JP2020009186A - Diagnosis support device, diagnosis support method and diagnosis support program - Google Patents

Abstract

To lessen a load of a diagnosis in interpreting images.SOLUTION: A diagnosis support device comprises: an image retrieval unit; an amount-of-characteristic extraction unit; and a correlation determination unit. The image retrieval unit is configured to acquire an examination object corresponding to a designated examination, and retrieve a medical image about an image category associated with the examination object from a plurality of medical images acquired in the examination. The amount-of-characteristic extraction unit is configured to extract an amount of characteristic of a designated portion with respect to the medical image obtained by the retrieval. The correlation determination unit is configured to acquire a diagnosis result to be derived from the medical image of the image category on the basis of the amount of characteristic extracted from the medical image of the image category.SELECTED DRAWING: Figure 2

Description

  Embodiments of the present invention relate to a diagnosis support device, a diagnosis support method, and a diagnosis support program.

  In MR (Magnetic Resonance) inspection, imaging is often performed at a time under a plurality of imaging conditions. For example, in the head MR examination, a plurality of MR images such as a DWI (Diffusion Weight Image), an ADC (Apparent Diffusion Coefficient) map, and a T2-weighted image are captured.

  The radiologist interprets the acquired plurality of MR images while performing a plurality of procedures each corresponding to a predetermined examination purpose. For example, in image reading related to head MR examination, an image reading doctor first diagnoses whether or not a patient has acute cerebral infarction. Then, it is diagnosed whether there is another cerebral infarction or stenosis of blood vessels, and finally, whether there is any structural abnormality (for example, inflammation). In each procedure, for example, the radiologist makes a diagnosis by displaying any one of the acquired plurality of MR images and determining the correlation between the displayed plurality of MR images.

  Generally, almost all of the selection of the MR images to be displayed and the determination of the correlation between the MR images need to be performed by a radiologist. 2. Description of the Related Art In recent years, with the development of medical image diagnostic apparatuses such as an MRI (Magnetic Resonance Imaging) apparatus and a medical image management system (PACS: Picture Archiving and Communication System), the amount of data that can be obtained by an interpreting physician has increased. For this reason, the burden imposed on the interpreting doctor during interpretation is increasing.

International Publication No. WO 2013/065090

  The problem to be solved by the invention is to reduce the burden of diagnosis in interpretation.

  According to the embodiment, the diagnosis support device includes an image search unit, a feature amount extraction unit, and a correlation determination unit. The image search unit acquires an examination purpose corresponding to the designated examination, and searches for a medical image of an image type associated with the examination purpose from the plurality of medical images acquired in the examination. The feature amount extraction unit extracts a feature amount of a part designated from the medical image obtained by the search. The correlation determination unit acquires a diagnosis result derived from the medical image of the image type based on the feature amount extracted from the medical image of the image type.

FIG. 1 is a block diagram illustrating a functional configuration of an in-hospital system provided with the diagnosis support apparatus according to the first embodiment. FIG. 2 is a block diagram illustrating a functional configuration of the diagnosis support device illustrated in FIG. FIG. 3 is a diagram for explaining processing when an image interpretation doctor creates an image interpretation report using the diagnosis support apparatus shown in FIG. 2. FIG. 4 is a diagram showing a relation table stored in the memory shown in FIG. FIG. 5 is a diagram showing a definition correlation table stored in the memory shown in FIG. FIG. 6 is a diagram for explaining the processing of the correlation determination function shown in FIG. FIG. 7 is a diagram illustrating a specific example of the processing of the correlation determination function illustrated in FIG. FIG. 8 is a diagram illustrating a specific example of the processing of the correlation determination function illustrated in FIG. FIG. 9 is a diagram illustrating a specific example of the processing of the correlation determination function illustrated in FIG. FIG. 10 is a diagram illustrating a specific example of the processing of the correlation determination function illustrated in FIG. FIG. 11 is a diagram illustrating a specific example of the processing of the correlation determination function illustrated in FIG. FIG. 12 is a diagram for explaining the processing of the finding creation function shown in FIG. FIG. 13 is a diagram illustrating a specific example of the processing of the finding creation function illustrated in FIG. 2. FIG. 14 is a block diagram illustrating a configuration of a diagnosis support system provided with a diagnosis support device according to the second embodiment. FIG. 15 is a block diagram illustrating a functional configuration of the diagnosis support device illustrated in FIG.

  Hereinafter, embodiments will be described with reference to the drawings.

(1st Embodiment)
FIG. 1 is a block diagram illustrating an example of a functional configuration of an in-hospital system provided with a diagnosis support apparatus 10 according to the first embodiment. The system shown in FIG. 1 includes a diagnosis support apparatus 10, a hospital information system (HIS: Hospital Information System) 20, a radiology information system (RIS: Radiology Information System) 30, and a medical image management system (PACS: Picture Archiving and Communication System). ) 40, a report system 50, a medical image diagnostic apparatus 60, and a communication terminal 70. The diagnosis support apparatus 10, the hospital information system 20, the radiation department information system 30, the medical image management system 40, the report system 50, the medical image diagnosis apparatus 60, and the communication terminal 70 are connected to a hospital network such as a LAN (Local Area Network). Connected so that data communication is possible. The connection to the in-hospital network may be a wired connection or a wireless connection. Also, the line to be connected is not limited to the hospital network as long as security is ensured. For example, it may be connected to a public communication line such as the Internet via a VPN (Virtual Private Network).

  The hospital information system 20 is a system that stores in-hospital information such as medical care information, patient information, and order information, and manages the stored in-hospital information. The medical care information includes, for example, information related to an electronic medical record, such as finding information, disease name information, vital information, examination stage information, and information on treatment details. The patient information includes, for example, a patient ID, a patient name, gender, age, and the like. The order information is, for example, information requested by a clinician or the like for a radiological examination, a specimen test, a physiological test, a prescription, a medication, an interpretation, and the like. When the order information is examination order information for requesting a radiological examination, or interpretation order information for requesting interpretation, the examination order information or the interpretation order information includes, for example, examination date, examination site, examination purpose, and medical image diagnostic apparatus. It includes information related to diagnosis by type, requesting doctor, request department, and the like.

  The hospital information system 20 includes, for example, a server device 21 and a communication terminal 22. The server device 21 and the communication terminal 22 are connected to be able to perform data communication via a hospital network. The server device 21 is an in-hospital information management server that stores medical information, patient information, order information, and the like in the hospital information system 20, and manages the stored medical information, patient information, order information, and the like.

  Although FIG. 1 shows an example in which the server included in the hospital information system 20 is only the server device 21, the present invention is not limited to this. A plurality of server devices 21 may be provided as necessary. For example, the server device 21 may be provided for each piece of information to be managed. Specifically, for example, the hospital information system 20 has an electronic medical record system that manages medical information and patient information, and an order entry system that issues order information to a medical department. May be provided with a server device.

