JP2017029461A - Medical image processor and medical image processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a medical image processor and a medical image processing method capable of creating various images for image reading training while considering the influence of a simulation focus even in an area other than the lungs.SOLUTION: The medical image processor includes a storage part for storing various simulation focuses, a selection part for receiving the selection of observation information, and a simulation medical image creation part for adding a simulation focus in the storage part to a reference image according to the observation information to create a simulation medical image. The simulation medical image creation part performs correction processing of the reference image on the basis of the simulation focus to be added and then adds the simulation focus to the reference image.SELECTED DRAWING: Figure 2

Description

  The present invention relates to a medical image processing apparatus and a medical image processing method for processing a medical image acquired by a medical image capturing apparatus such as an X-ray CT apparatus, an MRI apparatus, and an ultrasonic apparatus, and creates a medical image for interpretation training. Regarding technology.

  A medical image acquired by the medical image photographing apparatus is interpreted by an interpreting doctor and used for image diagnosis. It is desirable that an interpreting doctor interprets medical images of various cases and gains experience so that an image diagnosis can be made.

  In Patent Document 1, by synthesizing various virtual nodules and chest images simulating a lesion, as a teaching material for interpretation training, the physique of the subject, the nature, shape, size, position, number of the simulated lesion, It is disclosed that medical images having different imaging conditions are created.

JP 2008-48880 JP

Super Practical Manual CT / Medical Science Evaluation of coronary plaque tissue properties by diagnostic imaging Coronary Disease Journal 2011; 17: 107-113 Medical image analysis and surgical support computational anatomical approach / Yoshinobu Sato, Department of Radiological Medicine, Graduate School of Medicine, Osaka University

  However, Patent Document 1 is a simple one that mainly assumes a lung nodule, and the effect on the surrounding structure due to the addition of a simulated lesion is not considered. In other words, when adding virtual nodules to the lung, which is mostly in the air region, the influence on the surrounding structure can be ignored, but in other regions, it is necessary to consider deformation and dislocation of the surrounding structure. For example, it was not easy to make unstable plaques with lipid-rich eccentric plaques with positive remodeling of blood vessels and fine calcifications close to actual cases.

  Accordingly, an object of the present invention is to provide a medical image processing apparatus and a medical image processing method capable of creating various interpretation training images in consideration of the influence of a simulated lesion even in a region other than the lung.

  In order to achieve the above object, according to the present invention, a simulated lesion is added to the reference image in accordance with the finding information, and the reference image is corrected based on the added simulated lesion.

  More specifically, a storage unit in which various simulated lesions are stored, a selection unit that accepts selection of finding information, and a simulated lesion in the storage unit is added to a reference image according to the finding information for simulation A simulated medical image creating unit for creating a medical image, wherein the simulated medical image creating unit corrects the reference image based on the added simulated lesion and then adds the simulated lesion to the reference image. A medical image processing apparatus.

  A selection step for accepting selection of finding information; and a simulated medical image creation step for creating a simulated medical image by adding a simulated lesion to a reference image in accordance with the finding information, and adding in the simulated medical image creation step The medical image processing method is characterized in that after the correction process is performed on the reference image based on the simulated lesion, the simulated lesion is added to the reference image.

  According to the present invention, it is possible to provide a medical image processing apparatus and a medical image processing method capable of creating various interpretation training images in consideration of the influence of a simulated lesion even in a region other than the lung.

The figure which shows the hardware constitutions of the medical image processing apparatus of this invention The figure which shows the flow of a process of Example 1 of this invention. The figure which shows the example of the screen structure of the image data list display of Example 1 of this invention The figure which shows the flow of the process of the feature-value data creation of Example 1 of this invention The figure which shows an example of the simulation medical image creation condition database of Example 1 of this invention The figure which shows an example of the case classification data of Example 1 of this invention The figure which shows an example of the feature-value data of Example 1 of this invention The figure explaining the data 503 of the simulation medical image creation condition database of Example 1 of this invention The figure which shows the example of simulation medical image condition input of Example 1 of this invention, and simulation medical image data preparation The figure which shows the flow of a process of simulation medical image data preparation of Example 1 of this invention The figure which shows the example of the screen structure of the subject case list display of Example 1 of this invention The figure which shows the flow of a process of Example 2 of this invention.

