JP2021104193A - Image processing device, program and method - Google Patents

Abstract

To obtain information indicating a contour with the excellent accuracy.SOLUTION: An image processing device according to an embodiment comprises: a storage unit; and an output unit. The storage unit stores a learned model that outputs information indicating a contour of a detection object by receiving the input of second image data obtained by processing a window condition on the basis of a luminance value of the detection object of the first image data of a target position of the detection object. The output unit outputs the information indicating the contour of the detection object on the basis of the second image data by using the learned model.SELECTED DRAWING: Figure 1

Description

Embodiments of the present invention relate to image processing devices, programs and methods.

There are image processing devices and medical diagnostic imaging devices that analyze the site of the subject. For example, based on the image data of a plurality of time phases in which the outline of the myocardium is drawn, a myocardial perfusion is performed to obtain the blood flow in the heart muscle of the subject to which the contrast medium is injected. There are an image processing device and an X-ray CT (Computed Tomography) device. Myocardial perfusion generates a time-density curve for the contrast medium on a pixel-by-pixel or local basis in the myocardial region, and the amount of blood flow that moves during the period from the inflow of the contrast medium to the outflow of the contrast medium. Is an analysis that is calculated for each pixel or each local area based on the time density curve. In various analyzes including such myocardial perfusion, in order to perform the analysis with high accuracy, it is desired that the position and shape of the contour of the analysis target site used in the analysis have good accuracy. That is, in order to perform the analysis with high accuracy, it is desired to obtain information showing an accurate contour.

Japanese Unexamined Patent Publication No. 2016-116843 Japanese Unexamined Patent Publication No. 2010-20742 Japanese Unexamined Patent Publication No. 2018-20215

The problem to be solved by the present invention is to obtain information showing an accurate contour.

The image processing device of the embodiment includes a storage unit and an output unit. The storage unit inputs the outline of the detection target by inputting the second image data of the first image data of the target position of the detection target, which is the window condition processed based on the brightness value of the detection target. Store the trained model that outputs the indicated information. The output unit uses the trained model to output information indicating the outline of the detection target based on the second image data.

FIG. 1 is a diagram showing an example of a configuration of a medical processing system according to the first embodiment. FIG. 2 is a diagram for explaining an example of data transmitted by the X-ray CT apparatus according to the first embodiment. FIG. 3 is a diagram for explaining an example of data transmitted by the X-ray CT apparatus according to the first embodiment. FIG. 4 is a diagram showing an example of CT image data according to the first embodiment. FIG. 5 is a diagram for explaining an example of processing executed by the learning data set generation function according to the first embodiment. FIG. 6 is a diagram for explaining an example of processing executed by the learning data set generation function according to the first embodiment. FIG. 7 is a diagram showing an example of window conditions according to the first embodiment. FIG. 8 is a diagram for explaining an example of processing executed by the learning data set generation function according to the first embodiment. FIG. 9 is a diagram for explaining an example of processing executed by the trained model generation function according to the first embodiment. FIG. 10 is a diagram for explaining an example of processing executed by the estimation function according to the first embodiment. FIG. 11 is a flowchart showing an example of the flow of processing executed by the image processing apparatus according to the first embodiment during learning. FIG. 12 is a flowchart showing an example of a flow of processing executed by the image processing apparatus according to the first embodiment during operation. FIG. 13 is a diagram for explaining an example of processing executed by the trained model generation function according to the first modification of the first embodiment. FIG. 14 is a diagram for explaining an example of processing executed by the estimation function according to the first modification of the first embodiment. FIG. 15 is a diagram for explaining an example of processing executed by the image processing apparatus according to the second modification of the first embodiment. FIG. 16 is a diagram for explaining an example of processing executed by the image processing apparatus according to the second modification of the first embodiment. FIG. 17 is a diagram for explaining an example of processing executed by the learning data set generation function according to the second modification of the first embodiment. FIG. 18 is a diagram for explaining an example of processing executed by the learning data set generation function according to the second modification of the first embodiment. FIG. 19 is a diagram for explaining an example of processing executed by the learning data set generation function according to the second modification of the first embodiment. FIG. 20 is a diagram for explaining an example of processing executed by the learning data set generation function according to the second modification of the first embodiment. FIG. 21 is a diagram for explaining an example of processing executed by the estimation function according to the second modification of the first embodiment. FIG. 22 is a diagram showing an example of the data structure of the database according to the second modification of the first embodiment. FIG. 23 is a diagram for explaining an example of processing executed by the learning data set generation function according to the third modification of the first embodiment. FIG. 24 is a diagram showing an example of the configuration of the medical processing system according to the second embodiment. FIG. 25 is a diagram for explaining an example of processing executed by the learning data set generation function 235c according to the second embodiment during operation. FIG. 26 is a diagram for explaining an example of processing executed by the trained model generation function according to the second embodiment. FIG. 27 is a diagram for explaining an example of processing executed by the estimation function according to the second embodiment. FIG. 28 is a flowchart showing an example of the flow of processing executed by the image processing apparatus according to the second embodiment at the time of learning. FIG. 29 is a flowchart showing an example of the flow of processing executed by the image processing apparatus according to the second embodiment during operation. FIG. 30 is a diagram for explaining an example of processing executed by the trained model generation function according to the first modification of the second embodiment. FIG. 31 is a diagram for explaining an example of processing executed by the estimation function according to the first modification of the second embodiment. FIG. 32 is a diagram for explaining an example of processing executed by the image processing apparatus according to the second modification of the second embodiment. FIG. 33 is a diagram for explaining an example of processing executed by the estimation function according to the second embodiment. FIG. 34 is a diagram for explaining an example of processing executed by the image processing apparatus according to the third modification of the second embodiment. FIG. 35 is a diagram for explaining an example of processing executed by the estimation function according to the third modification of the second embodiment. FIG. 36 is a diagram for explaining an example of processing executed by the image processing apparatus according to the modified example 4 of the second embodiment. FIG. 37 is a diagram for explaining an example of the processing executed by the estimation function according to the modified example 4 of the second embodiment. FIG. 38 is a diagram for explaining another example of the analysis target.

Hereinafter, embodiments of an image processing apparatus, a program, and a method will be described in detail with reference to the drawings. The image processing apparatus, program, and method according to the present application are not limited to the embodiments shown below.

(First Embodiment)
FIG. 1 is a diagram showing an example of the configuration of the medical processing system 1 according to the first embodiment. As shown in FIG. 1, the medical processing system 1 includes an X-ray CT device 110, a terminal device 120, and an image processing device 130. Here, the X-ray CT device 110, the terminal device 120, and the image processing device 130 are communicably connected via the network 200.

In the medical processing system 1, there is a stage at the time of learning to generate a trained model and a stage at the time of operation to perform processing using the trained model.

The X-ray CT apparatus 110 captures an image of a subject and generates CT image data showing the CT image. For example, the X-ray CT apparatus 110 includes an X-ray tube, an X-ray detector, a DAS (Data Acquisition System), and a processing circuit. For example, an X-ray tube generates X-rays that irradiate a subject. The detector detects the X-rays that have passed through the subject and outputs an electric signal corresponding to the detected X-rays to the DAS. The DAS amplifies the electric signal output from the detector, converts the amplified electric signal which is an analog signal into a digital signal, generates detection data, and outputs the generated detection data to a processing circuit. The processing circuit performs preprocessing such as logarithmic conversion processing, offset correction processing, sensitivity correction processing between channels, and beam hardening correction processing on the detected data. Preprocessed detection data is referred to as raw data. Then, the processing circuit performs reconstruction processing using a filter correction back projection method, a successive approximation reconstruction method, or the like on the raw data to generate three-dimensional CT image data. CT values are set for each of the plurality of pixels (voxels) of the CT image data. For example, the CT value is a value within the range of "-1000" or more and "1000" or less.

Then, the X-ray CT apparatus 110 transmits various data to the image processing apparatus 130. 2 and 3 are diagrams for explaining an example of data transmitted by the X-ray CT apparatus 110 according to the first embodiment. For example, the X-ray CT apparatus 110 transmits the CT image data 11 for learning to the image processing apparatus 130 at the time of learning when the image processing apparatus 130 generates the learned model, as shown in FIG. As a result, the image processing device 130 generates the trained model using the CT image data 11.

Further, in the operation of the X-ray CT apparatus 110, when the image processing apparatus 130 executes the process of estimating the image data using the trained model, as shown in FIG. 3, the CT image data 12 for the operation is image-processed. It transmits to the device 130. As a result, the image processing device 130 executes processing using the CT image data 12 and the trained model.

The subject imaged when the CT image data 11 is generated and the subject imaged when the CT image data 12 is generated may be different or the same. However, the part imaged when the CT image data 11 is generated and the part imaged when the CT image data 12 is generated are the same. That is, the portion drawn in the CT image data 11 and the portion drawn in the CT image data 12 are the same. Further, for example, a contrast medium is administered to a subject imaged when the CT image data 11 is generated and a subject imaged when the CT image data 12 is generated. Specific examples of the process of generating the trained model at the time of learning executed by the image processing device 130 and the process of using the trained model at the time of operation will be described later.

Returning to the description of FIG. The terminal device 120 is a device for allowing a user such as a doctor or an examination engineer working in a hospital to view an image. For example, the terminal device 120 is realized by a personal computer (PC: Personal Computer), a tablet PC, a PDA (Personal Digital Assistant) operated by a user, a mobile phone, or the like. For example, the terminal device 120 receives various image data transmitted from the X-ray CT device 110 or the image processing device 130. Then, the terminal device 120 displays an image based on the received image data on the display provided in the terminal device 120, and receives various operations on the image via the input interface included in the terminal device 120.

The image processing device 130 acquires various data from the X-ray CT device 110, and executes various processes using the acquired data. For example, the image processing device 130 is realized by a computer device such as a server, a workstation, a personal computer, or a tablet terminal.

The image processing device 130 includes a communication interface 131, a memory 132, an input interface 133, a display 134, and a processing circuit 135.

The communication interface 131 is connected to the processing circuit 135, and performs communication performed between the image processing device 130 and the X-ray CT device 110 and communication performed between the image processing device 130 and the terminal device 120. Control. To give a specific example, the communication interface 131 receives various data and information from the X-ray CT device 110 and the terminal device 120, and transmits the received data and information to the processing circuit 135. For example, the communication interface 131 is realized by a network card, a network adapter, a NIC (Network Interface Controller), or the like.

The memory 132 is connected to the processing circuit 135 and stores various data and information. For example, the memory 132 stores the CT image data 11 and the CT image data 12 transmitted from the X-ray CT apparatus 110.

Further, the memory 132 stores a program for the processing circuit 135 to realize various functions. For example, the memory 132 is realized by a semiconductor memory element such as a flash memory, a hard disk, an optical disk, or the like. The memory 132 is an example of a storage unit.

The input interface 133 is connected to the processing circuit 135 and receives various instructions, requests, and information input operations from the user. For example, the input interface 133 converts the input operation received from the user into an electric signal and transmits the electric signal to the processing circuit 135. For example, the input interface 133 includes a trackball, a switch button, a mouse, a keyboard, 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 a non-optical sensor. It is realized by a contact input circuit, a voice input circuit, and the like. In the present specification, the input interface 133 is not limited to the one provided with physical operating parts such as a mouse and a keyboard. For example, an example of the input interface 133 includes an electric signal processing circuit that receives an electric signal corresponding to an input operation from an external input device provided separately from the device and transmits the electric signal to the processing circuit 135. ..

The display 134 is connected to the processing circuit 135 and displays various information and images. For example, the display 134 converts the information and data transmitted from the processing circuit 135 into an electrical signal for display and outputs it. For example, the display 134 is realized by a liquid crystal monitor, a CRT (Cathode Ray Tube) monitor, a touch panel, or the like.

The processing circuit 135 controls the operation of the image processing device 130 in response to an input operation received from the user via the input interface 133. For example, the processing circuit 135 is realized by a processor.

The example of the configuration of the medical processing system 1 according to the present embodiment has been described above. For example, the image processing device 130 included in the medical processing system 1 is installed in a medical institution such as a hospital or a clinic, and a CT generated by the X-ray CT device 110 is used as a subject for a patient who is admitted to or goes to the medical institution. It is used for various diagnostic imaging using image data. In the present embodiment, the image processing apparatus 130 executes various processes described below so that the contour of the part to be analyzed (for example, the position and shape of the contour with high accuracy) can be obtained with high accuracy. do.