  The communication terminal 22 is, for example, a terminal for a medical staff such as a medical doctor to input an instruction to the server device 21. Specifically, for example, the communication terminal 22 is operated by a clinician and outputs an instruction signal to the server device 21 so as to issue an examination order to the radiation department information system 30. When receiving the instruction signal transmitted from the communication terminal 22, the server device 21 transmits the examination order information and the patient information to the radiation department information system 30. The examination order information and the patient information are transmitted to the radiation department information system 30 in accordance with a preset standard, for example, HL7 (Health Level 7). Further, for example, the communication terminal 22 outputs an instruction signal to the server device 21 to issue an interpretation order to the report system 50. Upon receiving the instruction signal transmitted from the communication terminal 22, the server device 21 transmits the interpretation order information and the patient information to the report system 50. The interpretation order information and the patient information are transmitted to the report system 50 in accordance with a preset standard, for example, HL7.

  The radiation department information system 30 is a system that manages information such as examination reservations in the radiation department, which is one of the medical departments. The radiation department information system 30 includes, for example, a server device 31 and a communication terminal 32. The server device 31 and the communication terminal 32 are connected so as to be able to perform data communication via a hospital network.

  The server device 31 is a department information management server that manages information such as examination reservations in the radiation department. Specifically, for example, when receiving the examination order information and the patient information transmitted from the hospital information system 20, the server device 31 sets the examination order information in the radiation department. The examination order information in the radiology department includes detailed setting information required at the time of examination, for example, detailed information on the medical image diagnostic apparatus 60. The detailed examination order information set in the radiation department includes, for example, the presence or absence of a contrast agent, imaging conditions such as radiation dose when the medical image diagnostic apparatus 60 is an X-ray CT (Computed Tomography) apparatus, and an MRI apparatus. In the case of, imaging conditions and the like in various imaging protocols relating to preparatory imaging and main imaging are included. The detailed examination order information may be preset based on the received examination order information, patient information, past examination history, and the like. The server device 31 transmits detailed examination order information and patient information to the medical image diagnostic apparatus 60.

  Although FIG. 1 shows an example in which the server included in the radiation department information system 30 is only the server device 31, the present invention is not limited to this. A plurality of server devices 31 may be provided as needed.

  The communication terminal 32 is, for example, a terminal for medical staff such as a radiologist to input an instruction to the server device 31 of the radiation department information system 30. Specifically, for example, the communication terminal 32 is operated by a radiologist, and based on the examination order information transmitted from the hospital information system 20 and the patient information, transmits detailed examination order information in the radiation department to the server device 31. Enter

  The medical image diagnostic apparatus 60 is an apparatus that generates medical image data by imaging a subject. The medical image diagnostic apparatus 60 includes, for example, an X-ray diagnostic apparatus, an X-ray CT apparatus, an MRI apparatus, an ultrasonic diagnostic apparatus, a SPECT (Single Photon Emission Computed Tomography) apparatus, a PET (Positron Emission computed Tomography) apparatus, a SPECT apparatus, and an X-ray apparatus. SPECT-CT apparatus in which a line CT apparatus is integrated, PET-CT apparatus in which a PET apparatus and an X-ray CT apparatus are integrated, PET-MRI apparatus in which a PET apparatus and an MRI apparatus are integrated, or these And the like.

  The medical image diagnostic apparatus 60 images a patient based on examination order information and patient information managed by the radiation department information system 30, and generates medical image data. The medical image diagnostic apparatus 60 converts the generated medical image data into an image file conforming to DICOM (Digital Imaging and Communication Medicine) standard. The image file includes medical image data and supplementary information added to the medical image data. The supplementary information is information for managing medical image data, and includes, for example, an examination UID, a series UID, a patient ID, and the like. The test UID is an identifier that can uniquely specify a test. The series UID is, for example, an identifier that can be uniquely specified for a series of images acquired for each imaging region or each imaging condition.

  The medical image management system 40 is a system that stores medical image data and manages the stored medical image data. The medical image management system 40 has a server device 41, for example. The server device 41 stores, for example, an image file created by the medical image diagnostic apparatus 60 and converted according to the DICOM standard, and manages the stored image file. For example, the server device 31 transmits the requested image file among the stored image files to the interpreting doctor in response to a request from the interpreting doctor.

  FIG. 1 shows an example in which the server included in the medical image management system 40 is only the server device 41, but the present invention is not limited to this. A plurality of server devices 41 may be provided as necessary.

  The report system 50 is a system that manages the creation of an interpretation report by an interpretation doctor. The report system 50 has, for example, a server device 51.

  The server device 51 manages various kinds of information for an image interpretation doctor to create an image interpretation report. Specifically, for example, the server device 51 stores a management list that manages whether or not the interpretation report has been created for the interpretation order. Upon receiving the interpretation order information and the patient information transmitted from the hospital information system 20, the server device 51 updates the management list. That is, upon receiving the interpretation order information and the patient information, the server device 51 adds, for example, one uninterpreted interpretation order to the management list.

  Although FIG. 1 shows an example in which the server included in the report system 50 is only the server device 51, the present invention is not limited to this. A plurality of server devices 51 may be provided as necessary.

  The communication terminal 70 is a terminal, for example, a viewer for medical staff such as an interpreting doctor to access systems and devices connected to the LAN. More specifically, for example, the communication terminal 70 is operated by an image interpretation doctor and outputs a request signal for requesting information for creating an image interpretation report to the diagnosis support apparatus 10.

  The diagnosis support apparatus 10 is an apparatus that supports creation of an interpretation report by supporting diagnosis of a patient by an interpretation doctor or the like. FIG. 2 is a block diagram illustrating an example of a functional configuration of the diagnosis support device 10 illustrated in FIG. The diagnosis support device 10 illustrated in FIG. 2 has a processing circuit 11, a memory 12, and a communication interface 13. The processing circuit 11, the memory 12, and the communication interface 13 are communicably connected to each other via, for example, a bus.

  The processing circuit 11 is a processor that functions as a center of the diagnosis support device 10. The processing circuit 11 realizes a function corresponding to the program by executing a program stored in the memory 12 or the like.

  The memory 12 is a storage device such as a read only memory (ROM), a random access memory (RAM), a hard disk drive (HDD), a solid state drive (SSD), and an integrated circuit storage device for storing various information. Further, the memory 12 may be a drive device that reads and writes various information from and to a portable storage medium such as a CD-ROM drive, a DVD drive, and a flash memory. Note that the memory 12 does not necessarily need to be realized by a single storage device. For example, the memory 12 may be realized by a plurality of storage devices. Further, the memory 12 may be in another computer connected to the diagnosis support apparatus 10 via a network.

  The memory 12 stores a diagnosis support program and the like according to the present embodiment. Note that this diagnosis support program may be stored in the memory 12 in advance, for example. Further, for example, the program may be stored in a non-transitory storage medium and distributed, read from the non-transitory storage medium, and installed in the memory 12.

  The communication interface 13 performs data communication with the hospital information system 20, the radiation department information system 30, the medical image management system 40, and the report system 50. The communication interface 13 performs data communication in accordance with, for example, a known standard that is set in advance. Communication with the hospital information system 20 is performed, for example, in accordance with HL7. Communication with the medical image management system 40 is performed, for example, in accordance with DICOM.

