JP2021111283A - Medical information processing apparatus, learning data generation program, and learning data generation method - Google Patents

Abstract

To easily generate learning data.SOLUTION: A medical information processing apparatus includes an identification section, an acquisition section, an extraction section, and an integration section. The identification section identifies an inspection purpose of a first inspection on the basis of a first inspection result of a medical image acquired by the first inspection on a subject. The acquisition section acquires a second inspection result in a second inspection performed on the subject for the inspection purpose from a medical data server on the basis of order information on the first inspection result. The extraction section extracts first and second inspection data that can be used for learning a machine learning model from the first and second inspection results, respectively, and extracts a learning image that can be used for learning from the medical image. The integration section integrates the first and second inspection data and the learning image as learning data that can be used for learning.SELECTED DRAWING: Figure 1

Description

The present embodiment relates to a medical information processing device, a learning data generation program, and a learning data generation method.

Traditionally, radiologists select learning data in addition to the usual interpretation work. At this time, the image interpreting doctor needs to appropriately add relevant information to the selected learning data. In addition, the image interpretation doctor describes the image interpretation result in the image interpretation report, which is a product of the image interpretation. However, the interpretation result is not always correct. In this case, the interpreter needs to check medical documents such as pathological test results and medical records.

It is a heavy burden for the interpreting doctor to confirm such data assignment and diagnosis results one by one while carrying out busy work. That is, creating correct answer data in the learning data has a problem that a great deal of labor is required for the image interpreter.

Japanese Unexamined Patent Publication No. 2017-204041

The problem to be solved by the present invention is to easily generate learning data.

The medical information processing apparatus according to the present embodiment includes a specific unit, an acquisition unit, an extraction unit, and an integration unit. The specific unit specifies the inspection purpose of the first examination based on the first examination result regarding the medical image obtained by the first examination on the subject. The acquisition unit acquires the second test result in the second test performed on the subject for the test purpose from the medical data server based on the order information regarding the first test result. The extraction unit extracts the first test data that can be used for learning the machine learning model from the first test result, extracts the second test data that can be used for the learning from the second test result, and obtains the medical image. A learning image that can be used for the learning is extracted from. The integration unit integrates the first inspection data, the second inspection data, and the learning image as learning data that can be used for the learning.

FIG. 1 is a diagram showing a configuration example of a medical information processing device according to an embodiment together with a hospital information system. FIG. 2 is a diagram showing an example of the relationship between the acquisition target by the acquisition function and the extraction target by the extraction function according to the embodiment. FIG. 3 is a diagram showing an example of learning data generated by integrating various data by the integration function according to the embodiment. FIG. 4 is a diagram showing an example of a processing procedure of the learning data generation process according to the embodiment.

Hereinafter, embodiments of a medical information processing apparatus, a learning data generation program, and a learning data generation method will be described in detail with reference to the drawings. In the following embodiments, the parts with the same reference numerals perform the same operation, and duplicate description will be omitted as appropriate. The medical information processing device according to the present application is not limited to the embodiments shown below.

(Embodiment)
The overall configuration of the medical information processing apparatus according to the embodiment will be described. FIG. 1 is a diagram showing a configuration example of the medical information processing apparatus 1 according to the embodiment together with a hospital information system (HIS: Hospital Information System) 3. The medical information processing device 1 is connected to the hospital information system 3 via the communication network 31. First, the hospital information system 3 relating to the medical information processing device 1 will be described, and then the medical information processing device 1 will be described. The hospital information system 3 includes an electronic medical record system 33, a radiation department information system 35, a report creation device 36, a medical image diagnostic device 37, a medical image management system 39, a clinical examination department information system 41, and other department systems. 43 and. The electronic medical record system 33, the radiation department information system 35, the report creation device 36, the medical image diagnostic device 37, the medical image management system 39, the clinical examination department information system 41, and the other department system 43 are wired. Alternatively, they are connected to each other via a wireless communication network 31.

The electronic medical record system 33 is an information system that manages an electronic medical record for recording medical treatment contents. The electronic medical record system 33 includes an electronic medical record server device and an electronic medical record terminal device. The electronic medical record server device is a computer device that executes processing related to management of the electronic medical record. The electronic medical record server device is composed of one to a plurality of computer devices depending on the scale of the system. The electronic medical record terminal device is used by doctors, nurses, etc. who input and refer to the electronic medical record, and one to a plurality of electronic medical record terminal devices are provided depending on the scale of the system. The electronic medical record server device and the electronic medical record terminal device are connected to the communication network 31.

The electronic medical record terminal device inputs a test order for the subject according to the instruction of the user. When the test order is the imaging of a subject related to diagnostic imaging, the test order includes, for example, information on the test contents such as the test site (imaging target, disease name, test purpose, modality, etc.). Information such as the patient ID, patient name, test ID, and test date of the subject to be tested may also be included. The electronic chart server device responds to the input of the test order and provides information on the test order (hereinafter, order information). Is transmitted to the radiation department information system 35 together with the patient information of the subject in the order information. When the examination order is to carry out a clinical examination on the subject, the electronic chart server device inputs the examination order. Accordingly, the order information is transmitted to the clinical laboratory department information system 41.