Hereinafter, the details of the image processing apparatus 130 according to the present embodiment will be described. As shown in FIG. 1, the processing circuit 135 of the image processing apparatus 130 has a control function 135a, an acquisition function 135b, a training data set generation function 135c, a trained model generation function 135d, an estimation function 135e, an analysis function 135f, and an output function. Run 135g.

Here, for example, each processing function of the processing circuit 135 is stored in the memory 132 in the form of a program that can be executed by a computer. Each processing function of the control function 135a, the acquisition function 135b, the learning data set generation function 135c, the trained model generation function 135d, the estimation function 135e, the analysis function 135f, and the output function 135g, which are the components of the processing circuit 135 shown in FIG. Is stored in memory 132 in the form of a program that can be executed by a computer. The processing circuit 135 reads each program from the memory 132 and executes each read program to realize a function corresponding to each program. In other words, the processing circuit 135 in the state where each program is read has each function shown in the processing circuit 135 of FIG.

The control function 135a has a communication interface 131, a memory 132, a display 134, and each function 135b of the image processing device 130 so as to execute various instructions and processes corresponding to various requests input via the input interface 133. Control 135g. For example, the control function 135a controls transmission / reception of various data and various information via the communication interface 131, storage of data and information in the memory 132, display of various images on the display 134, and the like.

For example, the control function 135a stores the CT image data 11 transmitted from the X-ray CT apparatus 110 at the time of learning in the memory 132. Further, the control function 135a stores the CT image data 12 transmitted from the X-ray CT apparatus 110 during operation in the memory 132.

The acquisition function 135b acquires various types of data. For example, the acquisition function 135b acquires the CT image data 11 stored in the memory 132 at the time of learning. Further, the acquisition function 135b acquires the CT image data 12 stored in the memory 132 during operation.

FIG. 4 is a diagram showing an example of CT image data 11 according to the first embodiment. The CT image data 11 is data obtained by the X-ray CT apparatus 110 by imaging a three-dimensional region including the left ventricle of the heart of the subject. The CT image data 11 is three-dimensional CT image data.

As shown in FIG. 4, the CT image data 11 depicts the myocardium of the left ventricle 14 of the heart of the subject. In the following description, the case where the site to be analyzed is the myocardium of the left ventricle 14 and the image processing device 130 generates a trained model using the CT image data 11 in which the myocardium of the left ventricle 14 is visualized is taken as an example. I will explain it by listing it. However, when the site to be analyzed is a site other than the myocardium of the left ventricle, the image processing device 130 uses CT image data in which the site other than the myocardium of the left ventricle is visualized, and is the same as the process described below. Processing can be performed to generate a trained model similar to the trained model described below. The site to be analyzed is, for example, an example of an object to be detected.

The training data set generation function 135c generates a training data set used when generating a trained model at the time of training. Hereinafter, an example of a method of generating a set of training data by the learning data set generation function 135c will be described.

The learning data set generation function 135c generates short-axis cross-sectional image data under predetermined window conditions based on the CT image data 11. The window conditions include a window width and a window level.

For example, at the time of learning, the user inputs the position of the cross section (cross-sectional position of the myocardium) of the short-axis cross-sectional image in which the myocardium, which is the site to be analyzed in the left ventricle 14, is drawn, into the processing circuit 135 via the input interface 133. do. As a result, the cross-sectional position of the myocardium is input to the learning data set generation function 135c of the processing circuit 135. Here, the short-axis cross-sectional image is, for example, an image showing a cross-section orthogonal to the long axis of the left ventricle 14.

5 and 6 are diagrams for explaining an example of the processing executed by the learning data set generation function 135c according to the first embodiment. As shown in FIG. 5, the learning data set generation function 135c specifies the cross section 15a specified by the cross-sectional position of the myocardium. The cross section 15a is, for example, a cross section orthogonal to the long axis of the left ventricle 14. Then, as shown in FIG. 6, the learning data set generation function 135c generates the short-axis cross-section image data 15 showing the specified cross-section 15a based on the CT image data 11. For example, when the user inputs the cross-sectional positions of a plurality of myocardiums to the processing circuit 135 for the CT image data 11, the learning data set generation function 135c has a plurality of shorts corresponding to the cross-sectional positions of the plurality of myocardiums. Axial cross-section image data 15 is generated. In this case, each of the plurality of short-axis cross-sectional image data 15 is used in the process described below.

The position of the cross section 15a is, for example, an example of the target position.

The learning data set generation function 135c may automatically generate short-axis cross-sectional image data without accepting the cross-sectional position of the myocardium from the user. For example, the learning data set generation function 135c may detect the long axis of the left ventricle 14 and generate a plurality of short-axis cross-sectional image data 15 along the long axis. In this case, each of the plurality of short-axis cross-sectional image data 15 is used in the process described below.

As shown in FIG. 6, the short-axis cross-sectional image data 15 depicts the myocardium 16 which is the site to be analyzed, the lumen 17 of the left ventricle 14, and the background 18. Further, in the short-axis cross-sectional image data 15, the contour 16a of the outer wall of the myocardium 16 and the contour 16b of the inner wall of the myocardium 16 are depicted.

Luminance values are set for each of the plurality of pixels of the short-axis cross-sectional image data 15. For example, the luminance value is a value within the range of "0" or more and "255" or less.

Here, even when the same part of the same subject is imaged, the tube voltage of the X-ray tube at the time of imaging, the reconstruction conditions, and the concentration of the contrast medium in the subject that changes with the passage of time. Etc. may be different, and the CT value of each voxel constituting the CT image data 11 may be different. It should be noted that the concentration of the contrast medium is different when different parts are imaged even in the same subject. In addition, the concentration of the contrast medium is different even when the same site of different subjects is imaged. The tube voltage of the X-ray tube at the time of imaging, the reconstruction conditions, and the concentration of the contrast medium in the subject are factors that make the CT value different. Therefore, in the following description, the tube voltage of the X-ray tube at the time of imaging, the reconstruction conditions, the concentration of the contrast medium in the subject, and the like may be referred to as "difference factors". Therefore, even when the same part of the same subject is imaged, the brightness value of each pixel constituting the short-axis cross-sectional image data 15 may be different if the difference factor is different. Therefore, if the difference factor is different, the brightness value of each pixel constituting the contours 16a and 16b may also be different. Therefore, if the image processing device 130 executes an analysis such as myocardial perfume on the short-axis cross-sectional image data 15, the analysis result obtained by the analysis can be obtained even if the same site of the same subject is used. , The result may be different depending on the different factors. That is, if the different factors are different, the analysis result may be different even for the same myocardium of the same subject.

Then, a case will be described in which the image processing apparatus 130 uses only the short-axis cross-sectional image data 15 as input data and uses the image data in which the contour of the myocardium is accurately drawn as the teacher data to generate a trained model. In this case, even if the same myocardium of the same subject is used, the position and shape of the contour of the myocardium in the image data estimated by the trained model may be different if the different factors are different. Therefore, in this case, the image processing apparatus 130 cannot generate a trained model having good contour accuracy (for example, contour position and shape accuracy).

Therefore, the image processing device 130 performs the following processing in order to generate a trained model with good accuracy regarding the contour. For example, the learning data set generation function 135c displays the short-axis cross-sectional image based on the short-axis cross-sectional image data 15 on the display 134, and then receives the position of the myocardium from the user. For example, a case where the user specifies one point of the myocardium 16 drawn on the short-axis cross-sectional image displayed on the display 134 via the input interface 133 will be described. In this case, the learning data set generation function 135c accepts the position 19 of the myocardium 16 in the short-axis cross-sectional image data 15 as shown in FIG. The position 19 of the myocardium 16 corresponds to one point of the myocardium 16 designated by the user.

Then, the learning data set generation function 135c identifies the pixel corresponding to the position 19 of the myocardium 16 in the short-axis cross-sectional image data 15. Then, the learning data set generation function 135c specifies the brightness value of the specified pixel. Then, the learning data set generation function 135c specifies the CT value (CT value in the CT image data 11) corresponding to the specified luminance value. For example, the learning data set generation function 135c specifies the CT value corresponding to the specified brightness value based on the window condition used when the short-axis cross-sectional image data 15 is generated. For example, the learning data set generation function 135c may specify the voxels of the CT image data 11 corresponding to the specified pixels and specify the CT value of the specified voxels. In the following description, the CT value specified by the learning data set generation function 135c will be described as “α”.

Then, the training data set generation function 135c determines the window conditions including the window width and the window level. For example, the learning data set generation function 135c determines the window width by the following equation (1).

window_width = k × α (1)

However, in the equation (1), "window_width" indicates the window width. Further, in the equation (1), "k" indicates a coefficient. For example, "k" is a coefficient greater than 0 and less than 1. Further, in the equation (1), "x" is an operator indicating multiplication. The same applies to the following equation.

Then, the learning data set generation function 135c determines the window level by the following equation (2).

window_level ＝ window_width / 2 (2)

However, in the equation (2), "window_level" indicates the window level. Further, in the equation (2), "/" is an operator indicating division. The same applies to the following equation.

Then, the learning data set generation function 135c determines whether or not the window width determined by the above equation (1) is smaller than the predetermined first threshold value. When the learning data set generation function 135c determines that the determined window width is equal to or greater than the predetermined first threshold value, the window level determined by the above equation (2) is the predetermined second threshold value. Determine if it is less than.

Then, when the determination data set generation function 135c determines that the determined window width is smaller than the predetermined first threshold value, the first threshold value is set as the window width and the second threshold value is set as the window level. The window condition to be used is determined as the window condition used in the following processing. Similarly, when the training data set generation function 135c determines that the determined window level is smaller than the predetermined second threshold value, the first threshold value is set as the window width and the second threshold value is set as the window level. The window condition to be used is determined as the window condition used in the following processing.

On the other hand, the training data set generation function 135c determines that the determined window width is equal to or greater than the predetermined first threshold value, and determines that the determined window level is equal to or greater than the predetermined second threshold value. Determines the window condition including the determined window width and the determined window level as the window condition used in the following processing.

FIG. 7 is a diagram showing an example of the window condition 20 according to the first embodiment. In FIG. 7, the horizontal axis represents the CT value of the voxel of the CT image data 11, and the vertical axis represents the brightness value. FIG. 7 shows a case where the training data set generation function 135c determines the window condition 20 including the window width W and the window level L.

In the case shown in the example of FIG. 7, the luminance value corresponding to the CT value “−1000” is “0”. The brightness value corresponding to the CT value of "-1000 + W" or higher is "255". Further, the brightness value Vb corresponding to the CT value Vc larger than "-1000" and smaller than "-1000 + W" is represented by the following equation (3).

Vb = (255 / W) x Vc (3)

FIG. 8 is a diagram for explaining an example of processing executed by the learning data set generation function 135c according to the first embodiment. The learning data set generation function 135c generates short-axis cross-section image data 25 showing the same cross-section as the short-axis cross-section image data 15 from the CT image data 11 based on the determined window conditions.

Here, the coefficient k in the equation (1) is a coefficient obtained by an experiment or a simulation so that the CT value corresponding to the contour of the part to be analyzed is included in the window width W. That is, the coefficient k is a coefficient obtained so that the CT value corresponding to the brightness value set for the pixels constituting the contour (pixels of the image data) is included in the window width W.

For example, the coefficient k is calculated so that two CT values, a CT value corresponding to the contour of the outer wall of the myocardium of the left ventricle and a CT value corresponding to the contour of the inner wall, are included in the window width W. As a result, as shown in FIG. 8, the boundary between the myocardium 16 and the lumen 17 and the boundary between the myocardium 16 and the background 18 are clearly depicted in the short-axis cross-sectional image data 25. That is, in the short-axis cross-sectional image data 25, the difference between the average value of the brightness values of the pixels constituting the myocardium 16 and the average value of the brightness values of the pixels constituting the lumen 17 is equal to or greater than a predetermined value. Further, in the short-axis cross-sectional image data 25, the difference between the average value of the brightness values of the pixels constituting the myocardium 16 and the average value of the brightness values of the pixels constituting the background 18 is also a predetermined value or more. As described above, the short-axis cross-sectional image data 25 is image data in which the boundary between the myocardium 16 and the lumen 17 and the boundary between the myocardium 16 and the background 18 are clearly depicted regardless of the difference factor.

As described above, the training data set generation function 135c newly determines the window condition and is newly determined without using the above-mentioned predetermined window condition used when generating the short-axis cross-sectional image data 15. The short-axis cross-sectional image data 25 is generated by using the window condition. That is, the learning data set generation function 135c generates the short-axis cross-section image data 25 in which the window condition is processed based on the brightness value of the myocardium 16 of the short-axis cross-section image data 15 at the position of the cross-section 15a of the myocardium 16.