  Note that the diagnosis support device 10 may have an input interface. The input interface receives various input operations from a user, converts the received input operations into electric signals, and outputs the electric signals to the processing circuit 11. The input interface is connected to input devices such as a mouse, a keyboard, a trackball, a switch, a button, a joystick, a touchpad, and a touch panel on which an instruction is input by touching an operation surface. Further, the input device connected to the input interface may be an input device provided in another computer connected via a network or the like.

  Further, the diagnosis support device 10 may include a display. The display displays various information according to an instruction from the processing circuit 11. Further, the display may display a GUI (Graphical User Interface) for receiving various operations from the user. As the display, for example, an arbitrary display such as a CRT (Cathode Ray Tube) display, a liquid crystal display, an organic EL display, an LED display, and a plasma display can be appropriately used.

  The processing circuit 11 shown in FIG. 2 executes a diagnosis support program stored in the memory 12 to realize a function corresponding to the program. For example, by executing the diagnosis support program, the processing circuit 11 executes an image search function 111, an area extraction function 112, a feature amount extraction function 113, an image information acquisition function 114, a correlation determination function 115, a finding creation function 116, and a report. It has a creation function 117. In the present embodiment, the image search function 111, the region extraction function 112, the feature amount extraction function 113, the image information acquisition function 114, the correlation determination function 115, the finding creation function 116, and the report creation function 117 are performed by a single processor. A case in which this is realized will be described, but the present invention is not limited to this. For example, a processing circuit is configured by combining a plurality of independent processors, and each processor executes a program to execute an image search function 111, an area extraction function 112, a feature amount extraction function 113, an image information acquisition function 114, a correlation determination function 115, the finding creation function 116, and the report creation function 117 may be realized.

  The image search function 111 is a function of searching an image stored in the medical image management system 40 for an image necessary for interpretation. Specifically, for example, in the image search function 111, the processing circuit 11 acquires the examination purpose of the examination requested to be interpreted from the hospital information system 20 or the report system 50. The processing circuit 11 replaces the acquired inspection purpose with the interpretation purpose, and acquires information on the image type associated with the interpretation purpose.

  The relationship between the interpretation purpose and the image type is associated in, for example, a relationship table stored in the memory 12. In the relation table, the image types determined based on the clinical findings that the interpretation purpose needs to be confirmed are summarized in advance for each interpretation purpose. Note that the relation table can be updated as needed. Updating of the relation table may be performed by input from an operator, or may be automatically performed by a predetermined learning system.

  The processing circuit 11 accesses the medical image management system 40 and searches the image file created by the examination for an image file of the acquired image type. The processing circuit 11 transmits the searched image file to the radiologist via the communication interface 13.

  The region extraction function 112 is a function of extracting a target region to be determined from a medical image. Specifically, for example, in the region extraction function 112, the processing circuit 11 acquires a method for extracting a target region for each image type when at least a part of the medical image is specified by the radiologist. The relationship between each image type and the method for extracting a target area for each image type is associated, for example, with a definition table stored in the memory 12. In the definition table, an extraction method capable of extracting an effective target region for each image type is summarized in advance for each diagnosis result. Note that the definition table can be updated as needed. Updating of the definition table may be performed by input from an operator, or may be automatically performed by a predetermined learning system. The processing circuit 11 extracts a target region from a medical image of each image type based on the acquired extraction method.

  The feature amount extraction function 113 is a function of extracting a feature amount for each image type from the extracted target region. Specifically, for example, in the feature amount extraction function 113, the processing circuit 11 acquires a method for extracting the feature amount of the extracted target region. The relationship between each image type and the method of extracting a feature amount for each image type is associated with, for example, a definition table stored in the memory 12. In the definition table, for each image type, a method capable of extracting valid features is summarized in advance for each determinable diagnosis result. The processing circuit 11 extracts a feature amount of the target area using the acquired method.

  The image information acquisition function 114 is a function of acquiring information about a medical image designated by a radiologist. Specifically, for example, in the image information acquisition function 114, the processing circuit 11 acquires information on a medical image at least partially designated by the radiologist, that is, information on a key image. Note that the processing circuit 11 also acquires information about a medical image including a part identified as a part specified by the radiologist. The image information on the key image is, for example, link information, and includes, for example, an address for accessing a medical image stored in the medical image management system 40.

  The correlation determination function 115 is a function of determining a correlation between medical images of each image type based on the feature amount. Specifically, for example, in the correlation determination function 115, the processing circuit 11 checks the correlation between the medical images of each image type by comparing the extracted feature amount with a reference set in advance for each image type. judge. The criterion for the feature amount for determining the correlation between the medical images of each image type is set based on, for example, clinical findings, and is summarized in a correlation table stored in the memory 12 in advance.

  In the correlation determination function 115, the processing circuit 11 acquires a diagnosis result associated with the determined correlation between medical images. Specifically, for example, in the correlation table, a criterion for a feature amount for determining a correlation between medical images of each image type is associated with a diagnosis result. The processing circuit 11 acquires a diagnosis result associated with the determined correlation between medical images based on the correlation table. Note that the correlation table can be updated as needed. Updating of the correlation table may be performed by input from an operator, or may be automatically performed by a predetermined learning system. When the correlation table is derived from the correlation between medical images of each image type, for example, diagnosis results determined based on clinical findings are compiled in advance for each interpretation purpose.

  The finding creating function 116 is a function of creating findings in the interpretation report with reference to the determination result of the correlation determining function 115. Specifically, for example, in the finding creation function 116, the processing circuit 11 creates finding contents based on the information registered in the correlation table and the like. The processing circuit 11 transmits the content of the created finding to the radiologist via the communication interface 13.

  The report creation function 117 is a function for creating an interpretation report based on the created findings. Specifically, for example, in the report creation function 117, the processing circuit 11 creates an image interpretation report based on the findings, key images referred to when creating the findings, image information on the key images, and the like. The processing circuit 11 transmits the created interpretation report to the interpretation doctor via the communication interface 13.

  Next, the operation of the diagnosis support apparatus 10 configured as described above will be described in detail according to the processing procedure of the processing circuit 11.

  FIG. 3 is a diagram illustrating an example of processing when an image interpretation doctor creates an image interpretation report using the diagnosis support apparatus 10 via the communication terminal 70. In the description of the processing of the diagnosis support apparatus 10 using FIG. 3, it is assumed that the examination to be interpreted is, for example, “head MR examination” and the examination purpose is, for example, “cerebral infarction”. Further, with the examination of the interpretation target being “head MR examination” and the examination purpose being “cerebral infarction”, the memory 12 of the diagnosis support apparatus 10 has the relation table shown in FIG. It is assumed that the definition correlation table shown is stored.

  FIG. 4 is a diagram illustrating an example of a relation table in which interpretation purposes and image types are associated with each other. According to FIG. 4, for the purpose of image interpretation: “cerebral infarction”, image type 1: “DWI”, image type 2: “ADC map”, and image type 3: “T2-weighted image” is associated with the image. Further, the image type 1: “MRA X direction” and the image type 2: “MRA Y direction”, which are determined by clinical findings to require confirmation, are associated with the interpretation purpose: “vascular stenosis”. . Note that the relation table illustrated in FIG. 4 includes two interpretation purposes and image types associated with these interpretation purposes, but the relation table is not limited to this. The relation table may include another interpretation purpose and an image type related to the interpretation purpose. For example, in the relation table, the image type associated with the image interpretation purpose: “bleeding” and the image interpretation purpose: “bleeding”, and the image type associated with the image interpretation purpose: “inflammation” and the image interpretation purpose: “inflammation” Etc. may be included.