The test results for all the tests described in the order information are linked to the electronic medical record for the subject. At this time, the clinician who issued the test request makes a diagnosis result based on the test result, and describes the diagnosis result in the electronic medical record of the subject. Diagnosis results include information such as diagnosis name, stage, and stage severity. The stage and the severity of the disease are defined by a predetermined definition document for each stage and the severity of the disease. The electronic medical record terminal device inputs diagnostic results based on various tests on the subject, for example, according to the instruction of the clinician. The diagnosis result is stored in the electronic medical record server device. In addition, the electronic medical record terminal device inputs the result of the treatment executed based on the diagnosis result (hereinafter referred to as the treatment result) according to the instruction of the clinician. The treatment result is stored in the electronic medical record server device. Treatment results include, for example, the presence or absence of surgery, the area to be removed (for example, total resection, partial preservation), the availability of surgery after laparotomy, the presence or absence of radiation therapy, and a series of treatment events in chronological order. be.

The radiological department information system (RIS) 35 is an information system for managing information in the radiological department in a hospital, and includes a radiological department information server device and a radiological department information terminal device. The radiation department information server device is a computer device that executes processing related to information management in the radiation department. The radiation department information server device is composed of one to a plurality of computer devices depending on the scale of the system. The radiation department information terminal device is used by inspection engineers in the radiation department, and one or more units are provided depending on the scale of the system. The radiation department information server device and the radiation department information terminal device are connected to the communication network 31.

As an example of operation, the radiation department information server device receives order information from the electronic medical record system 33, and acquires information and the like of the medical image diagnostic device 37 from the order information. The radiation department information server device transmits the order information to the medical image diagnostic device 37 specified by the order information. When the order information is transmitted to the medical diagnostic imaging apparatus 37, the radiology department information server apparatus may add information such as a patient ID and an examination date to a DICOM (Digital Imaging and Communications in Medicine) tag or the like. ..

The report creation device 36 is a device for an image interpreter to create an image interpretation report for a medical image. The report creation device 36 may be realized by a PC (Personal Computer) terminal in the medical image management system 39. The report creation device 36 receives the order information and the patient information transmitted from the electronic medical record system 33 via the communication network 31. In response to receiving the order information and the patient information, the report creation device 36 creates a template of the interpretation report in which predetermined information such as the patient information, the examination information, and the request source information to be the target of the interpretation report is written.

The report creation device 36 acquires a medical image necessary for creating an interpretation report from the medical image management system 39. Further, when the report creation device 36 is instructed to create the interpretation report, the report creation device 36 displays a report creation screen for the template of the interpretation report in which the predetermined information is written. On the report creation screen, a "confirm" button indicating that the content of the interpretation report has been confirmed is displayed.

The report creation device 36 creates an interpretation report based on input of finding information and the like on the report creation screen. The content to be input to the interpretation report may be selected by the user from a plurality of items displayed on the report creation screen. The interpretation report describes the medical image used for diagnosis in the examination (hereinafter referred to as a key image) and the findings observed in the key image. An inspection ID and an image ID based on order information are attached to the key image. The key image may be associated with the interpretation report by using a technique such as Harperlink. In addition, marks and annotations may be superimposed on the key image to explain the findings.

The interpreter clicks the "Confirm" button to confirm the content displayed on the report creation screen. At this time, the report creation device 36 stores the created image interpretation report in, for example, the medical image management server device in the medical image management system 39. In addition, when the "confirm" button is clicked, the report creation device 36 transmits the created interpretation report to the medical information processing device 1. In addition, the report creation device 36 transmits the created interpretation report to the electronic medical record system 33 in response to a request from the medical examination doctor.

The medical image diagnostic apparatus 37 is an apparatus for acquiring medical images such as an X-ray diagnostic apparatus, an X-ray CT (Computed Tomography) apparatus, and a magnetic resonance imaging apparatus. A plurality of medical image diagnostic devices 37 may be connected via the communication network 31. Based on the examination order from the radiation department information system 35, the medical diagnostic imaging apparatus 37 executes imaging of the subject in response to an operation in which imaging is started by an engineer or the like. By executing the imaging, the medical diagnostic imaging apparatus 37 collects data indicating the state of the body tissue of the subject. Next, the medical image diagnostic apparatus 37 generates a medical image based on the data. The medical image diagnostic apparatus 37 transmits the generated medical image together with the order information to the medical image management system 39.

