No (fifth request)), the received request is the fifth request. Therefore, the image generation function 443 uses the trained model 53 to obtain the first spatial resolution raw data from the second spatial resolution raw data 65 collected by the main scan in the second acquisition mode. Generate (step S304). Specifically, the image generation function 443 inputs the raw data 65 into the trained model 53 to cause the trained model 53 to generate and output the raw data having the first spatial resolution. Then, the image generation function 443 acquires the generated (output) raw data. In this way, the image generation function 443 generates raw data.

Then, the image generation function 443 generates CT image data based on the generated raw data (step S305). Then, the image processing function 444 generates image data for display from the generated CT image data, and the output function 445 causes the image indicated by the image data for display to be displayed on the display 42 (step S306), and the process is performed. To finish.

That is, in the third embodiment, the image generation function 443 uses the common trained model 53 for the CT image data and the case where the second request is received and the case where the fifth request is received. Generate image data for display. That is, the trained model 53 used by the image generation function 443 is the same when the second request is received and when the fifth request is received. Therefore, the free space of the memory 41 can be increased as compared with the case where the memory 41 stores the trained model for each request.

The third embodiment has been described above. According to the third embodiment, the free space of the memory 41 can be increased as compared with the case where the memory 41 stores the trained model for each request. Further, according to the third embodiment, the same effect as that of the first embodiment can be obtained.

In each of the above-described embodiments and modifications, the case where the X-ray CT devices 1 and 1a execute various processes has been described. However, similar to the X-ray CT apparatus 1,1a, other medical diagnostic imaging apparatus capable of generating the first spatial resolution data and the second spatial resolution data lower than the first spatial resolution is available. , The same processing as the various processing executed by the X-ray CT apparatus 1 may be executed. For example, a medical image diagnostic device such as an MRI device or an X-ray CT device provided with a photon counting type detector may perform the same processing as the various processes performed by the X-ray CT device 1.

Each component of each device according to the above-described embodiment is a functional concept, and does not necessarily have to be physically configured as shown in the figure. That is, the specific form of distribution / integration of each device is not limited to the one shown in the figure, and all or part of the device is functionally or physically distributed in arbitrary units according to various loads and usage conditions. Can be integrated and configured. Further, each processing function performed by each device may be realized by a CPU and a program analyzed and executed by the CPU, or may be realized as hardware by wired logic.

Further, the various processes described in the above-described embodiment can be realized by executing a program prepared in advance on a computer such as a personal computer or a workstation. This program can be distributed via a network such as the Internet. The program can also be executed by being recorded on a computer-readable recording medium such as a hard disk, flexible disk (FD), CD-ROM, MO, or DVD, and read from the recording medium by the computer.

According to at least one embodiment or modification described above, it is possible to ensure image quality that does not interfere with image interpretation while reducing the data size of the raw data stored in the memory 41.

Although some 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 embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the gist of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, as well as in the scope of the invention described in the claims and the equivalent scope thereof.

1 X-ray CT device 41 Memory 443 Image generation function

Claims (10)

Partial data that is part of the first data corresponding to the first resolution collected by performing a scan on the subject is stored and downsampled to at least a part of the first data. A storage unit that stores the second data corresponding to the second resolution lower than the first resolution, which is acquired by executing the above.
When a request to generate image data based on the second data is received, the image is imaged using a trained model that generates third data corresponding to the first resolution based on the second data. When a request to generate data and generate image data based on the partial data is received, an image generation unit that generates image data without using the trained model and an image generation unit.
A medical processing system. The image generation unit
When a request for generating image data corresponding to the first resolution is received based on the second data, the trained model is used to generate the image data.
When a request for generating image data corresponding to the second resolution is received based on the second data, the image data is generated without using the trained model.
The medical processing system according to claim 1. The partial data is obtained from the analysis of the first image data based on the first data, the analysis of the second image data of the subject different from the first image data, and the analysis of the first image data from the user. Is identified from the first data by the operation of, the user's operation on the second image data, the scan protocol selected when performing the scan, or information about a predetermined data collection position. ,
The medical processing system according to claim 1 or 2. The medical use according to any one of claims 1 to 3, wherein the first data, the second data, and the third data are data corresponding to the intensity of X-rays transmitted through the subject. Processing system. The first resolution is a first spatial resolution, and the second resolution is a second spatial resolution lower than the first spatial resolution, according to any one of claims 1 to 4. The medical processing system described. Lower than the first resolution obtained by performing downsampling on at least a portion of the first data corresponding to the first resolution collected by performing a scan on the subject. A storage unit that stores the second data in association with the second trained model acquired by learning the first trained model using the second data corresponding to the second resolution. ,
When a request to generate image data based on the second data is received, the image is obtained by applying the second data to the second trained model associated with the second data. An image generator that generates data and
A medical processing system. The first scan obtained by performing a downsampling on at least a portion of the first data corresponding to the first resolution collected by performing the first scan on the subject. The second data corresponding to the second resolution lower than the resolution is stored, and the third data corresponding to the second resolution collected by performing the second scan on the subject is stored. Memory and
A common trained model is used between the case where the request for generating image data based on the second data is accepted and the case where the request for generating image data based on the third data is accepted. An image generator that generates image data and
A medical processing system. On the computer
More than the first resolution obtained by performing downsampling on at least a portion of the first data corresponding to the first resolution collected by performing a scan on the subject. Learning to generate the third data corresponding to the first resolution based on the second data when the request to generate the image data based on the second data corresponding to the lower second resolution is received. Generate image data using the ready model
When a request for generating image data based on partial data that is a part of the first data is received, the image data is generated without using the trained model.
A program to execute processing. On the computer
Lower than the first resolution obtained by performing downsampling on at least a portion of the first data corresponding to the first resolution collected by performing a scan on the subject. The second data is associated with the second trained model acquired by learning the first trained model using the second data corresponding to the second resolution and stored in the storage unit. ,
When a request to generate image data based on the second data is received, the image is obtained by applying the second data to the second trained model associated with the second data. Generate data,
A program to execute processing. On the computer
The first scan obtained by performing a downsampling on at least a portion of the first data corresponding to the first resolution collected by performing the first scan on the subject. Accepts a request to generate image data based on second data corresponding to a second resolution lower than the resolution
Accepting a request to generate image data based on the third data corresponding to the second resolution collected by performing a second scan on the subject.
An image using a trained model that is common to the case where the request for generating image data based on the second data is accepted and the case where the request for generating image data based on the third data is accepted. Generate data,
A program to execute processing.

Images