  FIG. 5 is a diagram illustrating an example of a definition correlation table used when creating an image interpretation report on cerebral infarction. The definition correlation table illustrated in FIG. 5 includes a definition table in which a method for extracting a target region and a method for extracting a feature amount are summarized, and a method for determining a correlation between medical images of each image type. The reference for the feature amount is described together with a correlation table in which the criteria are summarized for each diagnosis result. According to FIG. 5, the diagnosis result about cerebral infarction, which is expressed by the stage and the disease state, is associated with the criterion for the feature amount, the definition for extracting the region, and the definition for extracting the feature amount. I have. The criterion for the feature amount, the definition for extracting the region, and the definition for extracting the feature amount are set for each image type. In FIG. 5, in order to improve the determination accuracy of the diagnosis result, the diagnosis result of the cerebral infarction is associated with a feature amount reference and a method for region extraction in the CT image. Not necessarily required.

  FIG. 5 shows a definition correlation table for cerebral infarction, but the definition correlation table is not limited to this. The memory 12 may also store a definition correlation table suitable for another inspection purpose. FIG. 5 shows a definition correlation table in which the definition table and the correlation table are integrated, but the definition table and the correlation table may be stored in the memory 12 as separate tables. .

  In FIG. 3, first, the radiologist accesses the report system 50 via the communication terminal 70 and selects a “head MR examination” to be interpreted from the management list for managing the interpretation order (step). S31). A selection signal of “head MR examination” is transmitted from the communication terminal 70 to the diagnosis support apparatus 10 (step S32).

  When receiving the selection signal from the communication terminal 70, the processing circuit 11 of the diagnosis support device 10 executes the image search function 111. In the image search function 111, the processing circuit 11 acquires the inspection purpose for the selected “head MR inspection” from the hospital information system 20 (step S33). Specifically, for example, based on an instruction signal input from a physician or the like via the communication terminal 22 to the server device 21 of the hospital information system 20, the examination purpose for the “head MR examination” is “cerebral infarction”. Is stored. The processing circuit 11 reads out the inspection purpose: “cerebral infarction” from the server device 21. The processing circuit 11 may acquire the inspection purpose from the report system 50.

  When acquiring the examination purpose: “cerebral infarction”, the processing circuit 11 converts the acquired examination purpose: “cerebral infarction” into the interpretation purpose, and compares the converted interpretation purpose with the relation table stored in the memory 12. . The processing circuit 11 associates the image interpretation purpose: “cerebral infarction” in the relation table with the image type 1: “DWI”, the image type 2: “ADC map”, and the image type 3: “T2-weighted image”. Is acquired (step S34). When acquiring the purpose of inspection: “cerebral infarction”, the processing circuit 11 reads out the definition correlation table set for “cerebral infarction” from the memory 12 (step S35).

  The processing circuit 11 obtains a DWI, an ADC map, and a T2-weighted image corresponding to the image type acquired in step S34 from the medical image acquired in the examination: “head MR examination” and stored in the medical image management system 40. Is retrieved (step S36). The processing circuit 11 reads a plurality of DWIs, ADC maps, and T2-weighted images obtained as a result of the search from the medical image management system 40, and transmits them to the communication terminal 70 as medical image data (step S37). When the medical image management system 40 stores a CT image of the head of the same patient in the medical image management system 40, the processing circuit 11 may also read out the CT image and transmit the CT image to the communication terminal 70 as medical image data.

  Upon receiving the medical image data, the communication terminal 70 displays the DWI, the ADC map, and the T2-weighted image based on the received medical image data on a display layout for diagnosing “cerebral infarction” (step S38). The radiologist refers to the DWI, the ADC map, and the T2-weighted image displayed on the communication terminal 70, and searches for a lesion in the image. When a suspected lesion is found in the image, that is, for example, when a suspected lesion is found in one DWI (hereinafter, referred to as a key DWI) among a plurality of DWIs, the radiologist makes a discovery. Annotation is performed on the specified part (step S39). The communication terminal 70 outputs the annotation information about the annotation given by the radiologist to the diagnosis support apparatus 10 (step S310). When the medical image data includes a CT image, the communication terminal 70 may also display the CT image on a display layout for diagnosing “cerebral infarction”.

  Upon receiving the annotation information, the processing circuit 11 of the diagnosis support device 10 executes the area extracting function 112. In the area extracting function 112, the processing circuit 11 extracts a target area to be determined for an annotated image or the like based on the received annotation information (step S311). Specifically, first, the processing circuit 11 acquires a method for region extraction from the definition correlation table acquired in step S35 based on the annotation information. For example, when a part of the key DWI is annotated, the processing circuit 11 acquires “area” from the definition row for “DWI” in the definition correlation table. In the present embodiment, the “region” indicates that a pixel value near the annotated position is extracted. The range of the pixel values to be extracted differs depending on the focus point defined according to the type of the annotation. When the annotation is of an arrow type, the tip of the arrow becomes the focus point. For example, when an arrow-shaped annotation is added to the key DWI, a predetermined range at the tip of the arrow is a range from which a pixel value is extracted. The processing circuit 11 extracts, from the key DWI, pixel values in a range in the vicinity of the annotated position at a predetermined gradation. The processing circuit 11 stores the extracted pixel value in the memory 12.

  As the region extraction method, for example, an existing region extraction method such as bolus detection is used. Note that another area extraction method may be used. Further, in the definition correlation table shown in FIG. 5, a case where “region” is described is shown, but the present invention is not limited to this. In the definition correlation table, for example, an area extraction method may be defined instead of “area”. Further, in the definition correlation table, for example, in addition to the “region”, a hue, a state, and the like may be defined.

  Further, for example, when an annotation is added to a part of the key DWI, the processing circuit 11 uses the ADC map and the T2-weighted image for the same part as the key DWI from the plurality of ADC maps and the T2-weighted image. Are selected as a key ADC map and a key T2 emphasized image, respectively. The processing circuit 11 identifies a portion corresponding to the annotated portion in the key DWI from the key ADC map and the key T2 emphasized image. Identification of a part corresponding to the part annotated with the key DWI is performed using, for example, a physiological position of a patient, coordinates in the annotated DWI, and the like, such as an Anatomical Landmark. The processing circuit 11 extracts, from the key ADC map and the key T2 emphasized image, pixel values in a range in the vicinity of the identified part by a predetermined gradation. The processing circuit 11 stores the extracted pixel value in the memory 12.