The medical image management system (PACS: Picture Archiving and Communication Systems) 39 is an information system that manages medical images such as X-ray CT images and MR images. The medical image management system 39 may manage not only the image data after the reconstruction but also the data (raw data) before the reconstruction. The medical image management system 39 includes a medical image management server device. The medical image management server device is a computer device that executes processing related to medical image management. The medical image management server device is composed of one to a plurality of computer devices depending on the scale of the system. The medical image management server device is connected to the communication network 31. When the medical image management server device receives the medical image and the order information from the medical image diagnosis device 37, the medical image management server device stores the medical image and the order information in association with each other. Further, the medical image management server device stores the interpretation report created by the report creation device 36 and manages it as appropriate.

The clinical laboratory information system (LIS: Laboratory Information System) 41 is an information system that manages information in the clinical laboratory department in a hospital, and is equipped with a clinical laboratory department information server device, a clinical laboratory department information terminal device, and a laboratory device. ing. In the clinical laboratory department information system 41, a sample test and a physiological function test are performed by a clinical laboratory technician based on the order information transmitted from the electronic medical record system 33. The sample test may be outsourced to an external testing institution such as a sanitary laboratory. The clinical laboratory department information server device is a computer device that executes processes related to information management in clinical tests. The clinical laboratory department information server device is composed of one to a plurality of computer devices depending on the scale of the system. The clinical laboratory department information terminal device is used by a laboratory engineer or the like in the clinical laboratory department, and one or more units are provided depending on the scale of the system. The clinical laboratory department information server device and the clinical laboratory department information terminal device are connected to the communication network 31.

The testing device is a device related to the execution of a sample test in which a sample such as blood or urine obtained from a patient is analyzed by a clinical laboratory technician, or a physiological function test for measuring an electroencephalogram or an electrocardiogram of a patient. The sample test includes, for example, a pathological test, a blood / biochemical test, a general test for urine and stool, an immunoserocer test, a microbiological test, a blood transfusion and an organ transplant-related test, and the like. Physiological function tests include, for example, electroencephalogram tests, respiratory function tests, cardiac system tests, fundus photography tests, and the like. A plurality of testing devices are installed according to the type of sample test and the type of physiological function test. In addition, at least a part of the sample test may be carried out by an external testing institution. At this time, the clinical laboratory department information server device receives the test results related to the sample test from an external testing institution and stores them. In addition, the clinical examination department information server device transmits the examination result to the electronic medical record system 33.

The other department system 43 is, for example, a surgery department system, a rehabilitation department system, a dialysis department system, or the like. Similar to the above-mentioned radiation department information system 35 and clinical examination department information system 41, in other department systems 43, surgery implementation reports, rehabilitation implementation reports, etc. are issued according to the order issued by the doctor in the electronic medical record terminal device. Created. The other department system 43 transmits the prepared operation report, rehabilitation report, and the like to the electronic medical record system 33.

The medical information processing device 1 is connected to the hospital information system 3 via the communication interface 11 and the communication network 31. The medical information processing device 1 may be incorporated in the hospital information system 3. Further, the medical information processing device 1 may be used as the report creation device 36. The medical information processing device 1 shown in FIG. 1 includes a communication interface 11, a memory 13 (storage unit), a display 15 (display unit), an input interface 17 (input unit), and a processing circuit 19 (processing unit). Have. Data communication between the communication interface 11, the memory 13, the display 15, the input interface 17, and the processing circuit 19 is performed via a bus (BUS).

The communication interface 11 performs data communication with the hospital information system 3. The standard for communication between the communication interface 11 and the hospital information system 3 may be any standard, and examples thereof include HL7 (Hearth Level 7), DICOM, or both.

The memory 13 is a storage device such as an HDD (Hard disk Drive), an SSD (Solid State Drive), or an integrated circuit storage device that stores various information. The memory 13 stores, for example, data generated by executing the acquisition function 193, the specific function 195, the extraction function 197, the calculation function 199, the integration function 201, the selection function 203, and the like. The generated data and these functions will be described later. The memory 13 stores a plurality of learning data generated by the integrated function 201. The learning data is data that can be used for learning a machine learning model. In addition to the HDD and SSD, the memory 13 is connected to a portable storage medium such as a CD (Compact Disc), a DVD (Digital Versaille Disc), or a flash memory, or a semiconductor memory element such as a RAM (Random Access Memory). It may be a drive device that reads and writes various information. Further, the memory 13 stores the control program according to the present embodiment.

The display 15 displays various information. For example, the display 15 outputs a GUI (Graphical User Interface) or the like for receiving various operations from the user. For example, as the display 15, for example, a liquid crystal display (LCD: Liquid Crystal Display), a CRT (Casode Ray Tube) display, an organic EL display (OELD: Organic Electro Luminescence Display), a plasma display or any other display may be appropriately used. , Can be used. Further, the display 15 may be a desktop type, or may be composed of a tablet terminal or the like capable of wireless communication with the main body of the medical information processing device 1.