The coefficient k may be determined so that the CT value corresponding to at least one of the contour of the outer wall and the contour of the inner wall of the myocardium of the left ventricle is included in the window width W.

The short-axis cross-section image data 15 and the short-axis cross-section image data 25 are used as input data (input images) when the trained model is generated by the trained model generation function 135d.

Further, the training data set generation function 135c generates contour image data to be used as teacher data when generating a trained model. The contour image data is image data in which the contour of the outer wall and the contour of the inner wall of the myocardium in the cross section shown by the short-axis cross-sectional image data 15 are accurately depicted. An example of a method of generating contour image data will be described.

For example, the learning data set generation function 135c displays the short-axis cross-sectional image based on the short-axis cross-sectional image data 15 on the display 134, and then the user displays the accurate positions and shapes of the contours 16a and 16b. Accepts contour information indicating. Explaining by giving a specific example, the user operates the input interface 133 while checking the displayed contours 16a and 16b, and processes the contour information indicating the accurate position and shape of the contours 16a and 16b in the processing circuit 135. Enter in.

When the learning data set generation function 135c receives the contour information, the contours 16a and 16b generate the contour image data 26 drawn at an accurate position and in an accurate shape based on the contour information.

The trained model generation function 135d generates a trained model at the time of training. FIG. 9 is a diagram for explaining an example of processing executed by the trained model generation function 135d according to the first embodiment. FIG. 9 shows an example of a method of generating the trained model 28 at the time of learning of the first embodiment. As shown in FIG. 9, at the time of training, the trained model generation function 135d learns the relationship between the set of the short-axis cross-sectional image data 15 and the short-axis cross-sectional image data 25 and the contour image data 26 to train the trained model. 28 is generated.

As described above, the trained model generation function 135d generates the trained model 28 by learning the set of the short-axis cross-section image data 15 and the short-axis cross-section image data 25 in association with the contour image data 26. For example, at the time of learning, the image processing device 130 receives a plurality of CT image data 11. Then, the learning data set generation function 135c generates one set of the short-axis cross-sectional image data 15, the short-axis cross-sectional image data 25, and the contour image data 26 for each CT image data 11. That is, the learning data set generation function 135c generates a plurality of sets for the plurality of CT image data 11. Then, the trained model generation function 135d generates the trained model 28 using a plurality of sets.

For example, the trained model generation function 135d performs machine learning by inputting the short-axis cross-sectional image data 15 and the short-axis cross-sectional image data 25 as input data and the contour image data 26 as training data to the machine engine. For example, machine learning engines include Deep Learning, Neural Network, Logistic Regression Analysis, Nonlinear Discrimination Analysis, Support Vector Machine (SVM), Random Forest, etc. Machine learning is performed using various algorithms such as Naive Bays.

The trained model generation function 135d generates a trained model 28 as a result of such machine learning. The trained model 28 estimates the image data corresponding to the contour image data 26 by inputting a set of the image data corresponding to the short-axis cross-sectional image data 15 and the image data corresponding to the short-axis cross-sectional image data 25 ( Generate) and output.

Then, the trained model generation function 135d stores the generated trained model 28 in the memory 132.

The estimation function 135e uses the trained model 28 to estimate highly accurate image data regarding the contour of the part to be analyzed during operation, and outputs the estimated image data. An example of the processing executed by the estimation function 135e will be described with reference to FIG. FIG. 10 is a diagram for explaining an example of processing executed by the estimation function 135e according to the first embodiment. FIG. 10 shows an example of an estimation method of the contour image data 32 at the time of operation of the first embodiment.

The estimation function 135e uses the CT image data 12 acquired by the acquisition function 135b during operation to generate the short-axis cross-sectional image data 30 shown in FIG. For example, the estimation function 135e is the same method as the method in which the learning data set generation function 135c uses the CT image data 11 to generate the short-axis cross-sectional image data 15 at the time of learning, and the short-axis cross section is performed using the CT image data 12. Image data 30 is generated. For example, one or more short-axis cross-sectional image data 30 is generated from one CT image data 12. When a plurality of short-axis cross-sectional image data 30s are generated, each of the plurality of short-axis cross-sectional image data 30 is used in the process described below.

Similar to the short-axis cross-sectional image data 15, for example, the short-axis cross-sectional image data 30 depicts the myocardium, the lumen of the left ventricle, and the background, which are the sites to be analyzed. The short-axis cross-section image data 30 is image data corresponding to the short-axis cross-section image data 15. The short-axis cross-sectional image data 30 is image data of the position of the cross-section 15a of the myocardium 16 which is the site to be analyzed. The short-axis cross-sectional image data 30 is, for example, an example of the first image data.

Then, the estimation function 135e uses the short-axis cross-sectional image data 30 to generate the short-axis cross-sectional image data 31 shown in FIG. For example, the estimation function 135e uses the short-axis cross-section image data 30 in the same manner as the method in which the learning data set generation function 135c uses the short-axis cross-section image data 15 to generate the short-axis cross-section image data 25 during training. Generates short-axis cross-sectional image data 31.

In the short-axis cross-sectional image data 31, for example, as in the short-axis cross-sectional image data 25, the difference between the average value of the brightness values of the pixels constituting the myocardium and the average value of the brightness values of the pixels constituting the portion other than the myocardium. Is greater than or equal to the predetermined value. The short-axis cross-section image data 31 is image data corresponding to the short-axis cross-section image data 25. The short-axis cross-section image data 31 is short-axis cross-section image data obtained by processing window conditions based on the brightness value of the myocardium 16 of the short-axis cross-section image data 30 at the position of the cross-section 15a of the myocardium 16. The short-axis cross-sectional image data 31 is, for example, an example of the second image data.

Then, as shown in FIG. 10, the estimation function 135e causes the trained model 28 to estimate the contour image data 32 by inputting the short-axis cross-section image data 30 and the short-axis cross-section image data 31 into the trained model 28. To output. That is, the estimation function 135e estimates and outputs the contour image data 32 based on the short-axis cross-section image data 30 and the short-axis cross-section image data 31 by using the trained model 28. For example, the estimation function 135e outputs the contour image data 32 to the analysis function 135f. The contour image data 32 is image data corresponding to the contour image data 26. The contour image data 32 is, for example, an example of information indicating the contour 16a and the contour 16b of the myocardium 16. The estimation function 135e is an example of an output unit.

Here, the short-axis cross-sectional image data 31 is image data in which the boundary between the myocardium and a portion other than the myocardium is clearly depicted regardless of the difference factor. Further, the accuracy regarding the contour of the myocardium in the contour image data 32 is higher than the accuracy regarding the contour of the myocardium in the short-axis cross-sectional image data 30.

Therefore, the trained model 28 according to the first embodiment can estimate and output the contour image data 32 in which the contour of the myocardium, which is the site to be analyzed, is accurately depicted regardless of the difference factor. That is, according to the first embodiment, it is possible to obtain information indicating the contour of the part to be analyzed with high accuracy regardless of the difference factor.

The analysis function 135f performs various analyzes using the image data output from the trained model. For example, the analysis function 135f performs analysis using the contour of the myocardium to be analyzed drawn in the contour image data 32. An example of such an analysis is myocardial perfusion performed based on contour image data 32 of a plurality of time phases.

The output function 135g displays an image based on various image data, an analysis result obtained by analysis by the analysis function 135f, and the like on the display 134.

FIG. 11 is a flowchart showing an example of a flow of processing executed by the image processing apparatus 130 according to the first embodiment during learning. The process shown in FIG. 11 is executed, for example, when the user inputs to the processing circuit 135 an instruction for executing the process of generating the trained model 28 via the input interface 133.

As shown in FIG. 11, the acquisition function 135b acquires the CT image data 11 stored in the memory 132 (step S101). Then, the learning data set generation function 135c generates the short-axis cross-sectional image data 15 based on the CT image data 11 (step S102).

Then, the learning data set generation function 135c displays the short-axis cross-sectional image based on the short-axis cross-sectional image data 15 on the display 134 (step S103). Then, the learning data set generation function 135c determines whether or not the position 19 of the myocardium 16 has been accepted from the user (step S104). When the learning data set generation function 135c determines that the position 19 of the myocardium 16 is not accepted (step S104: No), the determination in step S104 is performed again.

When the learning data set generation function 135c determines that the position 19 of the myocardium 16 has been accepted (step S104: Yes), it specifies the brightness value of the pixel at the position 19 of the myocardium 16 in the short-axis cross-sectional image data 15 (step S104: Yes). Step S105). Then, the learning data set generation function 135c specifies the CT value “α” in the CT image data 11 corresponding to the specified luminance value (step S106).

Then, the learning data set generation function 135c determines the window condition including the window width W and the window level L by using the above equations (1) and (2) and the first threshold value and the second threshold value. (Step S107).

Then, the learning data set generation function 135c generates the short-axis cross-section image data 25 showing the same cross-section as the short-axis cross-section image data 15 from the CT image data 11 based on the determined window conditions. (Step S108).

Then, the learning data set generation function 135c determines whether or not the contour information has been received from the user (step S109). When the learning data set generation function 135c determines that the contour information is not accepted (step S109: No), the determination in step S109 is performed again.

When the learning data set generation function 135c determines that the contour information has been received (step S109: Yes), the contours 16a and 16b are drawn at the correct positions and in the correct shape based on the contour information. Image data 26 is generated (step S110).

Then, the trained model generation function 135d generates the trained model 28 by learning the set of the short-axis cross-section image data 15 and the short-axis cross-section image data 25 in association with the contour image data 26 (step S111). ). Then, the trained model generation function 135d stores the generated trained model 28 in the memory 132 (step S112), and ends the process.

FIG. 12 is a flowchart showing an example of a flow of processing executed by the image processing apparatus 130 according to the first embodiment during operation. The process shown in FIG. 12 is executed, for example, when the user inputs an instruction for executing the analysis by the analysis function 135f to the processing circuit 135 via the input interface 133.

As shown in FIG. 12, the acquisition function 135b acquires the CT image data 12 stored in the memory 132 (step S201). Then, the estimation function 135e generates the short-axis cross-sectional image data 30 using the CT image data 12 (step S202).

Then, the estimation function 135e generates the short-axis cross-section image data 31 using the short-axis cross-section image data 30 (step S203).

Then, the estimation function 135e estimates and outputs the contour image data 32 based on the short-axis cross-section image data 30 and the short-axis cross-section image data 31 using the trained model 28 (step S204).

Then, the analysis function 135f performs analysis using the contour of the myocardium to be analyzed drawn in the contour image data 32 (step S205). Then, the output function 135g displays the analysis result on the display 134 (step S206), and ends the process.

The first embodiment has been described above. According to the first embodiment, as described above, it is possible to obtain information indicating the contour of the part to be analyzed with high accuracy.

(Modification 1 of the first embodiment)
In the first embodiment, the case where the trained model generation function 135d generates the trained model 28 has been described. However, the trained model generation function 135d may generate another trained model instead of the trained model 28. Therefore, such a modification will be described as a modification 1 of the first embodiment. In the following description of the first embodiment, the points different from the first embodiment will be mainly described, and the description of the same points as the first embodiment may be omitted. Further, the same components as those in the first embodiment may be designated by the same reference numerals and the description thereof may be omitted.

FIG. 13 is a diagram for explaining an example of processing executed by the trained model generation function 135d according to the first modification of the first embodiment. FIG. 13 shows an example of a method of generating the trained model 28a at the time of learning the modified example 1. As shown in FIG. 13, at the time of learning, the trained model generation function 135d generates the trained model 28a by learning the relationship between the short-axis cross-sectional image data 25 and the contour image data 26.

In this way, the trained model generation function 135d generates the trained model 28a by learning the short-axis cross-sectional image data 25 and the contour image data 26 in association with each other. For example, at the time of learning, the learning data set generation function 135c generates a plurality of sets of the short-axis cross-sectional image data 25 and the contour image data 26. Then, the trained model generation function 135d generates the trained model 28a using the generated plurality of sets.

For example, the trained model generation function 135d performs machine learning by inputting the short-axis cross-sectional image data 25 as input data and the contour image data 26 as training data into the machine engine. The trained model generation function 135d generates a trained model 28a as a result of such machine learning.

The trained model 28a estimates (generates) and outputs image data corresponding to the contour image data 26 by inputting image data corresponding to the short-axis cross-sectional image data 25.

Then, the trained model generation function 135d stores the generated trained model 28a in the memory 132.