  Preferred embodiments of a medical image processing apparatus according to the present invention will be described below with reference to the accompanying drawings. In the following description and the accompanying drawings, the same reference numerals are given to the constituent elements having the same functional configuration, and redundant description will be omitted.

  FIG. 1 is a diagram illustrating a hardware configuration of the medical image processing apparatus 1. The medical image processing apparatus 1 includes a CPU (Central Processing Unit) 2, a main memory 3, a storage device 4, a display memory 5, a display device 6, a controller 7 connected to a mouse 8, a keyboard 9, a camera 10, and a network adapter 11. The system bus 12 is configured to be able to send and receive signals. The medical image processing apparatus 1 is connected to a medical image photographing apparatus 14, a medical image database 15, a problem case database 16, a simulated lesion database 17, and a training result database 18 through a network 13 so as to be able to send and receive signals. Here, “to enable signal transmission / reception” indicates a state in which signals can be transmitted / received to each other or from one to the other, regardless of whether they are electrically or optically wired or wireless.

  The CPU 2 is a device that controls the operation of each component. The CPU 2 loads a program stored in the storage device 4 and data necessary for program execution into the main memory 3 and executes it. The storage device 4 is medical image data photographed by the medical image photographing device 14, simulated medical image data created by the CPU 2 executing a program to be described later in response to input from the mouse 8 or the keyboard 9, and a problem case A device for storing data, specifically a hard disk or the like.

  The storage device 4 may be a device that exchanges data with a portable recording medium such as a flexible disk, an optical (magnetic) disk, a ZIP memory, or a USB memory. The medical image data is acquired from the medical image photographing device 14 or the medical image database 15 via a network 13 such as a LAN (Local Area Network). The storage device 4 stores a program executed by the CPU 2 and data necessary for program execution. The main memory 3 stores programs executed by the CPU 2 and the progress of arithmetic processing.

  The display memory 5 temporarily stores display data to be displayed on the display device 6 such as a liquid crystal display or a CRT (Cathode Ray Tube). The mouse 8 and the keyboard 9 are operation devices for an operator to give an operation instruction to the medical image processing apparatus 1. The mouse 8 may be another pointing device such as a trackpad or a trackball. The controller 7 detects the state of the mouse 8, acquires the position of the mouse pointer on the display device 6, and outputs the acquired position information and the like to the CPU 2. The camera 10 measures the line of sight of the operator. The network adapter 11 is for connecting the medical image processing apparatus 1 to a network 13 such as a LAN, a telephone line, or the Internet.

  The medical image capturing device 14 is a device that acquires medical image data such as a tomographic image of a subject. The medical imaging apparatus 14 is, for example, an MRI apparatus, an X-ray CT apparatus, an ultrasonic diagnostic apparatus, a scintillation camera apparatus, a PET apparatus, a SPECT apparatus, or the like. The medical image database 15 is a database system that stores medical image data captured by the medical image capturing device 14.

  The problem case database 16 includes the observation information input from the mouse 8 and the keyboard 9, and the simulated medical image data and the medical image generated by the CPU 2 executing a program described later in response to the input from the mouse 8 and the keyboard 9. This is a database system that stores medical image data in the database 15 in association with each other. The simulated lesion database 17 is a database system that stores feature data of simulated lesions created by the CPU 2 executing a program to be described later in response to input from the mouse 8 or the keyboard 9. The training result database 18 is a database system that stores data that records inputs from the mouse 8, keyboard 9, and camera 10, programs executed by the CPU 2, and the like.

  When the CPU 2 executes a method to be described later, the medical image data and the simulated medical image are displayed on the display device 6.

  FIG. 2 is a diagram showing the flow of processing in the first embodiment of the present invention. Hereinafter, each step of FIG. 2 will be described in detail.

(Step S201)
When the task case data creation mode is selected by operating the mouse 8 or the keyboard 9, the CPU 2 acquires the image data from the medical image database 15 and the problem case database 16, stores it in the storage device 4, and performs the process in step S202. Proceed.

  When the interpretation training mode is selected by operating the mouse 8 or the keyboard 9, the CPU 2 acquires image data from the problem case database 16, stores it in the storage device 4, and advances the process to step S218.