The input interface 17 receives various input operations from the user, converts the received input operations into electric signals, and outputs the received input operations to the processing circuit 19. For example, the input interface 17 receives selection instructions, various conditions, and the like from the user. As the input interface 17, for example, a mouse, a keyboard, a trackball, a switch, a button, a joystick, a touch pad, a touch panel display, and the like can be appropriately used.

In the present embodiment, the input interface 17 is not limited to the one provided with physical operation parts such as a mouse, a keyboard, a trackball, a switch, a button, a joystick, a touch pad, and a touch panel display. For example, an example of the input interface 17 includes an electric signal processing circuit that receives an electric signal corresponding to an input operation from an external input device provided separately from the device and outputs the electric signal to the processing circuit 19. .. Further, the input interface 17 may be composed of a tablet terminal or the like capable of wireless communication with the main body of the medical information processing device 1.

The processing circuit 19 controls the operation of the entire medical information processing apparatus 1 in response to the electrical signal of the input operation output from the input interface 17. For example, the processing circuit 19 has a processor such as a CPU (Central Processing Unit), an MPU (Micro Processing Unit), a GPU (Graphics Processing Unit), and a memory such as a ROM (Read Only Memory) or RAM as hardware resources. ..

Further, the processing circuit 19 includes an integrated circuit for a specific application (Application Specific Integrated Circuit: ASIC), a field programmable gate array (Field Programmable Gate Array: FPGA), and another composite programmable logic device (Complex Programmable). ), Simple Programmable Logical Device (SPLD) and the like.

The processing circuit 19 has a system control function 191 and an acquisition function 193, a specific function 195, an extraction function 197, a calculation function 199, an integration function 201, and a selection function 203. In the system control function 191 and the acquisition function 193, the specific function 195, the extraction function 197, the calculation function 199, the integrated function 201, and the selection function 203, each function is stored in the memory 13 in the form of a program that can be executed by a computer. .. The processing circuit 19 executes the system control function 191 and the acquisition function 193, the specific function 195, the extraction function 197, the calculation function 199, the integration function 201, and the selection function 203 by the processor that executes the program expanded in the memory.

That is, the processing circuit 19 corresponds to a processor that realizes a function corresponding to each program by reading a program from the memory 13 and executing the program. In other words, the processing circuit 19 in the state where each program is read has a function corresponding to the read program. It should be noted that each function (191, 193, 195, 197, 199, 201, 203) is not limited to the case where it is realized by a single processing circuit. A processing circuit may be formed by combining a plurality of independent processors, and each function (191, 193, 195, 197, 199, 201, 203) may be realized by executing a program by each processor. The processing circuit 19 that realizes the system control function 191 and the acquisition function 193, the specific function 195, the extraction function 197, the calculation function 199, the integrated function 201, and the selection function 203 is a system control unit, an acquisition unit, a specific unit, an extraction unit, and a calculation. This is an example of a department, an integration department, and a selection department.

The system control function 191 controls each function of the processing circuit 19 based on an input operation received from the user via the input interface 17. Specifically, the system control function 191 reads the control program stored in the memory 13 and expands it on the memory in the processing circuit 19, and controls each part of the medical information processing apparatus 1 according to the expanded control program. ..

The acquisition function 193 acquires the first examination result from the radiation department information system 35, triggered by the completion of the first examination result regarding the medical image acquired by the first examination on the subject. The first examination corresponds to the image examination by the medical image diagnostic apparatus 37. The first inspection result corresponds to the interpretation report. For example, the processing circuit 19 acquires the image interpretation report via the communication interface 11 when the content of the image interpretation report is determined by the image interpretation doctor by the acquisition function 193. The processing circuit 19 acquires order information regarding the interpretation report from the electronic medical record system 33 based on the acquired interpretation report.

The acquisition function 193 acquires the second test result in the second test performed on the subject for the test purpose of the first test from the medical data server based on the order information regarding the first test result. The inspection purpose of the first inspection is specified based on the acquired first inspection result by the specific function 195 described later. The second test is, for example, a sample test or a physiological function test. At this time, the medical data server corresponds to the clinical laboratory department information server device and the electronic medical record server device. Hereinafter, for the sake of specific explanation, the sample test shall be a pathological test and a blood / biochemical test. At this time, the second test result corresponds to each clinical test report of the pathological test, the blood / biochemical test, and the physiological function test. The processing circuit 19 acquires the electronic medical record of the subject from the electronic medical record server device, which is a kind of medical data server, based on the order information by the acquisition function 193.

The specific function 195 identifies the inspection purpose of the first examination based on the first examination result regarding the medical image obtained by the first examination for the subject. Specifically, the processing circuit 19 executes natural language processing based on a predefined dictionary on the acquired order information by the specific function 195. As a result, the processing circuit 19 identifies the inspection purpose from the order information. The processing circuit 19 assigns an order number and a patient ID at the time when the examination order is issued for the specified examination purpose, and also assigns a data ID which is a unique ID peculiar to the information processing apparatus 1 for medical use.