The estimation function 135e according to the first modification uses the trained model 28a to estimate highly accurate image data regarding the contour of the part to be analyzed, and outputs the estimated image data. An example of the process executed by the estimation function 135e will be described with reference to FIG. FIG. 14 is a diagram for explaining an example of processing executed by the estimation function 135e according to the first modification of the first embodiment. FIG. 14 shows an example of an estimation method of the contour image data 32a during the operation of the modified example 1.

The estimation function 135e uses the CT image data 12 acquired by the acquisition function 135b during operation to generate short-axis cross-sectional image data 30 (see FIG. 10) as in the first embodiment. Then, the estimation function 135e uses the short-axis cross-sectional image data 30 to generate the short-axis cross-sectional image data 31 as in the first embodiment. The short-axis cross-section image data 31 is image data corresponding to the short-axis cross-section image data 25.

Then, as shown in FIG. 14, the estimation function 135e inputs the short-axis cross-sectional image data 31 into the trained model 28a, so that the trained model 28a estimates the contour image data 32a and outputs it. That is, the estimation function 135e estimates and outputs the contour image data 32a based on the short-axis cross-sectional image data 31 using the trained model 28a. The contour image data 32a is image data corresponding to the contour image data 26. The contour image data 32a is, for example, an example of information indicating the contour 16a and the contour 16b of the myocardium 16.

Here, the short-axis cross-sectional image data 31 is image data in which the boundary between the myocardium and a portion other than the myocardium is clearly depicted regardless of the difference factor. Further, the accuracy regarding the contour of the myocardium in the contour image data 32a is higher than the accuracy regarding the contour of the myocardium in the short-axis cross-sectional image data 31.

Therefore, the trained model 28a according to the first modification can estimate and output the contour image data 32a in which the contour of the myocardium, which is the site to be analyzed, is accurately depicted regardless of the difference factor. That is, according to the first modification, it is possible to obtain information indicating the contour of the part to be analyzed with high accuracy regardless of the difference factor.

The modification 1 has been described above. According to the first modification, as described above, it is possible to obtain information indicating the contour of the part to be analyzed with high accuracy regardless of the difference factor, as in the first embodiment.

Further, in the first embodiment described above, the short-axis cross-section image data 15 and the short-axis cross-section image data 25 are used as input data when the trained model 28 is generated. On the other hand, in the modified example 1, the short-axis cross-sectional image data 25 is used as the input data when the trained model 28a is generated. Therefore, the modification example 1 has a smaller number and amount of image data used as input data when generating the trained model than the first embodiment. Therefore, according to the first modification, the convergence of machine learning when generating the trained model can be accelerated.

Further, it is conceivable that the short-axis cross-sectional image data 15 may be deleted from the image processing device 130 for some reason. However, according to the first modification, even in such a case, if the short-axis cross-sectional image data 25 is stored (held and stored) in the image processing device 130, the trained model 28a can be generated. can.

Similarly, it is conceivable that the short-axis cross-sectional image data 30 may be deleted from the image processing device 130 for some reason. According to the first modification, even in such a case, if the short-axis cross-sectional image data 31 is stored in the image processing device 130, the contour image data 32a can be estimated and output.

(Modification 2 of the first embodiment)
In the first embodiment, the case where the image processing device 130 generates the trained model 28 by using the short-axis cross-sectional image data 25 in which the contours 16a and 16b of the myocardium 16 are clearly depicted has been described. However, the image processing apparatus 130 may generate a trained model by using other image data instead of the short-axis cross-sectional image data 25. Therefore, such a modification will be described as a modification 2 of the first embodiment. In the following description of the modified example 2, the points different from the first embodiment will be mainly described, and the description of the same points as the first embodiment may be omitted. Further, the same components as those in the first embodiment may be designated by the same reference numerals and the description thereof may be omitted.

FIG. 15 is a diagram for explaining an example of processing executed by the image processing apparatus 130 according to the second modification of the first embodiment. FIG. 15 shows an example of a method of generating the trained model 28b at the time of learning the modified example 2. The learning data set generation function 135c according to the second modification generates the color image data 35 shown in FIG. RGB values are set for each of the plurality of pixels constituting the color image data 35.

R (R value) corresponds to red, G (G value) corresponds to green, and B (B value) corresponds to blue. For example, the RGB value set for each pixel is represented by "(R, G, B)". Each of R, G, and B is, for example, a value within the range of "0" or more and "255" or less.

An example of a method of determining G set in the pixels of the color image data 35 will be described. FIG. 16 is a diagram for explaining an example of processing executed by the image processing apparatus 130 according to the second modification of the first embodiment. For example, the learning data set generation function 135c displays the short-axis cross-sectional image based on the short-axis cross-sectional image data 15 on the display 134, and then receives the position 40 of the myocardium 16 from the user as shown in FIG. Further, the learning data set generation function 135c receives the position 41 of the lumen 17 and the position 42 of the background 18 from the user.

For example, a case where the user specifies one point of the myocardium 16, one point of the lumen 17, and one point of the background 18 drawn on the short-axis cross-sectional image via the input interface 133 will be described. In this case, as shown in FIG. 16, the learning data set generation function 135c accepts the position 40 of the myocardium 16 in the short-axis cross-sectional image data 15, the position 41 of the lumen 17, and the position 42 of the background 18. The position 40 of the myocardium 16 corresponds to one point of the designated myocardium 16. The position 41 of the lumen 17 corresponds to one point in the designated lumen 17. The position 42 of the background 18 corresponds to one point of the designated background 18.

Then, the learning data set generation function 135c identifies the pixel corresponding to the position 40 of the myocardium 16 in the short-axis cross-sectional image data 15. Then, the learning data set generation function 135c specifies the brightness value of the specified pixel. Then, the learning data set generation function 135c specifies the CT value (CT value in the CT image data 11) corresponding to the specified luminance value. For example, the learning data set generation function 135c specifies the CT value corresponding to the specified brightness value based on the window condition used when the short-axis cross-sectional image data 15 is generated. For example, the learning data set generation function 135c may specify the voxels of the CT image data 11 corresponding to the specified pixels and specify the CT value of the specified voxels. In this way, the learning data set generation function 135c identifies the CT value corresponding to the position 40 of the myocardium 16. In the following description, the CT value corresponding to the position 40 of the myocardium 16 will be described as “β”.

Further, the learning data set generation function 135c has a CT value corresponding to the position 41 of the lumen 17 and a background 18 in the same manner as the method for specifying the CT value “β” corresponding to the position 40 of the myocardium 16. The CT value corresponding to the position 42 of is specified. In the following description, the CT value corresponding to the position 41 of the lumen 17 will be described as “γ”, and the CT value corresponding to the position 42 of the background 18 will be described as “δ”.

Then, the learning data set generation function 135c determines the following window conditions in order to determine G set in the pixels of the color image data 35. For example, in the learning data set generation function 135c, the brightness value "255" corresponds to the CT value in the predetermined range including the CT value "β", and the brightness value "0" corresponds to the CT value in the predetermined range including the CT value "γ". Corresponds to, and determines the window condition in which the brightness value “0” corresponds to the CT value in a predetermined range including the CT value “δ”.

FIG. 17 is a diagram for explaining an example of processing executed by the learning data set generation function 135c according to the second modification of the first embodiment. For example, the learning data set generation function 135c determines the window condition 20a shown in FIG. 17 in order to determine G set in the pixels of the color image data 35. In FIG. 17, the horizontal axis represents the CT value of the voxel of the CT image data 11, and the vertical axis represents the G value.

The window condition 20a is indicated by a polygonal line 45. The polygonal line 45 is composed of a line segment 45a, a line segment 45b, a line segment 45c, a line segment 45d, and a line segment 45e. In FIG. 17, the horizontal axis and the line segments 45b and 45c overlap, but the range from one end to the other end of the line segment 45b is the range indicated by the reference numeral “45b_1”. The range from one end to the other end of the line segment 45c is the range indicated by the reference numeral "45c_1".

The line segment 45a indicates that the G value “255” corresponds to the CT value in a predetermined range including the CT value “β”. Further, the line segment 45b indicates that the G value "0" corresponds to the CT value in a predetermined range including the CT value "γ". Further, the line segment 45c indicates that the G value “0” corresponds to the CT value in a predetermined range including the CT value “δ”.

Further, the line segment 45d is a line segment connecting both ends of the line segment 45a on the side having a large CT value and both ends of the line segment 45b on the side having a small CT value. The line segment 45d indicates the relationship between the CT value and the G value that the G value decreases as the CT value increases within the range of the CT value corresponding to the line segment 45d.

Further, the line segment 45e is a line segment connecting both ends of the line segment 45a on the side having a small CT value and both ends of the line segment 45c on the side having a large CT value. The line segment 45e shows the relationship between the CT value and the G value that the G value also increases as the CT value increases within the range of the CT value corresponding to the line segment 45e.

FIG. 18 is a diagram for explaining an example of processing executed by the learning data set generation function 135c according to the second modification of the first embodiment. As shown in FIG. 18, the training data set generation function 135c generates color image data 46 showing the same cross section as the cross section shown by the short axis cross section image data 15 from the CT image data 11 based on the window condition 20a. do. RGB values are set for each of the plurality of pixels 46a constituting the color image data 46.

However, as shown in FIG. 18, the R value and the B value set in each pixel 46a of the color image data 46 are “0”. The G value set for each pixel 46a of the color image data 46 is a value based on the window condition 20a. That is, as shown in FIG. 18, the RGB values set for each pixel 46a are represented by (0, G, 0).

Here, in the color image data 46 generated based on the window condition 20a, the G value "255" is set for the pixel 46a corresponding to the myocardium 16, and the G value "255" is set for the pixel 46a corresponding to the lumen 17 and the background 18. 0 ”is set. Therefore, it can be said that the color image data 46 is image data showing the characteristics of the myocardium 16. More specifically, it can be said that the G value set in each pixel 46a of the color image data 46 indicates the characteristics of the myocardium 16.

Next, an example of a method of determining R set in the pixels of the color image data 35 will be described. The learning data set generation function 135c uses the position 41 of the lumen 17 to color in the same manner as the method of determining the G value set for each pixel of the color image data 35 using the position 40 of the myocardium 16. The R value set for each pixel of the image data 35 is determined. For example, the learning data set generation function 135c uses the position 41 of the lumen 17 in the same manner as the method of determining the window condition 20a showing the relationship between the CT value and the G value using the position 40 of the myocardium 16. Then, the window condition showing the relationship between the CT value and the R value is determined.

FIG. 19 is a diagram for explaining an example of processing executed by the learning data set generation function 135c according to the second modification of the first embodiment. The learning data set generation function 135c is a window showing the relationship between the CT value and the R value in the same manner as the method for generating the color image data 46 based on the window condition 20a showing the relationship between the CT value and the G value. Based on the conditions, the color image data 47 shown in FIG. 19 is generated.

RGB values are set for each of the plurality of pixels 47a constituting the color image data 47. However, as shown in FIG. 19, the G value and the B value set in each pixel 47a of the color image data 47 are “0”. Further, the R value set in each pixel 47a of the color image data 47 is a value based on a window condition indicating the relationship between the CT value and the R value. That is, as shown in FIG. 19, the RGB values set for each pixel 47a are represented by (R, 0, 0).

Here, in the color image data 47 generated based on the window condition showing the relationship between the CT value and the R value, the R value “255” is set in the pixel 47a corresponding to the lumen 17, and the myocardium 16 and the background 18 are set. The R value "0" is set in the pixel 47a corresponding to. Therefore, it can be said that the color image data 47 is image data showing the characteristics of the lumen 17. More specifically, it can be said that the R value set in each pixel 47a of the color image data 47 indicates the characteristics of the lumen 17.

Next, an example of a method for determining B to be set in the pixels of the color image data 35 will be described. The training data set generation function 135c is a color image using the position 42 of the background 18 in the same manner as the method of determining the G value set for each pixel of the color image data 35 using the position 40 of the myocardium 16. The B value set for each pixel of the data 35 is determined. For example, the learning data set generation function 135c uses the position 42 of the background 18 in the same manner as the method of determining the window condition 20a showing the relationship between the CT value and the G value using the position 40 of the myocardium 16. , Determine the window condition showing the relationship between the CT value and the B value.

FIG. 20 is a diagram for explaining an example of processing executed by the learning data set generation function 135c according to the second modification of the first embodiment. The learning data set generation function 135c is a window showing the relationship between the CT value and the B value in the same manner as the method for generating the color image data 46 based on the window condition 20a showing the relationship between the CT value and the G value. Based on the conditions, the color image data 48 shown in FIG. 20 is generated.