(Step S202)
The CPU 2 displays a list of image data stored in the storage device 4 as shown in FIG. 3, for example. Reference numeral 303 denotes a list of image information such as shooting conditions. Reference numeral 304 denotes a list of thumbnail images. The display range and display method of the image information list 303 and the thumbnail image list 304 can be changed by the search function 301 or the like. One or more image data are selected from the displayed image data by operating the mouse 8 or the keyboard 9.

(Step S203)
When the image data reading operation is selected by operating the mouse 8 or the keyboard 9, the CPU 2 advances the process to step S204. The image data reading operation is selected, for example, when the reading button 302 in FIG.

  When the finding information input operation is selected by operating the mouse 8 or the keyboard 9, the CPU 2 loads the image data selected in step S202 from the storage device 4 to the display memory 5 and advances the process to step S205. The finding information input operation is selected, for example, when the finding information input button 307 in FIG.

  When the simulated image creation operation is selected by operating the mouse 8 or the keyboard 9, the CPU 2 loads the image data selected in step S202 from the storage device 4 to the display memory 5 and advances the process to step S211. The simulated image creation operation is selected, for example, when the simulated image creation button 306 in FIG.

  When the condition setting operation is selected by the operation of the mouse 8 or the keyboard 9, the CPU 2 advances the process to step S216. The condition setting operation is selected, for example, when the condition setting button 305 in FIG.

(Step S204)
The CPU 2 reads the image data designated by the operation of the mouse 8 or the keyboard 9, stores it in the storage device 4, and returns the process to step S202. At this time, the read image may also be stored in the problem case database 16. The image data designation method includes, for example, a method in which a target drive is designated, a data structure in the designated drive is displayed in a tree, and one or a plurality are selected and read.
(Step S205)
The CPU 2 displays the image data loaded in the display memory 5 on the display device 6. The image data display method may be to display slice images side by side, or to display MPR (Multi Planar Reconstruction) images, volume rendering images, and virtual endoscopic images created by 3D image processing. good.

(Step S206)
The CPU 2 stores in the storage device 4 the coordinate information and the finding information on the image designated by the operation of the mouse 8 or the keyboard 9. The findings information includes basic information and history of the patient as a subject, family history, interview results and examination purpose, examination information obtained from examinations other than image diagnosis, and the like. Information such as the name and size of cases and lesions is stored in the storage device 4 in advance, and the CPU 2 displays the information as candidate findings information on the display device 6, and the operator is displayed on the display device 6. Finding information may be selected from candidates. The candidate for the finding information may be displayed based on the information acquired from the simulated lesion database 17. The finding information may be linked to a position on the medical image.

  The CPU 2 loads the program into the main memory 3, performs the area extraction process and various measurement processes using the coordinate information specified by the operation of the mouse 8 and the keyboard 9, and displays the results on the display device. Display in 6. The operator inputs the finding information based on the information displayed on the display device 6.

(Step S207)
The CPU 2 advances the process to step S208 when the creation of the feature data is selected by the operation of the mouse 8 or the keyboard 9, and advances the process to step S210 when it is selected not to create the feature data.

(Step S208)
CPU 2 performs region extraction processing and feature amount analysis processing based on the coordinate information obtained in step S206, creates feature amount data, and stores it in the storage device 4 in association with the finding information obtained in step S206. . The feature amount data creation processing will be described later with reference to FIG. An example of the configuration of the feature amount data and the simulated lesion database 17 is shown in FIG. 5 and will be described later.

(Step S209)
The CPU 2 associates the finding information stored in the storage device 4 with the feature amount data in step S208 and registers it in the simulated lesion database 17.

(Step S210)
The CPU 2 associates the finding information stored in the storage device 4 with the image data in step S206 and registers it in the problem case database 16.
Note that after the processing in step S210, the processing may be continued by making a transition to steps S201, S203, and S205.

(Step S211)
The CPU 2 displays the image data loaded in the display memory 5 on the display device 6. The image data display method may display slice images side by side, or may display an MPR image, volume rendering image, or virtual endoscopic image created by three-dimensional image processing.