The extraction function 197 extracts the first inspection data that can be used for learning the machine learning model from the first inspection result acquired by the acquisition function 193. The first examination data has information indicating the interpretation result such as the lesion characteristic, the size of the lesion, the suspicion of the disease, the corresponding next action, and the past examination result in the interpretation report. Specifically, the processing circuit 19 executes natural language processing based on a predefined dictionary on the information associated with the hyperlink or the medical image in the interpretation report by the extraction function 197. As a result, the processing circuit 19 extracts the first inspection data from the first inspection result. The extraction target by the extraction function 197 may be determined in advance for each layer in the structured interpretation report. Further, the extraction target may be selected by the user via the input interface 17. Further, the processing circuit 19 may display the contents of the extracted first inspection data on the display 15. At this time, the user selects a desired item to be used as learning data from the contents of the first inspection data via the input interface 17.

The extraction function 197 extracts a learning image that can be used for the learning from the medical image in the first test result acquired by the acquisition function 193. Specifically, the processing circuit 19 extracts the key image related to the interpretation report from the interpretation report by the extraction function 107. The processing circuit 19 may cut out the key image in a preset size. Further, the processing circuit 19 may extract a region designated by the image interpreter or a region set by a predetermined image processing in the key image in a preset size. Further, the processing circuit 19 may extract a key image in which a mark or annotation is written and has a preset size. The extracted key image, that is, the medical image, corresponds to the learning image.

The processing circuit 19 attaches a system inspection ID and a system image ID, which are different from the inspection ID and the image ID attached to the key image, to the learning image by the extraction function 197. The learning image corresponds to a lesion image. The processing circuit 19 temporarily stores the lesion image to which the system examination ID and the system image ID are attached in the memory 13.

The extraction function 197 is performed from the DICOM tag (including the private tag) attached to the medical image that is the source of the key image, for example, the type of modality, the manufacturer, the device name, the device ID, and the protocol information (protocol name, implementation procedure). Imaging parameters by modality (items described in DICOM Part 3 C.8 section such as exposure information, compression rate, tube voltage, tube current), image reconstruction parameters, contrast information, examination site, detector Extract tag information such as the number of columns, format, and model number of. When additional imaging of the subject or imaging by another medical diagnostic imaging apparatus 37 is ordered in the order information, the processing circuit 19 repeats extraction of the first examination data, the learning image, and the tag information. .. At this time, the processing circuit 19 extracts (calculates) the inspection intervals (durations) in the plurality of first inspections.

The extraction function 197 extracts the second inspection data that can be used for the learning from the second inspection result acquired by the acquisition function 193. Specifically, the processing circuit 19 is related to the first examination from the clinical examination department information system 41 and the other department system 43 based on the patient information, the order number in the order information, and the examination purpose by the extraction function 197. The second inspection result in the second inspection is extracted. More specifically, the processing circuit 19 obtains the result of the second test, in which the patient name and order number in the first test, the purpose of the test, and the record of the electronic medical record are the same, from the clinical laboratory department information system 41 and the other department system 43. Extract. When an additional sample test or the like for the subject is ordered in the order information, the processing circuit 19 calculates the test interval (period) in the plurality of second tests. Hereinafter, in order to make the explanation concrete, when the second test is a sample test, the second test result corresponds to a sample test report. At this time, the second test data corresponds to the sample test result.

When the sample test is a pathological test, the second test corresponds to the pathological test and the second test result corresponds to the histological examination report. In the pathological examination of cytology, the type of cancer, the malignancy of the cancer, the ability of cancer cells to proliferate, the presence or absence of hormone receptors, the overexpression of HER2 protein, the presence or absence of HER2 gene amplification, etc. contribute to the diagnosis. In the pathological examination of biopsy, the state and shape of distribution, classification of type, presence / absence of malignant cells at the excision boundary, size, presence / absence of infiltration, etc. contribute to diagnosis. The results of the pathological examination will be described in the histology report for the above items that contribute to the diagnosis. The processing circuit 19 executes natural language processing based on a predefined dictionary on the tissue diagnosis report by the extraction function 197. As a result, the processing circuit 19 extracts the pathological test results for the items contributing to the diagnosis from the histological diagnosis report as the second test data.

When the sample test is a blood / biochemical test, the second test corresponds to the blood / biochemical test, and the second test result corresponds to the blood / biochemical test report. The processing circuit 19 extracts numerical values in various items in the blood / biochemical test, that is, blood / biochemical test results as second test data from the blood / biochemical test report by the extraction function 197.

The extraction function 197 extracts the diagnostic result of the subject based on the first test result and the second test result from the electronic medical record acquired by the acquisition function 193. Specifically, the processing circuit 19 executes natural language processing based on a predefined dictionary on the electronic medical record by the extraction function 197. As a result, the processing circuit 19 extracts the diagnosis result from the electronic medical record. The source of the diagnosis result may be a part of the electronic medical record related to the order information.