RGB values are set for each of the plurality of pixels 48a constituting the color image data 48. However, as shown in FIG. 20, the R value and the G value set in each pixel 48a of the color image data 48 are “0”. Further, the B value set in each pixel 48a of the color image data 48 is a value based on a window condition indicating the relationship between the CT value and the B value. That is, as shown in FIG. 20, the RGB values set for each pixel 48a are represented by (0, 0, B).

Here, in the color image data 48 generated based on the window condition showing the relationship between the CT value and the B value, the B value "255" is set in the pixel 48a corresponding to the background 18, and the myocardium 16 and the lumen 17 are set. The B value "0" is set in the pixel 48a corresponding to. Therefore, it can be said that the color image data 48 is image data showing the characteristics of the background 18. More specifically, it can be said that the B value set in each pixel 48a of the color image data 48 indicates the characteristics of the background 18.

Then, the learning data set generation function 135c generates the color image data 35 by synthesizing the color image data 46, the color image data 47, and the color image data 48. As a result, the G value set in each pixel 46a of the color image data 46 is set in each corresponding pixel of the color image data 35. Further, the R value set in each pixel 47a of the color image data 47 is set in each corresponding pixel of the color image data 35. Further, the B value set in each pixel 48a of the color image data 48 is set in each corresponding pixel of the color image data 35.

Therefore, it can be said that the color image data 35 is image data showing the characteristics of the myocardium 16, the characteristics of the lumen 17, and the characteristics of the background 18.

Returning to the explanation of FIG. 15, at the time of learning, the trained model generation function 135d learns the relationship between the pair of the short-axis cross-sectional image data 15 and the color image data 35 and the contour image data 26, so that the trained model 28b To generate.

In this way, the trained model generation function 135d generates the trained model 28b by learning the short-axis cross-sectional image data 15, the color image data 35, and the contour image data 26 in association with each other. For example, at the time of learning, the learning data set generation function 135c generates a plurality of sets of short-axis cross-sectional image data 15, color image data 35, and contour image data 26. Then, the trained model generation function 135d generates the trained model 28b using the generated plurality of sets.

For example, the trained model generation function 135d performs machine learning by inputting the short-axis cross-sectional image data 15 and the color image data 35 as input data and the contour image data 26 as training data to the machine engine. The trained model generation function 135d generates a trained model 28b as a result of such machine learning.

As described above, the trained model generation function 135d generates the trained model 28b using the color image data 35 showing the characteristics of the myocardium 16, the characteristics of the lumen 17, and the characteristics of the background 18. Therefore, the trained model generation function 135d can more accurately generate a trained model 28b capable of discriminating between the myocardium 16 and the lumen 17 and discriminating between the myocardium 16 and the background 18. ..

The trained model 28b estimates (generates) image data corresponding to the contour image data 26 by inputting image data corresponding to the short-axis cross-sectional image data 15 and image data corresponding to the color image data 35. And output.

Then, the trained model generation function 135d stores the generated trained model 28b in the memory 132.

The estimation function 135e according to the second modification uses the trained model 28b to estimate highly accurate image data regarding the contour of the part to be analyzed, and outputs the estimated image data. An example of the process executed by the estimation function 135e will be described with reference to FIG. FIG. 21 is a diagram for explaining an example of processing executed by the estimation function 135e according to the second modification of the first embodiment. FIG. 21 shows an example of a method of estimating the contour image data 50 during the operation of the modified example 2.

The estimation function 135e uses the CT image data 12 acquired by the acquisition function 135b during operation to generate short-axis cross-sectional image data 30 (see FIG. 10) as in the first embodiment. Then, the estimation function 135e is the same method as the method in which the learning data set generation function 135c generated the color image data 35 using the short-axis cross-sectional image data 15, and the color image data using the short-axis cross-sectional image data 30. Generate 49. The color image data 49 is image data corresponding to the color image data 35.

Then, as shown in FIG. 21, the estimation function 135e inputs the short-axis cross-sectional image data 30 and the color image data 49 into the trained model 28b, so that the trained model 28b estimates the contour image data 50 and outputs the data. Let me. That is, the estimation function 135e estimates and outputs the contour image data 50 based on the short-axis cross-sectional image data 30 and the color image data 49 using the trained model 28b. The contour image data 50 is image data corresponding to the contour image data 26. The contour image data 50 is, for example, an example of information indicating the contour 16a and the contour 16b of the myocardium 16.

Here, the estimation function 135e inputs color image data 49 showing the characteristics of the myocardium, the characteristics of the lumen, and the characteristics of the background into the trained model 28b. Therefore, the estimation function 135e can output the contour image data 50 in which the myocardium and the lumen are discriminated and the myocardium and the background are discriminated with more accuracy. Therefore, according to the second modification, it is possible to obtain information showing the contour of the myocardium with high accuracy.

The modification 2 has been described above. According to the second modification, as described above, it is possible to obtain information indicating the contour of the part to be analyzed with high accuracy.

In the second modification, the learning data set generation function 135c has a CT value “β” corresponding to the myocardium 16 and a lumen 17 based on the window conditions used when the short-axis cross-sectional image data 15 is generated. The case of specifying the corresponding CT value “γ” and the CT value “δ” corresponding to the background 18 has been described. Further, the training data set generation function 135c describes a case where the voxels of the CT image data 11 corresponding to the specified pixels are specified and the CT values “β”, “γ” and “δ” of the specified voxels are specified. did. However, the learning data set generation function 135c may specify the CT value corresponding to the myocardium 16, the CT value corresponding to the lumen 17, and the CT value corresponding to the background 18 by other methods. Therefore, hereinafter, an example of the process in which the learning data set generation function 135c specifies the CT value corresponding to the myocardium 16 and the CT value corresponding to the lumen 17 and the CT value corresponding to the background 18 by another method will be described. do.

FIG. 22 is a diagram showing an example of the data structure of the database 51 according to the second modification of the first embodiment. The database 51 shown in FIG. 22 is stored in the memory 132. A plurality of records are registered in the database 51. Each record has a "target" item and a "CT value" item.

In the item of "object", an identifier (ID (Identification)) of an object for which the image processing apparatus 130 intends to obtain a CT value is registered. For example, as shown in FIG. 22, in the item of "object", the identifier "ID1" indicating the myocardium 16 and the identifier "ID2" indicating the lumen 17 and the identifier "ID3" indicating the background 18 are included in the item of "object". Is registered.

In the item of "CT value", the general CT value of the target indicated by the identifier registered in the item of "target" is registered. For example, as shown in FIG. 22, in the item of "CT value", the CT value "AA" of the myocardium 16, the CT value "BB" of the lumen 17, and the CT value "CC" of the background 18 are registered. ..

The learning data set generation function 135c according to the second modification uses the database 51 to specify the CT value corresponding to the myocardium 16, the CT value corresponding to the lumen 17, and the CT value corresponding to the background 18. Specifically, for example, the learning data set generation function 135c selects a record in which the identifier "ID1" indicating the myocardium 16 is registered in the "target" item from all the records registered in the database 51. search for.

Then, the learning data set generation function 135c acquires the CT value "AA" registered in the item of the "CT value" of the record obtained as a result of the search. In this way, the learning data set generation function 135c identifies the CT value “AA” corresponding to the myocardium 16. The learning data set generation function 135c specifies the CT value “BB” corresponding to the lumen 17 and the CT value “CC” corresponding to the background 18 by the same method.

According to the method as described above, the CT value corresponding to the myocardium 16, the CT value corresponding to the lumen 17, and the background are not specified by the user without specifying the position of the myocardium 16, the position of the lumen 17, and the position of the background 18. The CT value corresponding to 18 is specified. Therefore, the user can specify the CT value corresponding to the myocardium 16, the CT value corresponding to the lumen 17, and the CT value corresponding to the background 18 without performing a complicated manual operation. In the first embodiment and the first modification, the learning data set generation function 135c may specify the CT value “AA” corresponding to the myocardium 16 by the same method.

(Modification 3 of the first embodiment)
In the first embodiment and the first modification, the image processing device 130 generated the trained models 28 and 28a using the short-axis cross-sectional image data 25. However, the image processing apparatus 130 may generate a trained model using the added average image data of the short-axis cross-section image data 25 of a plurality of time phases instead of the short-axis cross-section image data 25. Therefore, such a modification will be described as a modification 3 of the first embodiment. In the following description of the modified example 3, the points different from the first embodiment and the modified example 1 will be mainly described, and the same points as the first embodiment and the modified example 1 may be omitted. Further, the same reference numerals may be given to the same configurations as those of the first embodiment and the first modification, and the description thereof may be omitted.

FIG. 23 is a diagram for explaining an example of processing executed by the learning data set generation function 135c according to the third modification of the first embodiment. FIG. 23 shows a plurality of time-phase short-axis cross-sectional image data 25. For example, in FIG. 23, the short-axis cross-sectional image data 25 shown on the far left is the image data of the time phase t1. Further, in FIG. 23, the short-axis cross-sectional image data 25 shown in the center is the image data of the time phase t2. Further, in FIG. 23, the short-axis cross-sectional image data 25 shown on the far right is the image data of the time phase t3. The time phase t1, the time phase t2, and the time phase t3 are different time phases.

Then, the learning data set generation function 135c executes the addition averaging process on the short-axis cross-section image data 25 of the plurality of time phases, thereby performing the addition averaging image data of the short-axis cross-section image data 25 of the plurality of time phases. 52 is generated.

Then, the trained model generation function 135d generates a trained model by using the added average image data 52 instead of the short-axis cross-sectional image data 25. For example, the trained model generation function 135d is a trained model using the added average image data 52 in the same manner as the method of generating the trained model 28 using the short-axis cross-sectional image data 25 in the first embodiment. To generate.

Alternatively, the trained model generation function 135d generates a trained model using the added average image data 52 in the same manner as the method of generating the trained model 28a using the short-axis cross-sectional image data 25 in the modified example 1. do. Then, the processing circuit 135 according to the modified example 3 uses the trained model generated in the modified example 3 at the time of operation, and is the same as the processing at the time of operation performed using the trained model 28 in the first embodiment. Is processed.

The modification 3 has been described above. According to the third modification, since the trained model is generated based on the information of a plurality of time phases, it is possible to generate a trained model capable of outputting a more accurate contour.

(Second embodiment)
Next, the image processing apparatus according to the second embodiment will be described. In the following description of the second embodiment, the points different from those of the first embodiment will be mainly described, and the description of the same points as those of the first embodiment may be omitted. Further, the same components as those in the first embodiment may be designated by the same reference numerals and the description thereof may be omitted.

FIG. 24 is a diagram showing an example of the configuration of the medical processing system 1a according to the second embodiment. The medical processing system 1a shown in FIG. 24 is different from the medical processing system 1 shown in FIG. 1 in that the image processing device 130a is provided in place of the image processing device 130. The image processing device 130a shown in FIG. 24 is different from the image processing device 130 shown in FIG. 1 in that the processing circuit 235 is provided in place of the processing circuit 135. The processing circuit 235 is realized by, for example, a processor. The processing circuit 235 includes a training data set generation function 235c, a trained model generation function 235d, and an estimation function 235e in place of the training data set generation function 135c, the trained model generation function 135d, and the estimation function 135e. It is different from the processing circuit 135 shown in FIG.

Hereinafter, among the processes executed by the training data set generation function 235c, the trained model generation function 235d, and the estimation function 235e, the processing executed by the training data set generation function 135c, the trained model generation function 135d, and the estimation function 135e. The differences will be mainly explained. FIG. 25 is a diagram for explaining an example of processing executed by the learning data set generation function 235c according to the second embodiment during operation.

As shown in FIG. 25, the learning data set generation function 235c detects the long axis 14a of the left ventricle 14 depicted in the CT image data 11. Then, the learning data set generation function 235c generates the short-axis cross-section image data 15 showing the cross-section 15a based on the CT image data 11 as in the first embodiment.

Further, the learning data set generation function 235c specifies a cross section 15b separated from the cross section 15a by a predetermined distance in the direction from the cross section 15a toward the heart base of the heart along the long axis 14a. Further, the learning data set generation function 235c specifies a cross section 15c separated from the cross section 15a by a predetermined distance in the direction from the cross section 15a toward the apex of the left ventricle along the long axis 14a. The cross section 15b and the cross section 15c are cross sections orthogonal to the long axis 14a, similarly to the cross section 15a. Further, each of the cross section 15a, the cross section 15b and the cross section 15c is a cross section at a different position.