(Step S212)
The CPU 2 uses the information such as the imaging part of the image data loaded in the display memory 5 and the part information determined from the coordinate information specified by the operation of the mouse 8 or the keyboard 9 to determine the conditions for creating the simulated medical image. Display candidates. Case classification data 501 in the simulated lesion database 17 is acquired as candidates for conditions for creating simulated medical images, and displayed as a list in context menus 602, 604, and 606, for example, as shown in FIG. For example, when the operator right-clicks on the image and the imaging region of the image is determined to be the heart and the mouse pointer is located in the coronary artery region, a list of conditions for creating a simulated lesion related to the coronary artery is shown in FIG. ) Is displayed on the context menu 604.

(Step S213)
When a creation condition is selected by the mouse 8 or the keyboard 9 from the candidates for the creation condition of the simulated medical image displayed in step S212, the CPU 2 selects the simulated medical image from the feature amount data 502 of the simulated lesion database 17. The creation condition is acquired and stored in 4 in the storage device.

(Step S214)
The CPU 2 creates simulated medical image data according to the simulated medical image creation conditions acquired in step S213. The simulated medical image is created by adding a simulated lesion to the reference image, and the created simulated medical image is used as an interpretation training image. The flow of processing for creating simulated medical image data will be described later with reference to FIG.

(Step S215)
The CPU 2 registers the simulated medical image data created in step S214 and the simulated medical image creation conditions in the problem case database 16 in association with the finding information.

  Note that after the processing in step S215, the processing may be continued by making a transition to steps S201, S203, S205, and S211.

(Step S216)
The CPU 2 accepts input of feature data of the simulated lesion by operating the mouse 8 and the keyboard 9. The feature amount of the simulated lesion is input using generally known information and model data such as Non-Patent Document 1, Non-Patent Document 2, and Non-Patent Document 3.

(Step S217)
The CPU 2 registers the feature data of the simulated lesion obtained in step S216 in the simulated lesion database 17. FIG. 5 shows an example of the configuration of the simulated lesion database 17. When the feature data of the simulated lesion is registered in the simulated lesion database 17 in advance, Step S216 and Step S217 may not be performed.

  Note that after the processing in step S217, the processing may be continued by making a transition to steps S201, S203, and S216.

(Step S218)
The CPU 2 acquires problem case data from the problem case database 16, and displays a list of problem case data, for example, as shown in FIG. 802 displays a list of image information such as imaging conditions of problem case data. Reference numeral 803 is a list of thumbnail images of problem case data. The display range and display method of the image information list 802 and the thumbnail image list 803 can be changed by a search function 801 or the like. One or a plurality of problem case data is selected from the displayed problem case data by operating the mouse 8 or the keyboard 9. After the interpretation training mode is selected in step S201, the operator inputs the operator information, and the CPU 2 performs the process of discriminating the operator, and the optimum task is based on the information recorded for each operator in step S218. Case data may be selected and displayed.

(Step S219)
The CPU 2 loads the problem case data selected in step S218 into the display memory 5, and displays the loaded problem case data on the display device 6. The image data display method may display slice images side by side, or may display an MPR image, volume rendering image, or virtual endoscopic image created by three-dimensional image processing. The operation for displaying the image data on the display device 6 is, for example, when the training start button 804 in FIG.

(Step S220)
The CPU 2 stores the coordinate information on the image designated by the operation of the mouse 8 and the keyboard 9 and the interpretation result in the storage device 4. Information such as names and sizes of cases and lesions may be stored in the storage device 4 in advance, and the CPU 2 may display the information on the display device 6 so that the operator can select them. Information registered in the simulated lesion database 17 may be used for information such as names and sizes of cases and lesions. The interpretation result may be linked to a position on the medical image.

  The CPU 2 loads the program into the main memory 3, performs the area extraction process and various measurement processes using the position information specified by the operation of the mouse 8 and the keyboard 9, and displays the result on the display device. Display in 6. The operator inputs the interpretation result based on the information displayed on the display device 6. Further, the CPU 2 records the image display conditions during training, the usage status of the tool, the execution status of the program, the data of the operator's line-of-sight position acquired by the camera 10 and the like as the operation record.