  When the pixel value of the target area is extracted, the processing circuit 11 executes the feature amount extraction function 113. In the feature amount extraction function 113, the processing circuit 11 acquires a feature amount based on the extracted pixel value (step S312). Specifically, for example, the processing circuit 11 calculates the intensity of the image detection signal based on the key DWI, the key ADC map, and the pixel value extracted from the key T2 emphasized image. The processing circuit 11 converts the signal strength from, for example, the extracted image gradation. The conversion from the image gradation is performed, for example, by defining a linear function between a preset minimum value and a maximum value, and substituting the extracted gradation for the defined linear function. In the conversion from the image gradation, a function specified by the user, for example, an nth-order function, etc., may be used in addition to the linear function.

  When the signal strength is calculated for the key DWI, the key ADC map, and the key T2-weighted image, the processing circuit 11 obtains a method for extracting a feature amount for the image for which the signal strength is calculated from the definition correlation table obtained in step S35. I do. For example, when the signal strength is calculated for the key DWI, the processing circuit 11 acquires “difference from ADC” from the definition row for “DWI” in the definition correlation table. In the present embodiment, the “difference from the ADC” is, for example, taking the difference between the signal strength calculated for the DWI and the signal strength calculated by inverting the ADC map between the DWI and the ADC map representing the same site. Represents The signal strength difference calculation method uses the existing signal strength difference calculation.

  Subsequently, the processing circuit 11 acquires the “ambient difference” from the definition row for “DWI” in the definition correlation table. In the present embodiment, the “surrounding difference” is calculated based on, for example, pixel values extracted from a vicinity range of a part specified by a radiologist or a vicinity range of a part identified as the part in the same image. Represents the difference between the calculated signal strength and the signal strength calculated based on the pixel values around the vicinity range. The processing circuit 11 calculates the signal strength for the key DWI based on the acquired “difference from ADC” and “surrounding difference”.

  Subsequently, the processing circuit 11 acquires a “threshold” for acquiring a feature value from the definition row for “DWI” in the definition correlation table. For example, in the definition of “DWI”, for example, a threshold for determining the feature amount: “equal signal”, “low signal”, and “high signal” is defined. The processing circuit 11 acquires a feature amount of “equal signal”, “low signal”, or “high signal” corresponding to the calculated signal strength based on the acquired “threshold value”. The processing circuit 11 stores the acquired feature amount in the memory 12.

  Further, for example, when the signal strength is calculated for the key ADC map, the processing circuit 11 acquires “difference from DWI” from the definition row for “ADC” in the definition correlation table. In the present embodiment, the “difference from DWI” is, for example, taking the difference between the signal strength calculated for the ADC map and the signal strength calculated by inverting the DWI between the DWI and the ADC map representing the same site. Represents

  Subsequently, the processing circuit 11 acquires a “temporal difference” from the definition row for “ADC” in the definition correlation table. In the present embodiment, the “temporal difference” is, for example, a difference between the signal intensity calculated for the latest image and the signal intensity calculated for the past image between the latest image and the past image representing the same site. It represents that. The past image is acquired with reference to, for example, an interpretation report created by a past examination as described later. The processing circuit 11 calculates a signal strength for the key ADC map based on the acquired “difference from DWI” and “difference over time”.

  Subsequently, the processing circuit 11 acquires a feature amount: “no change”, “decrease”, or “increase” based on the calculated temporal change of the signal intensity. Note that the processing circuit 11 may acquire new occurrences and disappearances as feature amounts in addition to the feature amounts: “no change”, “decrease”, and “increase”. The processing circuit 11 stores the acquired feature amount in the memory 12.

  Further, for example, when the signal strength is calculated for the key T2-weighted image, the processing circuit 11 acquires the “ambient difference” from the definition row for “T2” in the definition correlation table. The processing circuit 11 calculates the signal strength of the key T2-weighted image based on the acquired “surrounding difference”.

  Subsequently, the processing circuit 11 acquires a “threshold” for acquiring a feature amount from the definition row for “T2” in the definition correlation table. For example, in the definition of “T2”, for example, a threshold value for determining the feature amount: “high signal” is defined. The processing circuit 11 acquires a feature amount: “high signal” for the calculated signal strength based on the acquired “threshold”. The processing circuit 11 stores the acquired feature amount in the memory 12.

  Further, for example, when the signal intensity is calculated for the CT image, the processing circuit 11 acquires a “threshold” for acquiring a feature amount from the definition row for “CT” in the definition correlation table. For example, in the definition of “CT”, for example, a CT value for determining a feature amount: “low absorption” is defined as a threshold. The processing circuit 11 acquires a feature amount: “low absorption” for the calculated signal intensity based on the acquired “threshold”. The processing circuit 11 stores the acquired feature amount in the memory 12.

  The processing circuit 11 executes the correlation determination function 115 when acquiring the feature amount. When the correlation determination function 115 is executed, the processing circuit 11 determines the correlation between the key DWI, the key ADC map, and the key T2-weighted image by comparing the acquired feature amount with the reference set in the definition correlation table. (Step S313). Then, the processing circuit 11 acquires a diagnosis result associated with the correlation between the key DWI, the key ADC map, and the key T2-weighted image (step S314). That is, as shown in FIG. 6, the processing circuit 11 acquires a diagnosis result by performing a correlation determination using the correlation table and the acquired feature amount.

  A specific example will be described with reference to FIGS. For example, as shown in FIG. 7, it is assumed that the key DWI is annotated, and a portion corresponding to the annotation for the key DWI is identified from the key ADC map corresponding to the key DWI. It is assumed that a lesion cannot be confirmed for the key T2 emphasized image. At this time, the processing circuit 11 acquires the feature amount: “high signal” for the key DWI, acquires the feature amount: “degraded” for the key ADC map, and acquires the feature amount: Acquires "no disclaimer". The processing circuit 11 compares the acquired feature amounts with the definition correlation table, and as shown in FIG. 8, the DWI feature amount 1: “high signal”, the ADC feature amount 2: “decrease”, and T2 The characteristic amount of “3: Acquired” is associated with “stage:“ hyperacute stage (1 to 24 hours) ”” and pathological condition: “cellular edema” as a diagnostic result. The processing circuit 11 stores the disease stage: “super-acute stage (1-24 hours)” and the disease state: “cellular edema” in the memory 12 as a diagnosis result.

  For example, as shown in FIG. 9, it is assumed that an annotation is added to the key DWI and a part corresponding to the annotation for the key DWI is identified from a key ADC map corresponding to the key DWI. It is assumed that a lesion cannot be confirmed for the key T2 emphasized image. At this time, the processing circuit 11 acquires the feature amount: “high signal” for the key DWI, acquires the feature amount: “degraded” for the key ADC map, and acquires the feature amount: Acquires "no disclaimer". The processing circuit 11 compares the acquired feature amounts with the definition correlation table, and as shown in FIG. 8, the DWI feature amount 1: “high signal”, the ADC feature amount 2: “decrease”, and T2 The characteristic amount of “3: Acquired” is associated with “stage:“ hyperacute stage (1 to 24 hours) ”” and pathological condition: “cellular edema” as a diagnostic result.