Then, the learning data set generation function 235c generates the short-axis cross-section image data 55 showing the cross-section 15b based on the CT image data 11. A short-axis cross-section image data 56 showing a cross-section 15c is generated based on the CT image data 11.

Similar to the short-axis cross-sectional image data 15, the short-axis cross-sectional image data 55 and 56 depict the myocardium 16 which is the site to be analyzed, the lumen 17 of the left ventricle, and the background 18. Further, in the short-axis cross-sectional image data 55 and 56, the contour 16a of the outer wall of the myocardium 16 and the contour 16b of the inner wall of the myocardium 16 are drawn as in the short-axis cross-sectional image data 15.

In the second embodiment, the short-axis cross-section image data 15, the short-axis cross-section image data 55, and the short-axis cross-section image data 56 are used as input data when generating the trained model.

The trained model generation function 235d generates a trained model at the time of training. FIG. 26 is a diagram for explaining an example of processing executed by the trained model generation function 235d according to the second embodiment. FIG. 26 shows an example of a method of generating the trained model 58 at the time of learning of the second embodiment. As shown in FIG. 26, at the time of training, the trained model generation function 235d determines the relationship between the set of the short-axis cross-section image data 15, the short-axis cross-section image data 55, and the short-axis cross-section image data 56 and the contour image data 26. The trained model 58 is generated by training.

As described above, the trained model generation function 235d is learned by learning the set of the short-axis cross-section image data 15, the short-axis cross-section image data 55, and the short-axis cross-section image data 56 in association with the contour image data 26. Generate finished model 58. For example, at the time of learning, the learning data set generation function 235c generates a plurality of sets of the short-axis cross-sectional image data 15, the short-axis cross-sectional image data 55, the short-axis cross-sectional image data 56, and the contour image data 26. Then, the trained model generation function 235d generates the trained model 58 using the generated plurality of sets.

For example, the trained model generation function 235d inputs the short-axis cross-sectional image data 15, the short-axis cross-sectional image data 55, and the short-axis cross-sectional image data 56 as input data, and inputs the contour image data 26 as teacher data to the machine engine. , Do machine learning.

The trained model generation function 235d generates a trained model 58 as a result of such machine learning. The trained model 58 is input by inputting a set of image data corresponding to the short-axis cross-sectional image data 15, image data corresponding to the short-axis cross-sectional image data 55, and image data corresponding to the short-axis cross-sectional image data 56. Image data corresponding to the contour image data 26 is estimated (generated) and output.

Then, the trained model generation function 235d stores the generated trained model 58 in the memory 132.

Here, the site to be analyzed (for example, myocardium) is not a two-dimensional structure but a three-dimensional structure. In the second embodiment, the trained model generation function 235d is not the two-dimensional information about the cross section 15a shown by the short axis cross section image data 15, but the analysis target of the three cross sections of the cross section 15a, the cross section 15b, and the cross section 15c. Machine learning is performed using three-dimensional information about the part. Therefore, according to the second embodiment, the learning accuracy and the learning convergence speed can be improved as compared with the case where machine learning is performed using two-dimensional information.

The estimation function 235e uses the trained model 58 to estimate highly accurate image data regarding the contour of the part to be analyzed during operation, and outputs the estimated image data. An example of the process executed by the estimation function 235e will be described with reference to FIG. 27. FIG. 27 is a diagram for explaining an example of processing executed by the estimation function 235e according to the second embodiment. FIG. 27 shows an example of an estimation method of the contour image data 62 at the time of operation of the second embodiment.

The estimation function 235e uses the CT image data 12 acquired by the acquisition function 135b during operation to generate the short-axis cross-sectional image data 30 shown in FIG. 27. For example, the estimation function 235e is the same method as the method in which the learning data set generation function 235c generates the short-axis cross-sectional image data 15 using the CT image data 11 at the time of learning, and the short-axis cross section is performed using the CT image data 12. Image data 30 is generated.

Further, the estimation function 235e uses the CT image data 12 to generate the short-axis cross-sectional image data 60 shown in FIG. 27. For example, the estimation function 235e is the same method as the method in which the learning data set generation function 235c generates the short-axis cross-sectional image data 55 using the CT image data 11 at the time of learning, and the short-axis cross section is performed using the CT image data 12. Image data 60 is generated.

Further, the estimation function 235e uses the CT image data 12 to generate the short-axis cross-sectional image data 61 shown in FIG. 27. For example, the estimation function 235e is the same method as the method in which the learning data set generation function 235c generates the short-axis cross-sectional image data 56 using the CT image data 11 at the time of learning, and the short-axis cross section is performed using the CT image data 12. Image data 61 is generated.

In the short-axis cross-sectional image data 60 and 61, for example, the myocardium, the lumen of the left ventricle, and the background, which are the sites to be analyzed, are depicted in the same manner as in the short-axis cross-sectional image data 15. The short-axis cross-section image data 30 is image data corresponding to the short-axis cross-section image data 15. The short-axis cross-section image data 60 is image data corresponding to the short-axis cross-section image data 55. The short-axis cross-section image data 61 is image data corresponding to the short-axis cross-section image data 56. The short-axis cross-section image data 30, the short-axis cross-section image data 60, and the short-axis cross-section image data 61 are examples of the first image data.

Then, as shown in FIG. 27, the estimation function 235e inputs the short-axis cross-section image data 30, the short-axis cross-section image data 60, and the short-axis cross-section image data 61 into the trained model 58, thereby inputting the trained model 58. The contour image data 62 is estimated and output. That is, the estimation function 235e estimates and outputs the contour image data 62 based on the short-axis cross-section image data 30, the short-axis cross-section image data 60, and the short-axis cross-section image data 61 using the trained model 58. For example, the estimation function 235e outputs the contour image data 62 to the analysis function 135f. The contour image data 62 is image data corresponding to the contour image data 26. The estimation function 235e is an example of an output unit. The contour image data 62 is, for example, an example of information indicating the contour 16a and the contour 16b of the myocardium 16.

The trained model 58 according to the second embodiment can estimate and output contour image data 62 in which the contour of the myocardium, which is the site to be analyzed, is accurately depicted. That is, according to the second embodiment, it is possible to obtain information indicating the contour of the portion to be analyzed with high accuracy.

FIG. 28 is a flowchart showing an example of a flow of processing executed by the image processing apparatus 130a according to the second embodiment during learning. The process shown in FIG. 28 is executed, for example, when the user inputs to the processing circuit 235 an instruction for executing the process of generating the trained model 58 via the input interface 133.

As shown in FIG. 28, the acquisition function 135b acquires the CT image data 11 stored in the memory 132 (step S301). Then, the learning data set generation function 235c generates the short-axis cross-sectional image data 15 based on the CT image data 11 (step S302).

Then, the learning data set generation function 235c generates the short-axis cross-sectional image data 55 based on the CT image data 11 (step S303). Then, the learning data set generation function 235c generates the short-axis cross-sectional image data 56 based on the CT image data 11 (step S304).

Then, the trained model generation function 235d has been trained by machine learning by associating the set of the short-axis cross-sectional image data 15, the short-axis cross-sectional image data 55, and the short-axis cross-sectional image data 56 with the contour image data 26. Model 58 is generated (step S305). Then, the trained model generation function 235d stores the generated trained model 58 in the memory 132 (step S306), and ends the process.

FIG. 29 is a flowchart showing an example of a flow of processing executed by the image processing apparatus 130a according to the second embodiment during operation. The process shown in FIG. 29 is executed, for example, when the user inputs an instruction for executing the analysis by the analysis function 135f to the process circuit 235 via the input interface 133.

As shown in FIG. 29, the acquisition function 135b acquires the CT image data 12 stored in the memory 132 (step S401). Then, the estimation function 235e generates the short-axis cross-sectional image data 30 using the CT image data 12 (step S402).

Then, the estimation function 235e generates the short-axis cross-sectional image data 60 based on the CT image data 12 (step S403). Then, the estimation function 235e generates the short-axis cross-sectional image data 61 based on the CT image data 12 (step S404).

Then, the estimation function 235e estimates and outputs the contour image data 62 based on the short-axis cross-section image data 30, the short-axis cross-section image data 60, and the short-axis cross-section image data 61 using the trained model 58 (step). S405).

Then, the analysis function 135f performs analysis using the contour of the myocardium to be analyzed drawn in the contour image data 62 (step S406). Then, the output function 135g displays the analysis result on the display 134 (step S407), and ends the process.

The second embodiment has been described above. According to the second embodiment, as described above, it is possible to obtain information indicating the contour of the part to be analyzed with high accuracy.

(Modification 1 of the second embodiment)
In the second embodiment, the case where the trained model generation function 235d generates the trained model 58 has been described. However, the trained model generation function 235d may generate another trained model instead of the trained model 58. Therefore, such a modification will be described as a modification 1 of the second embodiment. In the following description of the first embodiment, the points different from the second embodiment will be mainly described, and the description of the same points as the second embodiment may be omitted. Further, the same reference numerals may be given to the same configurations as those of the second embodiment, and the description thereof may be omitted.

FIG. 30 is a diagram for explaining an example of processing executed by the trained model generation function 235d according to the first modification of the second embodiment. FIG. 30 shows an example of a method of generating the trained model 65 at the time of learning the modified example 1. As shown in FIG. 30, at the time of training, the trained model generation function 235d trains the trained model by learning the relationship between the set of the short-axis cross-sectional image data 55 and the short-axis cross-sectional image data 56 and the contour image data 26. Generate 65.

As described above, the trained model generation function 235d generates the trained model 65 by learning the set of the short-axis cross-section image data 55 and the short-axis cross-section image data 56 in association with the contour image data 26. For example, at the time of learning, the learning data set generation function 235c generates a plurality of sets of the short-axis cross-section image data 55, the short-axis cross-section image data 56, and the contour image data 26. Then, the trained model generation function 235d generates the trained model 65 by using the generated plurality of sets.

For example, the trained model generation function 235d performs machine learning by inputting the short-axis cross-sectional image data 55 and the short-axis cross-sectional image data 56 as input data and the contour image data 26 as training data to the machine engine. The trained model generation function 235d generates a trained model 65 as a result of such machine learning.

The trained model 65 estimates (generates) image data corresponding to the contour image data 26 by inputting image data corresponding to the short-axis cross-sectional image data 55 and image data corresponding to the short-axis cross-sectional image data 56. And output.

Then, the trained model generation function 235d stores the generated trained model 65 in the memory 132.

The estimation function 235e according to the first modification uses the trained model 65 to estimate highly accurate image data regarding the contour of the part to be analyzed during operation, and outputs the estimated image data. An example of the process executed by the estimation function 235e will be described with reference to FIG. 31. FIG. 31 is a diagram for explaining an example of processing executed by the estimation function 235e according to the first modification of the second embodiment. FIG. 31 shows an example of a method of estimating the contour image data 66 during the operation of the modified example 1.

The estimation function 235e uses the CT image data 12 acquired by the acquisition function 235b during operation to generate the short-axis cross-sectional image data 60 as in the second embodiment. Further, the estimation function 235e uses the CT image data 12 to generate the short-axis cross-sectional image data 61 as in the second embodiment. The short-axis cross-section image data 60 is image data corresponding to the short-axis cross-section image data 55. The short-axis cross-section image data 61 is image data corresponding to the short-axis cross-section image data 56.

Then, as shown in FIG. 31, the estimation function 235e causes the trained model 65 to estimate the contour image data 66 by inputting the short-axis cross-section image data 60 and the short-axis cross-section image data 61 into the trained model 65. To output. That is, the estimation function 235e estimates and outputs the contour image data 66 based on the short-axis cross-section image data 60 and the short-axis cross-section image data 61 by using the trained model 65. The contour image data 66 is image data corresponding to the contour image data 26.

Similar to the trained model 58 according to the second embodiment, the trained model 65 according to the first modification estimates and outputs contour image data 66 in which the contour of the myocardium, which is the site to be analyzed, is accurately depicted. be able to. That is, according to the first modification, it is possible to obtain information indicating the contour of the portion to be analyzed with high accuracy.

The modification 1 of the second embodiment has been described above. According to the first modification, as described above, it is possible to obtain information indicating the contour of the portion to be analyzed with high accuracy, as in the second embodiment.

Further, in the second embodiment described above, the short-axis cross-section image data 15, the short-axis cross-section image data 55, and the short-axis cross-section image data 56 are used as input data when the trained model 58 is generated. On the other hand, in the modified example 1, the short-axis cross-section image data 55 and the short-axis cross-section image data 56 are used as input data when the trained model 65 is generated. Therefore, the modification example 1 has a smaller number and amount of image data used as input data when generating the trained model than the second embodiment. Therefore, according to the first modification, the convergence of machine learning when generating the trained model can be accelerated.