(Step S221)
When the interpretation completion operation is performed by the mouse 8 or the keyboard 9, the CPU 2 interprets the interpretation result obtained by the input from the mouse 8 or the keyboard 9 in step S220 and the findings in the problem case data being displayed on the display device 6. The information is comparatively displayed on the display device 6. The comparison display may simply display the results side by side, or may display whether the answer is correct or incorrect. It is also possible to display the results obtained by analyzing the image display conditions during training, the usage status of tools, the execution status of the program, the data of the operator's line of sight, etc. so that they can be compared between operators. good.

(Step S222)
The CPU 2 stores the finding information obtained in step S221 and the comparison information between the interpretation results in the training result database 18. If an operator is selected in step S201, the interpretation result is recorded for each operator. The record of the interpretation result includes information such as having been trained.

  Note that after the processing in step S222, the processing may be continued by making a transition to steps S201, S203, and S218.

  FIG. 4 is a diagram showing a flow of processing for creating simulated lesion data according to the first embodiment of the present invention. Hereinafter, each step of FIG. 4 will be described in detail.

(Step S401)
CPU2 advances the process to step S402 when the extraction result of the lesion area does not exist in storage device 4, and acquires the area information when the extraction result of the lesion area exists and performs the process at step S403. Proceed.

(Step S402)
The CPU 2 acquires the finding information from the storage device 4, determines the extraction processing condition based on the acquired finding information, and extracts the lesion area based on the coordinate information of the finding information. The extraction processing conditions and coordinate information may be input by the operator from the mouse 8 or keyboard 9, or the conditions for each case or lesion are stored in the storage device 4 in advance, and the CPU 2 displays these information on the display device 6. And the operator may select them. The extraction conditions for each case and lesion may be acquired from the simulated lesion database 17. Instead of the CPU 2 executing the extraction process, the operator may input an extraction area from the mouse 8 or the keyboard 9.

(Step S403)
The CPU 2 calculates the statistical data of the shape and density value distribution for the lesion area obtained in step S402. As the shape statistical data, for example, the number of pixels, the area, the perimeter, the circularity, the distance from the center of gravity to each contour point, and the like are calculated. As the statistical data of the density value distribution, for example, an average value, a standard deviation, and the like are calculated.

(Step S404)
The CPU 2 extracts peripheral areas based on the lesion area information obtained up to step S402. For example, when attention is paid to the unstable plaque of the coronary artery as shown in FIG. 6B, the concentration value and shape information of the coronary artery region of the portion where the unstable plaque is not generated are acquired. In the peripheral region extraction process, a region obtained by excluding a lesion region from a region within a contour line obtained by simply extending a contour region contour line outward by a certain amount may be used as the peripheral region. May be input.

(Step S405)
The CPU 2 calculates statistical data of the shape and density value distribution for the peripheral area. As the shape statistical data, for example, the number of pixels, the area, the perimeter, the circularity, the distance from the center of gravity to each contour point, and the like are calculated. As the statistical data of the density value distribution, for example, an average value, a standard deviation, and the like are calculated. If it is determined from the finding information that there are a plurality of necessary peripheral areas, the processes in steps S404 and S405 are repeated as many times as necessary.

(Step S406)
The CPU 2 cuts out image data of the lesion area and the peripheral area obtained up to step S404. For example, the volume data in the lesion area and the surrounding area may be cut out, or the polygon data of the lesion area created may be cut out.

(Step S407)
The CPU 2 stores the image data cut out in step S406 in the storage device 4.

(Step S408)
The CPU 2 stores the finding information and the statistical data of the lesion area and the peripheral area obtained up to step S405 in the storage device 4 in association with the image data registered in step S407. Image capturing conditions such as tube voltage, tube current, slice thickness, FOV, and information such as an image reconstruction filter may be stored together.
(Step S409)
CPU2 is feature value data obtained by acquiring feature value data having the same finding information from the simulated lesion database 17 and performing modeling processing and machine learning processing as disclosed in Non-Patent Document 3. Create average feature data and simulated lesion model data.

(Step S410)
The CPU 2 registers the average feature amount data and the simulated lesion model data created in step S409 in the simulated lesion database 17.