The extraction function 197 extracts the treatment result corresponding to the diagnosis result from the electronic medical record acquired by the acquisition function 193. Specifically, the processing circuit 19 executes natural language processing based on a predefined dictionary on the electronic medical record by the extraction function 197. As a result, the processing circuit 19 extracts the treatment result from the electronic medical record. The source of the treatment result may be a part of the electronic medical record related to the diagnosis result.

The extraction function 197 extracts patient information regarding the subject from the electronic medical record acquired by the acquisition function 193. Patient information includes, for example, height / weight / BMI, blood pressure, body fat, gender, age, medical history, race (race), genomic information (for example, genetic diagnosis results), occupation, diet / drinking history, smoking history. , Exercise habits, family history, for women: birth history, first menstrual age, menopausal age, menstrual status, etc. Specifically, the processing circuit 19 executes natural language processing based on a predefined dictionary on the electronic medical record by the extraction function 197. As a result, the processing circuit 19 extracts patient information from the electronic medical record. The extraction source of the patient information is not limited to the electronic medical record, and may be a questionnaire regarding the subject, health examination information, or the like. The questionnaire and the health examination information are associated with, for example, an electronic medical record.

FIG. 2 is a diagram showing an example of the relationship between the acquisition target by the acquisition function 193 and the extraction target by the extraction function 197. As shown in FIG. 2, the inspection purpose is specified from the inspection order of the interpretation report, triggered by the acquisition of the interpretation report. Next, extraction targets such as test results, diagnosis results, and treatment results for the same test purpose as the interpretation report are extracted.

The calculation function 199 calculates the test reliability indicating the reliability (or credibility) of the first test result based on the first test result and the diagnosis result. The test reliability is, for example, a numerical value of the specificity (true positive, true negative, false positive, false negative) for the first test result as a value from 0 to 1 or a percentage. The processing circuit 19 uses the calculation function 199 to calculate, for example, the concordance rate between the diagnosis result and the findings in the interpretation report as the inspection reliability. The processing circuit 19 calculates the diagnostic reliability indicating the reliability (or credibility) of the diagnostic result based on the diagnostic result and the treatment result. The diagnostic reliability is, for example, a numerical value of specificity (true positive, true negative, false positive, false negative) for a diagnosis result as a value from 0 to 1 or a percentage. The processing circuit 19 calculates, for example, the concordance rate between the treatment result and the diagnosis result as the diagnosis reliability.

The integration function 201 integrates the data extracted by the extraction function 197 and the examination purpose specified by the specific function 195 as learning data that can be used for learning, based on the order number and the patient ID. Specifically, the processing circuit 19 generates learning data by integrating the first inspection data, the second inspection data, the learning image, and the inspection purpose by the integrated function 201. More specifically, the second inspection data is weighted in advance according to the type of the second inspection. The weight may be set individually according to the clinical type such as the test site and the test purpose. By default, the weight of the pathological test is the largest of the second tests, because the results of the pathological test contribute the most to the diagnostic results. Further, the weight is displayed on the display 15 and can be changed by a user's instruction via the input interface 17.

The processing circuit 19 assigns the same data ID as the inspection purpose to the learning data by the integrated function 201. In selecting learning data, a data ID is used instead of the patient name from the viewpoint of anonymity and security. The processing circuit 19 may further integrate at least one of the patient information, the test reliability, the diagnosis result, the treatment result, and the diagnosis reliability into the learning data.

Further, the processing circuit 19 may integrate the procedure name and the like in the operation implementation report into the learning data by the integrated function 201. At this time, the processing circuit 19 first acquires the operation execution report from the operation department system based on the order information by the acquisition function 193. Next, the processing circuit 19 uses the extraction function 197 to extract the procedure name and the like from the operation implementation report using natural language processing. Subsequently, the processing circuit 19 integrates the procedure name and the like into the learning data by the integration function 201.

The learning data is stored in the memory 13 in a text format such as CSV (Comma Separated Values). The storage format of the learning data is not limited to CSV, and may be another storage format such as a database method or an XML (XML (eXtensible Markup Language) method). In addition, the generation of learning data may be updated again as appropriate after the electronic medical record, which is the final stage of medical treatment for the subject, is confirmed. For example, after the electronic medical record is confirmed, all the inspection results may be collectively extracted after the medical record is confirmed by the user's instruction via the input interface 17. At this time, the user appropriately updates the learning data by inputting through the input interface 17.

FIG. 3 is a diagram showing an example of learning data generated by integrating various data by the integration function 201. As shown in FIG. 3, the processing circuit 19 integrates the data extracted by the extraction function 197 with the inspection purpose specified by the specific function 195 and the calculation result calculated by the calculation function 199 by the integration function 201. Automatically generates learning data to which a data ID is assigned. Of the plurality of weights shown in FIG. 3, the weight given to the pathological test result is the largest. The processing circuit 19 stores the generated learning data in the memory 13. The processing circuit 19 stores the learning data in the memory 13 as described above according to the completion of the interpretation report. That is, the memory 13 stores a group of learning data (hereinafter, referred to as a learning data set).