Further, it is conceivable that the short-axis cross-sectional image data 15 may be deleted from the image processing device 130a for some reason. However, according to the first modification, even in such a case, if the short-axis cross-section image data 55 and the short-axis cross-section image data 56 are stored in the image processing device 130a, the trained model 65 is generated. be able to.

Similarly, it is conceivable that the short-axis cross-sectional image data 30 may be deleted from the image processing device 130a for some reason. However, according to the first modification, even in such a case, if the short-axis cross-section image data 60 and the short-axis cross-section image data 61 are stored in the image processing device 130a, the contour image data 66 is estimated. , Can be output.

(Modification 2 of the second embodiment)
In the second embodiment, the case where the image processing apparatus 130a generates the trained model 58 by using the short-axis cross-section image data 15, the short-axis cross-section image data 55, and the short-axis cross-section image data 56 has been described. However, the image processing apparatus 130a may generate a trained model by using the color image data in which the short-axis cross-sectional image data 15, the short-axis cross-sectional image data 55, and the short-axis cross-sectional image data 56 are combined into one. good. Therefore, such a modification will be described as a modification 2 of the second embodiment. In the following description of the second embodiment, the points different from the second embodiment will be mainly described, and the description of the same points as the second embodiment may be omitted. Further, the same reference numerals may be given to the same configurations as those of the second embodiment, and the description thereof may be omitted.

FIG. 32 is a diagram for explaining an example of processing executed by the image processing apparatus 130a according to the second modification of the second embodiment. FIG. 32 shows an example of a method of generating the trained model 71 at the time of learning the modified example 2. The learning data set generation function 235c according to the second modification generates the color image data 70 shown in FIG. 32. RGB values are set for each of the plurality of pixels constituting the color image data 70.

The plurality of pixels constituting the color image data 70 correspond to the plurality of pixels constituting the short-axis cross-sectional image data 15. Further, the plurality of pixels constituting the color image data 70 also correspond to a plurality of pixels constituting the short-axis cross-sectional image data 55. Further, the plurality of pixels constituting the color image data 70 also correspond to a plurality of pixels constituting the short-axis cross-sectional image data 56.

An example of a method of generating the color image data 70 will be described. For example, the learning data set generation function 235c sets the brightness value set for each pixel of the short-axis cross-sectional image data 15 as the G value for each corresponding pixel of the color image data 70. Further, the learning data set generation function 235c sets the brightness value set for each pixel of the short-axis cross-sectional image data 55 as the R value for each corresponding pixel of the color image data 70. Further, the learning data set generation function 235c sets the brightness value set for each pixel of the short-axis cross-sectional image data 56 as the B value for each corresponding pixel of the color image data 70. In this way, the learning data set generation function 235c generates color image data 70 in which RGB values are set for each pixel.

Here, the G value set for each pixel of the color image data 70 indicates the feature of the cross section 15a. Further, the R value set for each pixel of the color image data 70 indicates the characteristics of the cross section 15b on the core base side of the cross section 15a. Further, the B value set for each pixel of the color image data 70 indicates the characteristic of the cross section 15c on the apex side of the cross section 15a. Therefore, it can be said that the color image data 70 is image data showing the features of the cross section 15a, the features of the cross section 15b, and the features of the cross section 15c. That is, the color image data 70 shows the myocardium 16 of the left ventricle 14 at a plurality of positions (the position of the cross section 15a, the position of the cross section 15b, and the position of the cross section 15c) of the myocardium 16 of the left ventricle 14.

Returning to the description of FIG. 32, at the time of learning, the trained model generation function 235d generates the trained model 71 by learning the relationship between the color image data 70 and the contour image data 26.

In this way, the trained model generation function 235d generates the trained model 71 by learning the color image data 70 and the contour image data 26 in association with each other. For example, at the time of learning, the learning data set generation function 235c generates a plurality of sets of color image data 70 and contour image data 26. Then, the trained model generation function 235d generates the trained model 71 using the generated plurality of sets.

For example, the trained model generation function 235d performs machine learning by inputting the color image data 70 as input data and the contour image data 26 as training data into the machine engine. The trained model generation function 235d generates a trained model 71 as a result of such machine learning.

In the second modification, the trained model generation function 235d performs machine learning using three-dimensional information regarding the parts to be analyzed in the three cross sections of the cross section 15a, the cross section 15b, and the cross section 15c shown by the color image data 70. Therefore, according to the second modification, the learning accuracy and the learning convergence speed can be improved as compared with the case where machine learning is performed using two-dimensional information.

The trained model 71 estimates (generates) and outputs the image data corresponding to the contour image data 26 by inputting the image data corresponding to the color image data 70.

Then, the trained model generation function 235d stores the generated trained model 71 in the memory 132.

The estimation function 235e according to the second modification uses the trained model 71 to estimate highly accurate image data regarding the contour of the part to be analyzed during operation, and outputs the estimated image data. An example of the process executed by the estimation function 235e will be described with reference to FIG. 33. FIG. 33 is a diagram for explaining an example of processing executed by the estimation function 235e according to the second modification of the second embodiment. FIG. 33 shows an example of an estimation method of the contour image data 73 during the operation of the modified example 2.

The estimation function 235e is the same method as the method in which the training data set generation function 235c generated the color image data 70 using the CT image data 11, and the color image data 72 shown in FIG. 33 is used by using the CT image data 12. To generate. The color image data 72 is image data corresponding to the color image data 70. Therefore, it can be said that the color image data 72 is image data showing the features of the cross section 15a, the features of the cross section 15b, and the features of the cross section 15c. That is, the color image data 72 shows the myocardium 16 of the left ventricle 14 at a plurality of positions (the position of the cross section 15a, the position of the cross section 15b, and the position of the cross section 15c) of the myocardium 16 of the left ventricle 14. The color image data 72 is, for example, an example of the first image data and the fifth image data.

Then, as shown in FIG. 33, the estimation function 235e inputs the color image data 72 into the trained model 71 to cause the trained model 71 to estimate and output the contour image data 73. That is, the estimation function 235e estimates and outputs the contour image data 73 based on the color image data 72 by using the trained model 71. The contour image data 73 is image data corresponding to the contour image data 26.

The modification 2 has been described above. The trained model 71 according to the second modification can estimate and output the contour image data 73 in which the contour of the myocardium, which is the site to be analyzed, is accurately depicted. Therefore, according to the second modification, it is possible to obtain information indicating the contour of the portion to be analyzed with high accuracy.

In the second modification, the trained model generation function 235d may generate a trained model using the color image data 70 and the short-axis cross-section image data 25. In this case, the estimation function 235e may output the contour image data by inputting the color image data 72 and the short-axis cross-sectional image data 31 into the trained model. This contour image data is, for example, an example of information indicating the contour 16a and the contour 16b of the myocardium 16.

(Modification 3 of the second embodiment)
In the second embodiment, the image processing device 130a generates the trained model 58 by using the short-axis cross-sectional image data 15, the short-axis cross-sectional image data 55, the short-axis cross-sectional image data 56, and the contour image data 26. Was explained. However, the image processing apparatus 130a may generate a trained model by using the short-axis cross-section image data 25 instead of the short-axis cross-section image data 15. Therefore, such a modification will be described as a modification 3 of the second embodiment. In the following description of the third embodiment, the points different from the second embodiment will be mainly described, and the description of the same points as the second embodiment may be omitted. Further, the same reference numerals may be given to the same configurations as those of the second embodiment, and the description thereof may be omitted.

FIG. 34 is a diagram for explaining an example of processing executed by the image processing apparatus 130a according to the third modification of the second embodiment. FIG. 34 shows an example of a method of generating the trained model 85 at the time of learning the modified example 3.

At the time of training, the trained model generation function 235d has been trained by learning the relationship between the set of the short-axis cross-section image data 25, the short-axis cross-section image data 55, and the short-axis cross-section image data 56, and the contour image data 26. Generate model 85.

As described above, the trained model generation function 235d learns by associating the pair of the short-axis cross-section image data 25, the short-axis cross-section image data 55, and the short-axis cross-section image data 56 with the contour image data 26. Generate the trained model 85. For example, at the time of learning, the learning data set generation function 235c generates a plurality of sets of the short-axis cross-sectional image data 25, the short-axis cross-sectional image data 55, the short-axis cross-sectional image data 56, and the contour image data 26. Then, the trained model generation function 235d generates the trained model 85 by using the generated plurality of sets.

For example, the trained model generation function 235d inputs the short-axis cross-sectional image data 25, the short-axis cross-sectional image data 55, and the short-axis cross-sectional image data 56 as input data, and inputs the contour image data 26 as teacher data to the machine engine. , Do machine learning. The trained model generation function 235d generates a trained model 85 as a result of such machine learning.

The short-axis cross-section image data 25 shows the characteristics of the cross-section 15a. Therefore, in the third modification, as in the second embodiment, machine learning is performed using three-dimensional information regarding the parts to be analyzed in the three cross sections of the cross section 15a, the cross section 15b, and the cross section 15c. Therefore, according to the modified example 3, the learning accuracy and the learning convergence speed can be improved as compared with the case where machine learning is performed using two-dimensional information.

The trained model 85 is a contour image by inputting image data corresponding to the short-axis cross-sectional image data 25, image data corresponding to the short-axis cross-sectional image data 55, and image data corresponding to the short-axis cross-sectional image data 56. Image data corresponding to the data 26 is estimated (generated) and output.

Then, the trained model generation function 235d stores the generated trained model 85 in the memory 132.

The estimation function 235e according to the third modification estimates the image data with high accuracy regarding the contour of the part to be analyzed by using the trained model 85 at the time of operation, and outputs the estimated image data. An example of the process executed by the estimation function 235e will be described with reference to FIG. 35. FIG. 35 is a diagram for explaining an example of processing executed by the estimation function 235e according to the third modification of the second embodiment. FIG. 35 shows an example of a method of estimating the contour image data 86 during the operation of the modified example 3.

As shown in FIG. 35, the estimation function 235e inputs the short-axis cross-section image data 31, the short-axis cross-section image data 60, and the short-axis cross-section image data 61 into the trained model 85, so that the contour image is input to the trained model 85. Data 86 is estimated and output. That is, the estimation function 235e estimates and outputs the contour image data 86 based on the short-axis cross-section image data 31, the short-axis cross-section image data 60, and the short-axis cross-section image data 61 by using the trained model 85. The contour image data 86 is image data corresponding to the contour image data 26. The contour image data 86 is, for example, an example of information indicating the contour 16a and the contour 16b of the myocardium 16. The short-axis cross-sectional image data 60 and the short-axis cross-sectional image data 61 are examples of, for example, a third image data.

The modification 3 has been described above. The trained model 85 according to the third modification can estimate and output the contour image data 86 in which the contour of the myocardium, which is the site to be analyzed, is accurately depicted. Therefore, according to the modified example 3, it is possible to obtain information indicating the contour of the portion to be analyzed with high accuracy.

In the modified example 3, the trained model generation function 235d uses the short-axis cross-sectional image data 55, the short-axis cross-sectional image data 56, at least one of the short-axis cross-sectional image data 56, the short-axis cross-sectional image data 25, and the contour image. A trained model may be generated using the data 26. In this case, the estimation function 235e inputs at least one of the short-axis cross-sectional image data 60 and the short-axis cross-sectional image data 61 and the short-axis cross-sectional image data 31 into the trained model. The contour image data may be output. This contour image data is, for example, an example of information indicating the contour 16a and the contour 16b of the myocardium 16.

(Modification 4 of the second embodiment)
In the third modification of the second embodiment, the image processing apparatus 130a uses the short-axis cross-sectional image data 25, the short-axis cross-sectional image data 55, the short-axis cross-sectional image data 56, and the contour image data 26 to train the model 85. The case of generating is described. However, the image processing device 130a may generate a trained model by using the contour image data instead of the short-axis cross-section image data 55 and the short-axis cross-section image data 56. Therefore, such a modification will be described as a modification 4 of the second embodiment. In the following description of the modified example 4, the points different from the second embodiment will be mainly described, and the description of the same points as the second embodiment may be omitted. Further, the same reference numerals may be given to the same configurations as those of the second embodiment, and the description thereof may be omitted.

FIG. 36 is a diagram for explaining an example of processing executed by the image processing apparatus 130a according to the fourth modification of the second embodiment. FIG. 36 shows an example of a method of generating the trained model 90 at the time of learning the modified example 4.