  FIG. 5 is a diagram showing an example of the configuration of the simulated lesion feature data and the simulated lesion database 17 according to the first embodiment of the present invention. As shown in FIG. 5A, the simulated lesion database 17 is managed by being divided into, for example, case classification data 501, feature data 502, and simulated lesion model data 503. Data in which names and sizes of cases and lesions are hierarchized is managed as case classification data 501 as shown in FIG. 5B.

  The feature amount data 502 is managed as a plurality of feature amount data tables as shown in FIG. 5C. Each feature amount data table includes a case classification ID, a processing type, a processing condition, and a use when creating a simulated lesion. ID of simulated lesion model data, statistical data of simulated lesion shape and density value distribution, related information such as medical image shooting conditions and resolution, information such as processing method and feature amount used for reference image correction processing, etc. Is included.

  In addition, the patient information entered in step S206, such as the patient history, family history, interview results and examination purpose, and observation information including examination information obtained from examinations other than image diagnosis are also stored in the feature data table. Also good. The simulated lesion model data 503 is managed as a simulated lesion model cut out and modeled from a plurality of past cases. The simulated lesion model data 503 may be created based on a known feature amount related to a lesion as shown in FIG. 5D.

  The case classification data 501 also has a role of managing a plurality of related lesions associated with one case. For example, in the case of a case having lesions in a plurality of organs, feature amount data 502 and simulated lesion model data 503 of each lesion are individually created, and feature amount data of the plurality of simulated lesions is obtained from the case classification data 501. 502 and simulated lesion model data 503 are collectively managed.

  FIG. 7 is a diagram showing a flow of processing for creating simulated medical image data according to the first embodiment of the present invention. Hereinafter, each step of FIG. 7 will be described in detail.

(Step S701)
The CPU 2 loads the coordinate information of the position or area obtained by operating the mouse 8 or the keyboard 9 into the main memory 3. For example, the position information of the mouse pointer 601 in FIG. 6A, the position information of the mouse pointer 603 in FIG. 6B, and the coordinate information in the ROI 605 in FIG. 6C are acquired.

(Step S702)
The CPU 2 performs region extraction processing based on the position information acquired in step S701 and the simulated medical image creation conditions acquired in step S213. For example, in the example of FIG. 6A, the liver region specified in step S701 is extracted by region growing processing or the like.

(Step S703)
The CPU 2 acquires the type and conditions of the simulated medical image creation process from the simulated medical image creation conditions obtained in step S213, and if density conversion processing is required for the reference image, the reference image is displayed in step S704. If shape conversion processing is necessary, the process proceeds to step S705. If addition of lesion model data is necessary, the process proceeds to step S706.

(Step S704)
The CPU 2 obtains density information from the simulated medical image creation conditions obtained in step S213, and performs density conversion processing on the extraction region obtained in step S702. For example, in FIG. 6 (a), the average CT value is obtained from the feature data of the mild fatty fat of the liver selected in the context menu 602, and the average CT value in the extraction region obtained in step S702 is the feature value. The density value of each pixel in the extraction area is increased or decreased by a certain amount so as to match the average CT value acquired from the data. Processing such as flattening or expansion of the histogram may be performed to convert the density distribution to match the feature amount data.

(Step S705)
The CPU 2 acquires shape and position information from the creation conditions of the simulated medical image obtained in step S213, and performs shape conversion on the extraction region obtained in step S702 by matrix calculation, expansion / contraction processing, etc. . For example, in FIG. 6 (c), information such as the degree of atrophy is acquired from the feature data of moderate atrophy selected in the context menu 606, and the contraction process of the extraction region obtained in step S702 is performed. .

(Step S706)
The CPU 2 obtains related information such as simulated lesion shape information, density information, imaging conditions and image conditions and simulated lesion model data 503 from the simulated medical image creation conditions obtained in step S213, and is loaded into the display memory 5. The simulated lesion model data is deformed so as to match the shooting conditions, image conditions, and the like of the current image. For example, when medium-sized unstable plaque is added to the coronary artery as shown in Fig. 6 (b), the difference between the FOV of the acquired simulated lesion model data and the FOV of the image loaded in the display memory 5 is acquired. The deformation process of the simulated lesion model data is performed so as to reduce the difference between the blood vessel diameter of the simulated lesion model data and the diameter of the blood vessel region extracted in step S702.