  For example, as shown in FIG. 10, it is assumed that a key DWI is annotated, and a part corresponding to the annotation for the key DWI is identified from the key ADC map and the key T2 emphasized image corresponding to the key DWI. At this time, the processing circuit 11 acquires the feature amount: “high signal” for the key DWI, acquires the feature amount: “degraded” for the key ADC map, and acquires the feature amount: Get "high signal". The processing circuit 11 compares the acquired feature amounts with the definition correlation table, and as shown in FIG. 11, the feature amount 1 for DWI: “high signal”, the feature amount 2 for ADC: “decrease”, and T2 The characteristic amount 3: The stage: “Acute stage (1-7 days)” and the condition: “Cellular edema + angioedema” associated with “High signal” are acquired as the diagnosis results.

  When acquiring the diagnosis result, the processing circuit 11 executes the finding creation function 116. When the finding creating function 116 is executed, the processing circuit 11 obtains the obtained diagnosis result, the annotated physiological position, the protocol name / series description used when obtaining the diagnosis result, as shown in FIG. Then, the content of the finding is created based on the feature amount acquired by this protocol (step S315).

  Specifically, for example, the annotated physiological position is recognized as “left frontal lobe”, and as shown in FIG. 13, as a diagnosis result, the stage: “super-acute stage (1-24 hours)”, And pathological condition: “cellular edema” was acquired, the protocol name / series description used in the judgment was “DWI, ADC, T2, CT”, and the feature amounts acquired by this protocol were “high signal” In the case of “low, presumed, early sign”, the processing circuit 11 determines that “a high signal is present in the DWI in the left frontal lobe, the ADC map is low, the T2 weighted image is presumed, and the early sign is in the CT image. Therefore, cell edema is suspected, and it is considered to be a cerebral infarction in the hyperacute stage. " The processing circuit 11 may create the findings using, for example, a technique such as a structured report. The processing circuit 11 transmits the created finding content to the communication terminal 70 as finding data (step S316).

  Upon receiving the finding data, the communication terminal 70 displays the finding based on the finding data on the display (step S317). The radiologist checks the findings displayed on the communication terminal 70, and corrects the findings as necessary (step S318). For example, the radiologist deletes unnecessary and redundant information in the findings. Accordingly, the content of the finding created by the diagnosis support apparatus 10: “In the left frontal lobe, there is a high signal in the DWI, the ADC map is reduced, the T2 weighted image is considered, and the CT image has an early sign. For example, the phrase "essential edema is suspected as a cerebral infarction in the superacute phase." Is rewritten as "a hyperacute cerebral infarction in the left frontal lobe with a high signal in DWI." Upon completion of the correction of the findings by the radiologist, the communication terminal 70 transmits the corrected findings to the diagnosis support apparatus 10 as the findings data.

  When receiving the corrected finding data, the processing circuit 11 executes the image information acquisition function 114. When the image information obtaining function 114 is executed, the processing circuit 11 obtains, as image information, the key DWI, the key ADC map, and the link information on the key T2 emphasized image referred to when creating the finding content (step S320).

  When acquiring the link information about the key DWI, the key ADC map, and the key T2 emphasized image, the processing circuit 11 executes the report creation function 117. When the report creation function 117 is executed, the processing circuit 11 creates an image interpretation report based on the corrected findings (step S321). Specifically, for example, the processing circuit 11 writes an examination date, an examination purpose, patient information, and the like in a predetermined report format. The processing circuit 11 reflects the corrected findings in a predetermined position of the report format. The processing circuit 11 pastes the key DWI, the key ADC map, and the key T2 emphasized image referred to when creating the finding content. Then, the processing circuit 11 associates the key DWI, the key ADC map, and the key T2 emphasized image with the link information on the key DWI, the key ADC map, and the key T2 emphasized image, respectively. As a result, a hyperlink to the key DWI, the key ADC map, and the key T2 emphasized image is provided. The processing circuit 11 transmits the created interpretation report to the communication terminal 70 as report data (step S322).

  Upon receiving the report data, the communication terminal 70 displays an interpretation report based on the report data on a display (step S323). The interpreting physician checks the interpretation report displayed on the communication terminal 70 and determines whether or not the examination purpose for which the findings should be described remains. If there remains, the radiologist designates the next examination purpose to the diagnosis support apparatus 10 and causes the processing from step S34 to be repeated. If there is no examination purpose for which a finding should be described, the radiologist completes the creation of a radiological report (step S324). The communication terminal 70 transmits the report data to the report system 50 when the interpretation doctor completes the creation of the interpretation report. The processing circuit of the communication terminal 70 may determine whether or not the inspection purpose for which a finding should be described remains. If there is a test remaining, the communication terminal 70 causes the diagnosis support apparatus 10 to specify the next test purpose and to repeat the processing from step S34.

  As described above, in the first embodiment, the diagnosis support apparatus 10 acquires the purpose of the examination when the radiologist designates the examination. The diagnosis support apparatus 10 acquires an image type associated with the inspection purpose. The diagnosis support apparatus 10 displays a medical image of the acquired image type among the medical images acquired by the examination to the interpreting physician. The diagnosis support apparatus 10 extracts a feature amount of a part specified by the radiologist from the displayed medical image. Then, the diagnosis support apparatus 10 acquires a diagnosis result associated with a feature amount extracted for each image type. This allows the radiologist to confirm only the medical images of the image types necessary for the diagnosis, so that it is possible to reduce the time spent searching for the medical images.

  Further, in the first embodiment, when the interpreting physician specifies a part for one medical image of the displayed plurality of medical images, the diagnosis support apparatus 10 is at the same position as the medical image, and the image type is Select a different medical image. The diagnosis support apparatus 10 identifies the same part as the part specified by the radiologist in the selected medical image. Then, the diagnosis support apparatus 10 extracts the feature amounts of the part specified by the radiologist and the identified part. Thus, it is possible to acquire the feature amount of this part and the feature amount of the part of another image type corresponding to this part simply by designating any one part of the displayed medical image. Becomes

  In the first embodiment, the diagnosis support apparatus 10 creates a finding based on a diagnosis result, an image type referred to when acquiring the diagnosis result, and a feature amount extracted from a medical image of the image type. I am trying to do it. Thus, the interpreting doctor can acquire the finding content only by specifying the part of the displayed medical image.

  In the first embodiment, the diagnosis support apparatus 10 reflects the findings corrected by the interpreting doctor in a predetermined format, and pastes the medical image referred to when the findings were created as a key image in the format. An image interpretation report is created by embedding image information on a key image in a format. That is, the process of associating evidence with the diagnosis result is automatically performed. This allows the interpreting doctor to obtain an interpretation report in which comprehensive judgment for the examination purpose is described, simply by specifying the site of the displayed medical image and correcting the findings.

  In the first embodiment, an example has been described in which the relation table, the definition table, and the correlation table are stored in the memory 12 and necessary information is acquired by referring to these tables. However, it is not limited to this. The memory 12 may store, instead of the relation table, a model that has been learned so as to provide an image type that needs to be confirmed when the interpretation purpose is input. Further, the memory 12 stores a model learned so as to extract a feature amount according to the input medical image when a medical image of a predetermined image type is input instead of the definition table. Is also good. Further, the memory 12 may store a model learned so as to output a diagnosis result according to the input characteristic amount when a predetermined characteristic amount is input instead of the correlation table.