The selection function 203 learns from a group of learning data (learning data set) stored in the memory 13 based on at least one of the inspection reliability, the diagnostic reliability, and the learning content for the machine learning model. Select the learning data to be used. For example, according to the user's instruction via the input interface 17, the processing circuit 19 selects the case and learning data desired by the user by the selection function 203. Specifically, the processing circuit 19 applies a selection method such as a filter set by the user to the learning data set to select target / non-target data related to learning. The processing circuit 19 may shape the selected learning data according to a learning system (not shown) and output it in a specific form. The output format of the learning data can be selected by the user according to the input method of the learning system. The input method of the learning system is, for example, CSV or text, image data, a database method, or an XML method.

The overall configuration of the medical information processing device 1 and the hospital information system 3 has been described above. Hereinafter, the processing procedure of the learning data generation process for automatically generating the learning data will be described. In order to make the explanation concrete, it is assumed that the first test result is an interpretation report and the second test result is a sample test report. FIG. 4 is a diagram showing an example of a processing procedure of the learning data generation process.

(Learning data generation process)
(Step S401)
The processing circuit 19 receives an image interpretation report from the medical image management system 39 in response to a click of the "confirm" button on the report creation screen, that is, when the image interpretation report corresponding to the first inspection result is completed by the acquisition function 193. get. The processing circuit 19 acquires order information regarding the interpretation report from the electronic medical record system 33 based on the acquired interpretation report. The acquired interpretation report and order information are temporarily stored in the memory 13.

(Step S402)
The processing circuit 19 specifies the inspection purpose in the order information by the specific function 195. Specifically, the processing circuit 19 specifies the inspection purpose described in the order information by applying natural language processing to the order information. The processing circuit 19 assigns a data ID for the specified inspection purpose. The inspection purpose to which the data ID is assigned is temporarily stored in the memory 13.

(Step S403)
The processing circuit 19 acquires a sample test report for the specified test purpose from the clinical laboratory department information system 41 by the acquisition function 193. For example, the processing circuit 19 stores a sample test report having the same patient ID as the patient ID described in the order information and having the same test purpose as the specified test purpose in the clinical laboratory department information system 41. Get from. The processing circuit 19 acquires the electronic medical record of the subject from the electronic medical record system 33 based on the order information by the acquisition function 193. The acquired certification inspection report and electronic medical record are temporarily stored in the memory 13.

(Step S404)
The processing circuit 19 extracts patient information, a diagnosis result, and a treatment result from an electronic medical record by an extraction function 197. When the diagnosis result and the treatment result are not reflected in the electronic medical record, the processing circuit 19 receives the diagnosis result and the treatment from the electronic medical record acquired after the lapse of a predetermined period, for example, the day before the re-examination date and time for the subject. Extract the results. The processing circuit 19 extracts the image interpretation result and the lesion image from the image interpretation report, and extracts the sample test result from the sample test report. The extracted patient information, diagnosis result, treatment result, image interpretation result, lesion image, and sample test result are temporarily stored in the memory 13. When the first inspection and the second inspection are further scheduled, the processing circuit 19 repeats the processing of steps S401 to S404 after the date and time when these inspections are performed.

(Step S405)
The processing circuit 19 calculates the inspection reliability based on the interpretation report and the diagnosis result by the calculation function 199. Further, the processing circuit 19 calculates the diagnostic reliability based on the diagnosis result and the treatment result. The calculated inspection reliability and diagnostic reliability are temporarily stored in the memory 13.

(Step S406)
The processing circuit 19 uses the integrated function 201 to obtain patient information, a diagnosis result, a treatment result, an image interpretation result, a lesion image, a sample test result, a test reliability, and a diagnosis reliability as learning data. Integrate. Next, the processing circuit 19 assigns a system ID to the learning data. The processing circuit 19 stores the learning data to which the system ID is assigned in the memory 13. The processing circuit 19 repeats the processing of steps S401 to S406 according to the completion of the interpretation report. As a result, the memory 13 stores a group of learning data (learning data set) according to the total number of interpretation reports. At this time, the memory 13 may function as a database of learning data. This step completes the learning data generation process.

When using a group of learning data stored in the memory 13, the processing circuit 19 is based on at least one of the inspection reliability, the diagnostic reliability, and the learning content for the machine learning model by the selection function 203. Then, the learning data used for learning is selected from the group of learning data stored in the memory 13. For example, when learning is performed on an opportunity learning model that inputs a lesion image and outputs a diagnosis result, the processing circuit 19 selects learning data having high diagnostic reliability and examination reliability by the selection function 203. ..