(Step S707)
Based on the simulated lesion model information 503 and the related information such as the concentration information of the simulated lesion acquired in step S706, the imaging conditions and image conditions, and the simulated lesion model data 503, the CPU 2 captures the imaging conditions and image conditions of the image loaded in the display memory 5. Adjust the concentration of the simulated lesion model data so that it fits. For example, when a medium-sized unstable plaque is added to the coronary artery as shown in FIG. 6 (b), it is caused by the difference between the imaging conditions of the acquired simulated lesion model data and the imaging conditions of the image loaded in the display memory 5. The concentration adjustment process of the simulated lesion model data is performed so as to eliminate the difference between the concentration characteristics and the difference between the concentration distribution of the acquired simulated lesion model data and the concentration distribution of the blood vessel region extracted in step S702.

(Step S708)
The CPU 2 adds the simulated lesion model data that has been subjected to the simulated lesion correction process such as the deformation process and the density adjustment process in steps S706 and S707 to the position specified in step S701. An object may be added by creating a polygon mesh.

(Step S709)
Based on the simulated medical image creation conditions obtained in step S213, it is determined that density conversion processing of the reference image is necessary as the shape or density value of the region in the reference image is changed or an object is added. If YES, the process proceeds to step S710. If it is determined that the reference image needs to be deformed, the process proceeds to step S711. When it is determined that the reference image correction process is not necessary, the simulated medical image creation process ends.

  To determine the correction processing of the reference image, data on the processing method of the surrounding area in the feature data table in the simulated lesion data acquired as the creation condition of the simulation image, basic patient information, past history, family history, interview results Finding information including examination information obtained from examinations other than image diagnosis, examination purpose, and image diagnosis is used.

  For example, if the data of the processing method of the peripheral region in the feature amount data table indicates the dislocation of the adjacent organ, it is determined that the dislocation of the peripheral region along the shape of the added simulated lesion is necessary, and the processing is performed in step S711. Advances. In addition, for example, when the liver fattening process is performed as shown in FIG. 6 (a), the process of increasing subcutaneous fat based on the patient's body fat percentage, waist circumference, blood test information, etc. It is determined that the fatification process is necessary, and the process proceeds to step S710.

(Step S710)
Information used for correcting the density value of the correction target area is acquired from the simulated medical image creation conditions obtained in step S213, and the density conversion process of the correction target area is performed. Here, the correction target region extraction condition is acquired from the simulated medical image creation condition as necessary, and the correction target region is extracted. For example, for the difference area after brain atrophy as shown in FIG. 6 (c), the density range of the peripheral area acquired from the feature data, the average value, the normal random number using the standard deviation, the uniform Processing such as replacing the density value of each pixel with a random number is performed. For example, when unstable plaque is added to the coronary artery as shown in FIG. 6 (b), the presence of myocardial fat is determined from the feature data obtained in step S213, and the concentration value of the myocardial portion is changed. .

(Step S711)
Information used for the deformation processing of the correction target area is acquired from the simulated medical image creation conditions obtained in step S213, and the correction target area is deformed according to the shape of the simulated lesion model added in step S708. Is called. Deformation processing of adjacent organs and related organs extracted in step S702 is performed. If a plurality of feature amount data is associated with the feature amount data acquired in step S213, steps S702 to S711 are repeated for the number of feature amount data.

  In the first embodiment of the present invention, simulated medical image data close to an actual case can be easily created by performing simulated degeneration processing, simulated lesion correction processing, and reference image correction processing, and the created simulated medical image data is used for training. By using it as data, it is possible to accumulate interpretation experience and accumulate necessary knowledge effectively and efficiently.

  FIG. 9 is a diagram showing a flow of processing of the second embodiment of the present invention. Hereinafter, each step of FIG. 9 will be described in detail. The same steps as those in the first embodiment are denoted by the same step symbols, and the description thereof is omitted.

(Step S901)
When the image data reading operation is selected by operating the mouse 8 or the keyboard 9, the CPU 2 advances the process to step S204.

  When the task case data creation operation is selected by operating the mouse 8 or the keyboard 9, the CPU 2 acquires the image data from the medical image database 15 and the problem case database 16, stores it in the storage device 4, and performs the process in step S202. Proceed.