  In the first embodiment, an example has been described in which an interpretation report for an MR examination is created. The examination for which the interpretation report is created is not limited to the MR examination. The examination for which an interpretation report is to be created may be, for example, examination by dual energy imaging using an X-ray CT apparatus, photon counting CT imaging, or the like.

  In the first embodiment, the case where the processing circuit 11 obtains a plurality of types of related images by using the image search function 111 with reference to, for example, the relation table illustrated in FIG. 4 is described. However, the data defined in the relation table is not limited to an image. The relation table may define information on past examinations, for example, an interpretation report. When the information regarding the past examination is defined in the relation table, the processing circuit 11 of the diagnosis support apparatus 10 operates as follows, for example.

  The processing circuit 11 compares a predetermined interpretation purpose with a relation table stored in the memory 12 and acquires, for example, a “T2-weighted image” and a “past interpretation report” defined in the relation table. I do. The processing circuit 11 reads out, from the hospital information system 20, the radiology department information system 30, or the like, an interpretation report corresponding to the acquired “past interpretation report” and created by a past examination for the same patient. The read interpretation report includes, for example, a statement that "in the T2-weighted image, old bleeding is present in the left putamen." In addition, a T2-weighted image referred to by the radiologist when inputting this finding is attached to the image interpretation report as a key T2-weighted image (key image). When reading the interpretation report, the processing circuit 11 extracts the key image attached to the interpretation report. When the key image is not attached to the interpretation report, the processing circuit 11 may read the image acquired in the past examination from the medical image management system 40.

  Further, the processing circuit 11 reads from the medical image management system 40 medical image data corresponding to the “T2-weighted image” acquired with reference to the association table. The processing circuit 11 transmits, to the communication terminal 70, medical image data on the key image extracted from the image interpretation report and medical image data read from the medical image management system 40.

  When receiving the medical image data, the communication terminal 70 displays an image based on the received medical image data, for example, a T2-weighted image for the current examination and a T2-weighted image for the past examination on the display layout. In the following description, the T2-weighted image for the current test is referred to as a current T2-weighted image, and the T2-weighted image for a past test is referred to as a past T2-weighted image. The radiologist refers to the current T2-weighted image and the past T2-weighted image displayed on the communication terminal 70 and searches for a lesion in the image. When a site suspected of being a lesion is found in the current T2-weighted image, the radiologist annotates the found site. The communication terminal 70 outputs the annotation information on the annotation given by the radiologist to the diagnosis support apparatus 10.

  Upon receiving the annotation information, the processing circuit 11 of the diagnosis support device 10 executes the area extracting function 112. In the region extraction function 112, the processing circuit 11 extracts a target region to be determined for an annotated image or the like based on the received annotation information. The processing circuit 11 extracts, from the current T2-weighted image, pixel values in a range near the annotated position by a predetermined gradation.

  When an annotation is given to a part of the current T2-weighted image, the processing circuit 11 identifies a part corresponding to the annotated part in the current T2-weighted image from the past T2-weighted image. The processing circuit 11 extracts, from the past T2-weighted image, pixel values in a range in the vicinity of the identified part by a predetermined gradation. The processing circuit 11 stores the extracted pixel value in the memory 12.

  When the pixel value of the target area is extracted, the processing circuit 11 executes the feature amount extraction function 113. In the feature amount extraction function 113, the processing circuit 11 acquires a feature amount based on the extracted pixel value. Specifically, for example, the processing circuit 11 calculates the intensity of the detection signal of the image based on the pixel values extracted from the current T2-weighted image and the past T2-weighted image. The processing circuit 11 converts the signal strength from, for example, the extracted image gradation.

  When the signal strength of the current T2-weighted image and the signal strength of the past T2-weighted image are obtained, the processing circuit 11 compares the obtained signal strengths. Specifically, for example, the difference (gradation difference) between the signal strength calculated for the current T2-weighted image and the signal strength calculated for the past T2-weighted image is calculated. The processing circuit 11 determines whether or not the difference between the calculated signal strengths exceeds a plurality of preset thresholds. Here, the plurality of threshold values are, for example, threshold values for determining a feature amount of a change of the current T2-weighted image with respect to the past T2-weighted image: “increase”, “decrease”, or “no change”.

  When the feature amount of the change of the current T2-weighted image with respect to the past T2-weighted image is acquired, the processing circuit 11 executes the correlation determination function 115. When the correlation determination function 115 is executed, the processing circuit 11 determines the correlation between the current T2-weighted image and the past T2-weighted image based on the acquired feature amount. Then, the processing circuit 11 acquires a diagnosis result associated with the correlation between the current T2-weighted image and the past T2-weighted image.

  When acquiring the diagnosis result, the processing circuit 11 executes the finding creation function 116. When the finding creating function 116 is executed, the processing circuit 11 creates finding contents based on the acquired diagnosis result, the annotated physiological position, the read past interpretation report, and the acquired feature amount. Specifically, for example, when the processing circuit 11 determines that the feature amount of the change of the current T2-weighted image with respect to the past T2-weighted image is “no change”, “the left putamen, old bleeding, and the previous T2-weighted Obtain the content of "No change from image".

  The processing circuit 11 transmits the created finding content to the communication terminal 70 as finding data. Upon receiving the finding data, the communication terminal 70 displays findings on the display based on the finding data. The radiologist checks the findings displayed on the communication terminal 70, and corrects the findings as necessary. For example, the content of the finding made by the diagnosis support apparatus 10: “Old bleeding in the left putamen, no change from the previous T2-weighted image” is “No significant change in the left putamen from the previous old bleeding.” Is rewritten as Upon completion of the correction of the findings by the radiologist, the communication terminal 70 transmits the corrected findings to the diagnosis support apparatus 10 as the findings data. As a result, the finding content considering the temporal difference is created.

(Second embodiment)
In the first embodiment, the case where the diagnosis support apparatus 10 is accommodated in a system in a hospital has been described as an example. In the second embodiment, a case where the diagnosis support apparatus 10a is provided outside a hospital will be described as an example.

  FIG. 14 is a block diagram illustrating a configuration example of a diagnosis support system provided with the diagnosis support device 10a according to the second embodiment. The diagnosis support apparatus 10a shown in FIG. 14 is connected to a hospital information system via the Internet and a network such as a communication network provided by a communication carrier. The diagnosis support device 10a may be realized by one server device provided on the network, or may be realized by a plurality of server devices.

  FIG. 15 is a block diagram illustrating an example of a functional configuration of the diagnosis support device 10a illustrated in FIG. The diagnosis support device 10a illustrated in FIG. 15 includes a processing circuit 11a, a memory 12, and a communication interface 13. The processing circuit 11a, the memory 12, and the communication interface 13 are communicably connected to each other via, for example, a bus.