The learning data set generation function 235c according to the modification 4 generates the contour image data 75 and the contour image data 76 shown in FIG. 36. The contour image data 75 is image data in which the contour 16a and the contour 16b of the myocardium 16 in the cross section 15b are drawn. Further, the contour image data 76 is image data in which the contour 16a and the contour 16b of the myocardium 16 in the cross section 15c are drawn.

An example of a method of generating the contour image data 75 will be described. For example, the learning data set generation function 235c detects the contour 16a and the contour 16b of the myocardium 16 using the short-axis cross-sectional image data 55, and generates the contour image data 75 in which the detected contour 16a and the contour 16b are drawn. do. For example, the learning data set generation function 235c detects the contour 16a and the contour 16b of the myocardium 16 by a known technique using the short-axis cross-sectional image data 55.

For example, the learning data set generation function 235c detects the contour 16b by utilizing the distribution of the luminance values set in the pixels of the short-axis cross-sectional image data 55. To explain with a specific example, when a contrast medium is administered to a subject, the brightness value corresponding to the lumen of the left ventricle of the heart is higher than the brightness value corresponding to the myocardium of the left ventricle due to the influence of the contrast medium. It tends to be higher. Therefore, the learning data set generation function 235c searches for pixels having the same or substantially the same brightness value as the brightness value set for the pixel corresponding to the start point, starting from the center of the myocardium 16 in the short-axis cross-sectional image data 55. do. The center of the myocardium 16 referred to here is, for example, the center of the region surrounded by the myocardium 16 and refers to one point in the lumen 17. Then, the learning data set generation function 235c obtains the region of the lumen 17 as a result of such a search. Then, the learning data set generation function 235c detects the end of the region of the lumen 17 as the contour 16b of the inner wall of the myocardium 16.

Further, the learning data set generation function 235c generates the contour image data 76 using the short-axis cross-sectional image data 56 in the same manner as the method of generating the contour image data 75 using the short-axis cross-sectional image data 55. ..

The contour image data 75 and the contour image data 76 may be generated by the user instead of being generated by the image processing device 130a.

Further, as will be described later, during operation, the contour image data 92 is estimated at a plurality of positions along the long axis 40a by the trained model 90. Therefore, when the contour image data 92 showing the cross section 15b has already been estimated by the trained model 90, the training data set generation function 235c uses the contour image data 92 showing the cross section 15b as the contour image data 75. It may be treated as. Similarly, when the trained model 90 has already estimated the contour image data 92 showing the cross section 15c, the training data set generation function 235c uses the contour image data 92 showing the cross section 15c as the contour image data 76. It may be treated as.

At the time of training, the trained model generation function 235d generates the trained model 90 by learning the relationship between the short-axis cross-section image data 25, the contour image data 75, the contour image data 76, and the contour image data 26. do.

As described above, the trained model generation function 235d learns by associating the pair of the short-axis cross-sectional image data 25, the contour image data 75, and the contour image data 76 with the contour image data 26, and learns the trained model 90. To generate. For example, at the time of learning, the learning data set generation function 235c generates a plurality of sets of short-axis cross-sectional image data 25, contour image data 75, contour image data 76, and contour image data 26. Then, the trained model generation function 235d generates the trained model 90 by using the generated plurality of sets.

For example, the trained model generation function 235d performs machine learning by inputting the short-axis cross-sectional image data 25, the contour image data 75, and the contour image data 76 as input data and inputting the contour image data 26 as teacher data into the machine engine. conduct. The trained model generation function 235d generates a trained model 90 as a result of such machine learning.

Here, the short-axis cross-section image data 25 shows the characteristics of the cross-section 15a. For example, the short-axis cross-section image data 25 shows the characteristics of the contour 16a and the contour 16b in the cross-section 15a. Further, the contour image data 75 shows the characteristics of the contour 16a and the contour 16b in the cross section 15b. Further, the contour image data 76 shows the characteristics of the contour 16a and the contour 16b in the cross section 15c. Therefore, in the modified example 4, machine learning is performed using the three-dimensional information regarding the parts to be analyzed in the three cross sections of the cross section 15a, the cross section 15b, and the cross section 15c, as in the second embodiment. Therefore, according to the modified example 4, the learning accuracy and the learning convergence speed can be improved as compared with the case where machine learning is performed using two-dimensional information.

The trained model 90 corresponds to the contour image data 26 by inputting the image data corresponding to the short-axis cross-sectional image data 25, the image data corresponding to the contour image data 75, and the image data corresponding to the contour image data 76. Estimate (generate) the image data to be output.

Then, the trained model generation function 235d stores the generated trained model 90 in the memory 132.

The estimation function 235e according to the modification 4 estimates the image data with high accuracy regarding the contour of the part to be analyzed by using the trained model 90 at the time of operation, and outputs the estimated image data. An example of the process executed by the estimation function 235e will be described with reference to FIG. 37. FIG. 37 is a diagram for explaining an example of the processing executed by the estimation function 235e according to the modified example 4 of the second embodiment. FIG. 37 shows an example of an estimation method of the contour image data 91 during the operation of the modified example 4.

At the time of learning, the image processing device 130a obtains the contour image data 81 shown in FIG. 37 by the same method as the method of obtaining the contour image data 75. For example, the estimation function 235e is the same method as the method in which the learning data set generation function 235c generates the contour image data 75 using the short-axis cross-sectional image data 55, and the contour image data using the short-axis cross-sectional image data 60. 81 is generated. The contour image data 81 is image data corresponding to the contour image data 75.

Further, the image processing device 130a obtains the contour image data 82 shown in FIG. 37 at the time of learning by the same method as the method of obtaining the contour image data 76. For example, the estimation function 235e is the same method as the method in which the learning data set generation function 235c generates the contour image data 76 using the short-axis cross-sectional image data 56, and the contour image data using the short-axis cross-sectional image data 61. Generate 82. The contour image data 82 is image data corresponding to the contour image data 76.

The contour image data 81 and the contour image data 82 are examples of, for example, the first image data and the fourth image data.

As shown in FIG. 37, the estimation function 235e estimates the contour image data 91 in the trained model 90 by inputting the short-axis cross-sectional image data 31, the contour image data 81, and the contour image data 82 into the trained model 90. Let it output. That is, the estimation function 235e estimates and outputs the contour image data 91 based on the short-axis cross-sectional image data 31, the contour image data 81, and the contour image data 82 using the trained model 90. The contour image data 91 is image data corresponding to the contour image data 26. The contour image data 91 is, for example, an example of information indicating the contour 16a and the contour 16b of the myocardium 16.

The modification 4 has been described above. The trained model 90 according to the modified example 4 can estimate and output the contour image data 91 in which the contour of the myocardium, which is the site to be analyzed, is accurately depicted. Therefore, according to the modified example 4, it is possible to obtain information indicating the contour of the portion to be analyzed with high accuracy.

In the modified example 4, the trained model generation function 235d uses at least one of the contour image data 75 and the contour image data 76, the short-axis cross-sectional image data 25, and the contour image data 26. All you have to do is generate a trained model. In this case, the estimation function 235e outputs the contour image data by inputting at least one contour image data of the contour image data 81 and the contour image data 82 and the short-axis cross-sectional image data 31 into the trained model. Just do it. This contour image data is, for example, an example of information indicating the contour 16a and the contour 16b of the myocardium 16.

Further, in the modified example 4, the trained model generation function 235d may generate a trained model without using the short-axis cross-sectional image data 25. That is, the trained model generation function 235d may generate a trained model by learning the relationship between the set of the contour image data 75 and the contour image data 76 and the contour image data 26. In this case, the estimation function 235e estimates the contour image data without inputting the short-axis cross-section image data 31 into the trained model. That is, the estimation function 235e estimates the contour image data by inputting the contour image data 81 and the contour image data 82 into the trained model. This contour image data is, for example, an example of information indicating the contour 16a and the contour 16b of the myocardium 16.

Further, in each of the above-described embodiments and modifications, the case where the site to be analyzed is the myocardium 16 has been described. However, the site to be analyzed is not limited to this. FIG. 38 is a diagram for explaining another example of the analysis target. For example, as shown in FIG. 38, the site to be analyzed may be a tubular structure 95 such as a blood vessel (for example, an aorta or a coronary artery). When the site to be analyzed is the tubular structure 95, a core wire is used instead of the long axis 14a. Further, instead of the cross section 15b on the core base side, for example, a cross section on the upstream side in the blood flow direction is used. Further, instead of the cross section 15c on the apex side, for example, a cross section on the downstream side in the blood flow direction is used.

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 132.

In addition, in FIG. 1 and FIG. 24, it has been described that a single memory 132 stores each program corresponding to each processing function. However, a plurality of memories 132 may be distributed and arranged, and the processing circuits 135 and 235 may be configured to read corresponding programs from the individual memories 132. Further, instead of storing the program in the memory 132, the program may be directly embedded in the circuit of the processor. In this case, the processor realizes the function by reading and executing the program embedded in the circuit.

According to at least one embodiment or modification described above, it is possible to obtain information showing an accurate contour.

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.

132 Memory 135e, 235e Estimate function

Claims (13)

Information indicating the outline of the detection target by inputting the second image data of the first image data of the target position of the detection target, which is the window condition processed based on the brightness value of the detection target. A storage unit that stores the trained model that outputs
Using the trained model, an output unit that outputs information indicating the outline of the detection target based on the second image data, and an output unit.
An image processing device. The trained model further outputs information indicating the contour by inputting the first image data.
The output unit uses the trained model to output information indicating the contour of the detection target based on the first image data and the second image data.
The image processing apparatus according to claim 1. The trained model further outputs information indicating the contour of the detection target by inputting a third image data at at least one position different from the target position of the detection target.
The output unit uses the trained model to output information indicating the contour of the detection target based on the second image data and the third image data.
The image processing apparatus according to claim 1. The trained model further receives the fourth image data in which the outline of the detection target is drawn at at least one position different from the target position of the detection target, whereby the detection target is input. Outputs information showing the outline of
The output unit uses the trained model to output information indicating the contour of the detection target based on the second image data and the fourth image data.
The image processing apparatus according to claim 1. The trained model further receives information indicating the contour of the detection target by inputting fifth image data indicating the detection target at a plurality of positions including the target position of the detection target. Output and
The output unit uses the trained model to output information indicating the contour of the detection target based on the second image data and the fifth image data.
The image processing apparatus according to claim 1. A storage unit that stores a trained model that outputs information indicating the contour of the detection target by inputting at least one first image data indicating the detection target at a plurality of positions of the detection target. ,
Using the trained model, an output unit that outputs information indicating the outline of the detection target based on the first image data, and an output unit.
An image processing device. The trained model outputs information indicating the contour of the detection target by inputting a plurality of the first image data indicating the detection target at the plurality of positions.
The output unit uses the trained model to output information indicating the contour of the detection target based on the plurality of first image data.
The image processing apparatus according to claim 6. The trained model outputs information indicating the contour of the detection target by inputting a plurality of the first image data in which the contour of the detection target is drawn at the plurality of positions.
The output unit uses the trained model to output information indicating the contour of the detection target based on the plurality of first image data.
The image processing apparatus according to claim 6. The trained model outputs information indicating the contour of the detection target by inputting one first image data indicating the detection target at the plurality of positions.
The output unit uses the trained model to output information indicating the contour of the detection target based on the one first image data.
The image processing apparatus according to claim 6. On the computer
Information indicating the outline of the detection target by inputting the second image data of the first image data of the target position of the detection target, which is the window condition processed based on the brightness value of the detection target. The trained model is acquired from the storage unit in which the trained model is stored, and the trained model is acquired.
Using the acquired trained model, information indicating the outline of the detection target is output based on the second image data.
A program to execute processing. On the computer
A storage unit in which a learned model that outputs information indicating the outline of the detection target is stored by inputting at least one first image data indicating the detection target at a plurality of positions of the detection target. Obtain the trained model from
Using the acquired trained model, information indicating the outline of the detection target is output based on the first image data.
A program to execute processing. The computer
Information indicating the outline of the detection target by inputting the second image data of the first image data of the target position of the detection target, which is the window condition processed based on the brightness value of the detection target. The trained model is acquired from the storage unit in which the trained model is stored, and the trained model is acquired.
Using the acquired trained model, information indicating the outline of the detection target is output based on the second image data.
How to perform the process. The computer
A storage unit in which a learned model that outputs information indicating the outline of the detection target is stored by inputting at least one first image data indicating the detection target at a plurality of positions of the detection target. Obtain the trained model from
Using the acquired trained model, information indicating the outline of the detection target is output based on the first image data.
How to perform the process.

Images