  When the condition setting operation is selected by operating the mouse 8 or the keyboard 9, the CPU 2 advances the process to step S903.

  When an interpretation training operation is selected by operating the mouse 8 or the keyboard 9, the CPU 2 acquires image data from the problem case database 16, stores it in the storage device 4, and advances the process to step S217.

(Step S202, S204)
Same as Example 1.

(Step S702)
When the finding information input operation is selected by operating the mouse 8 or the keyboard 9, the CPU 2 advances the process to step S205.

  When the simulated image creation operation is selected by operating the mouse 8 or the keyboard 9, the CPU 2 advances the process to step S211.

(Steps S205 to S206, S210 to S214)
Same as Example 1.
(Step S903)
The CPU 2 reads the shape data designated by the operation of the mouse 8 or the keyboard 9 and stores it in the storage device 4. The shape data is not limited to medical image data, but is generated from image data acquired by a device other than a general three-dimensional model creation tool or a medical image imaging device. As a method for specifying shape data, for example, there is a method in which a target drive is specified, a data structure in the specified drive is displayed in a tree, and one or a plurality are selected and read.

(Step S904)
The CPU 2 performs feature amount analysis based on the coordinate information obtained in step S206, and stores the feature amount data obtained by the feature amount analysis in the storage device 4. The feature amount data is the same as that in the first embodiment.

(Step S905)
The CPU 2 registers the feature amount stored in the storage device 4 in the simulated lesion database 17 in step S904.

  Note that after the processing in step S905, the processing may be continued by making a transition to steps S901 and S217.

(Steps S218 to S222)
Same as Example 1.

  In Embodiment 2 of the present invention, simulated medical image data can be created more flexibly by using various information other than medical image data, and it can be used as training data effectively and efficiently. You can gain experience in interpretation and accumulate necessary knowledge. Further, even for cases that are difficult to experience in practice, training can be performed by creating and using simulated medical image data.

  As mentioned above, although embodiment of this invention was described, this invention is not limited to these.

  1 Medical image processing device, 2 CPU, 3 Main memory, 4 Storage device, 5 Display memory, 6 Display device, 7 Controller, 8 Mouse, 9 Keyboard, 10 Camera, 11 Network adapter, 12 System bus, 13 Network, 14 Medical Imaging device, 15 Medical image database, 16 Problem case database, 17 Simulated lesion database, 18 Training result database, 301 Search function, 302 Load button, 303 Image information list, 304 Thumbnail image list, 305 Condition setting button, 306 Simulated image Create button, 307 Finding information input button, 501 Case classification data, 502 Feature data, 503 Simulated lesion model data, 601, 603 Mouse pointer, 602, 604, 606 Context menu, 605 ROI, 801 Search function, 802 Image information list , 803 thumbnail image list

Claims (6)

A storage unit for storing various simulated lesions;
A selection unit that accepts selection of findings information;
A simulated medical image creation unit that creates a simulated medical image by adding a simulated lesion in the storage unit to a reference image according to the findings information, and
The medical image processing apparatus, wherein the simulated medical image creation unit performs a correction process on the reference image based on the added simulated lesion and then adds the simulated lesion to the reference image. The medical image processing apparatus according to claim 1,
The medical image processing apparatus according to claim 1, wherein the correction process is a density conversion process for converting a density of a peripheral region of a simulated lesion to be added. The medical image processing apparatus according to claim 1,
The medical image processing apparatus according to claim 1, wherein the correction process is a deformation process for deforming a peripheral region of a simulated lesion to be added. The medical image processing apparatus according to claim 1,
An operation recording unit that records the operator's line-of-sight position or operation content as an operation record while the operator interprets the simulated medical image;
A medical image processing apparatus, further comprising: a comparison display unit that compares and displays operation records for each operator. The medical image processing apparatus according to claim 4, wherein
The comparison display unit further compares and displays an interpretation result for each operator, and a medical image processing apparatus. A selection step for accepting selection of findings information;
A simulated medical image creating step of creating a simulated medical image by adding a simulated lesion to a reference image according to the finding information,
In the simulated medical image creation step, a correction process is performed on the reference image based on the added simulated lesion, and then the simulated lesion is added to the reference image.

Images