  The processing circuit 11a is a processor that functions as a center of the diagnosis support device 10a. The processing circuit 11a realizes a function corresponding to the diagnosis support program by executing the diagnosis support program stored in the memory 12 or the like. For example, by executing the diagnosis support program, the processing circuit 11a executes an image search function 111a, an area extraction function 112, a feature amount extraction function 113, an image information acquisition function 114a, a correlation determination function 115, a finding creation function 116, and a report. It has a creation function 117. In the present embodiment, the image search function 111a, the area extraction function 112, the feature amount extraction function 113, the image information acquisition function 114a, the correlation determination function 115, the finding creation function 116, and the report creation function 117 are performed by a single processor. A case in which this is realized will be described, but the present invention is not limited to this. For example, a processing circuit is configured by combining a plurality of independent processors, and each processor executes a program to execute an image search function 111a, an area extraction function 112, a feature amount extraction function 113, an image information acquisition function 114a, and a correlation determination function. 115, the finding creation function 116, and the report creation function 117 may be realized.

  The image search function 111a is a function of searching a plurality of transmitted images for an image necessary for image interpretation. Specifically, for example, in the image search function 111a, the processing circuit 11a accepts an input for the inspection purpose of the inspection requested to be interpreted. The processing circuit 11a replaces the input inspection purpose with the interpretation purpose, and acquires information on the image type associated with the interpretation purpose.

  The relationship between the interpretation purpose and the image type is associated in, for example, a relationship table stored in the memory 12. The processing circuit 11a searches for a medical image of the acquired image type from a plurality of medical images for one examination transmitted from the client via the communication terminal 70 or the communication terminal 80. The processing circuit 11a transmits the searched medical image to the client via the communication interface 13.

  The image information acquisition function 114a is a function for acquiring information about a medical image specified by the client. Specifically, for example, in the image information acquisition function 114a, the processing circuit 11a acquires information of a medical image, at least a part of which is designated by the client, that is, information of a key image. The processing circuit 11a also acquires information about a medical image including a part identified as a part specified by the client. The acquired image information is, for example, link information, and includes, for example, an address for accessing a medical image stored in the memory 12 of the diagnosis support apparatus 10a.

  As described above, in the second embodiment, the diagnosis support device 10a is provided on a so-called cloud. The diagnosis support apparatus 10a receives a medical image for a predetermined examination together with a request for creating an interpretation report from the communication terminals 70 and 80 connected to the network. The diagnosis support device 10a receives an input for a test purpose from a client. The diagnosis support apparatus 10a acquires an image type associated with the inspection purpose. The diagnosis support apparatus 10a displays, to the client, a medical image of the acquired image type among the plurality of received medical images. The diagnosis support apparatus 10a extracts a feature amount of a part specified by the client from the displayed medical image. Then, the diagnosis support apparatus 10a acquires a diagnosis result associated with a feature amount extracted for each image type. Thus, if the client inputs the examination purpose, it is only necessary to confirm the medical images of the image types necessary for the diagnosis, so that the time spent searching for the medical images can be suppressed.

  According to at least one embodiment described above, the diagnosis support apparatus can reduce the burden of diagnosis in interpretation.

  The term “processor” used in the description of the embodiments may be, for example, a CPU (central processing unit), a GPU (Graphics Processing Unit), or an application specific integrated circuit (ASIC), a programmable logic device. (For example, a circuit such as a Simple Programmable Logic Device (SPLD), a Complex Programmable Logic Device (CPLD), and a Field Programmable Gate Array (FPGA)). The processor realizes functions by reading and executing a program stored in the storage circuit. Instead of storing the program in the storage circuit, the program may be directly incorporated in the circuit of the processor. In this case, the processor realizes a function by reading and executing a program incorporated in the circuit. Each processor of the above embodiments is not limited to being configured as a single circuit for each processor, but may be configured as a single processor by combining a plurality of independent circuits to realize its functions. Is also good. Furthermore, a plurality of components in each of the above embodiments may be integrated into one processor to realize the function.

  While some embodiments of the invention have been described, these embodiments have been presented by way of example only, and are not intended to limit the scope of the inventions. These embodiments can be implemented in other various forms, and various omissions, replacements, and changes can be made without departing from the spirit of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are also included in the invention described in the claims and the equivalents thereof.

10, 10a Diagnostic support device 11, 11a Processing circuit 12 Memory 13 Communication interface 20 Hospital information system 21 Server device 22 Communication terminal 30 Radiation department information system 31 Server device 32 Communication terminal 40 Medical Image management system 41 Server device 50 Report system 51 Server device 60 Medical image diagnostic device 70 Communication terminal 80 Communication terminals 111 and 111a Image search function 112 Area extraction function 113 Feature extraction function 114 and 114a ... image information acquisition function 115 ... correlation determination function 116 ... finding creation function 117 ... report creation function

Claims (8)

Acquiring an inspection purpose corresponding to the specified inspection, a medical image for an image type associated with the inspection purpose, an image search unit that searches from a plurality of medical images obtained in the inspection,
A feature amount extraction unit that extracts a feature amount of a part specified with respect to the medical image obtained by the search,
A diagnosis support apparatus comprising: a correlation determination unit configured to acquire a diagnosis result derived from a medical image of the image type based on a feature amount extracted from a medical image of the image type.   The image search unit refers to a relation table in which an interpretation purpose derived from the inspection purpose and an image type requiring confirmation for the interpretation purpose are associated with each other, so that an image associated with the inspection purpose is obtained. The diagnosis support device according to claim 1, wherein the seed is obtained.   When a part is designated for one of the plurality of medical images obtained by the search, the feature amount extraction unit is located at the same position as the first medical image and has a different image type. Selecting a second medical image, identifying, in the selected second medical image, the same site as the site specified for the first medical image, and comparing the specified site with the identified site; 3. The diagnosis support apparatus according to claim 1, wherein a feature amount is extracted.   The correlation determining unit refers to a correlation table in which the reference for the feature amount and the diagnosis result are associated with each other, and based on the feature amount extracted from the medical image of the image type, determines the diagnosis result. The diagnosis support apparatus according to claim 1, wherein the diagnosis support apparatus acquires the information.   5. The apparatus according to claim 1, further comprising: a finding creating unit configured to create a finding based on the diagnosis result, an image type referred to when acquiring the diagnosis result, and a feature amount extracted from a medical image of the image type. 6. The diagnosis support device according to any one of the above.   By reflecting the created findings in a predetermined format, pasting the medical image referred to when creating the findings as a key image in the format, and embedding image information about the key image in the format, The diagnosis support apparatus according to claim 5, further comprising a report creation unit that creates an interpretation report. Obtain the inspection purpose corresponding to the specified inspection,
A medical image for an image type associated with the inspection purpose is searched from a plurality of medical images acquired in the inspection,
Extracting a feature amount of a part specified for the medical image obtained by the search,
Based on the feature amount extracted from the medical image of the image type, obtain a diagnosis result derived from the medical image of the image type,
Diagnosis support method. Acquiring an inspection purpose corresponding to the specified inspection, a process of searching for a medical image of an image type associated with the inspection purpose from a plurality of medical images acquired in the inspection,
A process of extracting a feature amount of a part specified with respect to the medical image obtained by the search;
A diagnosis support program for causing a processor to execute a process of acquiring a diagnosis result derived from a medical image of the image type based on a feature amount extracted from the medical image of the image type.