According to the medical information processing apparatus 1 according to the embodiment described above, the inspection purpose of the first inspection is specified based on the first inspection result regarding the medical image acquired by the first inspection on the subject, and the inspection purpose is used. The second test result in the second test performed on the subject is acquired from the medical data server based on the order information regarding the first test result, and the first test result can be used for learning the machine learning model. The first test data is extracted, the second test data that can be used for the learning is extracted from the second test result, the learning image that can be used for the learning is extracted from the medical image, and the first test data and the second test are used. The data, the learning image, and the inspection purpose are integrated as learning data that can be used for the learning. As a result, according to the information processing apparatus 1 for medical use, learning data can be automatically generated without bothering the image interpreting doctor, triggered by the acquisition of the first examination result.

Further, according to the medical information processing apparatus 1, the electronic medical record of the subject is further acquired from the medical data server based on the order information, and the diagnosis result and diagnosis of the subject are made based on the first test result and the second test result. At least one of the treatment result and the patient information corresponding to the result is extracted from the electronic medical record, and at least one of the diagnosis result, the treatment result and the patient information is further integrated into the learning data. Further, according to the information processing apparatus 1 for medical use, the test reliability indicating the reliability of the first test result is calculated based on the first test result and the diagnosis result, and the test reliability is further integrated into the learning data. do. Further, according to the information processing apparatus 1 for medical use, the diagnostic reliability indicating the degree of reliability of the diagnostic result is calculated based on the diagnostic result and the treatment result, and the diagnostic reliability is further integrated into the learning data. As a result, according to the medical information processing apparatus 1, at least one of the diagnosis result, the treatment result, the patient information, the examination reliability, and the diagnosis reliability can be converted into learning data without bothering the user. It can be further integrated automatically.

From the above, according to the information processing apparatus 1 for medical use, various data related to learning data (second examination result) based on the interpretation report (first examination result) by the radiologist in the normal image interpretation work. Can be collected automatically. As a result, it is not necessary for the interpreting doctor to confirm the pathological examination result and the second examination result, which is a medical document such as a medical record, when creating the learning data. It is possible to improve the efficiency of the creation of. In addition, according to the medical information processing apparatus 1, the diagnosis result, the treatment result, the patient information, the examination reliability, and the diagnosis reliability can be further utilized in the selection of the learning data, and the machine learning model can be used. It is possible to perform more accurate learning.

(First application example)
The first application example is to update the past test reliability (hereinafter referred to as the past test reliability) for the same test purpose as the test purpose to the calculated test reliability based on the treatment result. The first application example corresponds to a situation in which a set of examination, diagnosis, and treatment for a subject is performed, and then a re-examination or follow-up (follow-up, follow-up) is performed on the same subject.

The processing circuit 19 has the same examination purpose as the examination purpose specified by the acquisition function 193, and the past learning data regarding the same patient information as the extracted patient information (hereinafter, referred to as past learning data). Is obtained from the training data set. The processing circuit 19 acquires the past inspection reliability from the past learning data by the acquisition function 193. The processing circuit 19 performs natural language processing based on a predefined dictionary on the acquired electronic medical record by the extraction function 197, and determines whether or not there is a change in the state such as the diagnosis name, stage, and severity of the disease, and treatment. Extract the presence or absence of.

The processing circuit 19 recalculates the past diagnosis reliability, which indicates the credibility of the past inspection, based on the extracted information by the calculation function 199. When performing the recalculation for the past inspection reliability, the display 15 may indicate the necessity of reviewing the inspection reliability, that is, the recalculation. At this time, the user selects whether or not the review is necessary via the input interface 17. The processing circuit 19 adds or subtracts a predetermined value to the past test reliability based on the treatment result and the diagnosis result (hereinafter referred to as the past diagnosis result) in the past learning data. Recalculate past inspection reliability. The predetermined value may be stored in the memory 13 in advance, or may be set by a user such as a doctor or an administrator via the input interface 17. Further, the predetermined value may be different between addition and subtraction.

Specifically, when the treatment results in reducing the stage and severity of the disease in the comparison between the treatment result and the past diagnosis result, that is, when the past diagnosis result is determined to be valid, the processing circuit 19 determines. The calculation function 199 adds a predetermined value to the past inspection reliability. Further, when the treatment does not reduce or worsen the severity of the stage and the disease in the comparison between the treatment result and the past diagnosis result, the processing circuit 19 subtracts a predetermined value from the past test reliability. Further, when the diagnosis name is changed in the comparison between the treatment result and the past diagnosis result, the processing circuit 19 subtracts a predetermined value from the past examination reliability.

The display 15 may display a comparison between the treatment result and the past diagnosis result and the reliability of the past examination. At this time, the past inspection reliability may be changed according to the instruction of the user via the input interface 17.

The processing circuit 19 updates the past inspection reliability with the recalculated inspection reliability by the integrated function 201. The processing circuit 19 updates the past learning data by integrating the updated past inspection reliability into the past learning data. The processing circuit 19 incorporates the updated past learning data into the learning data set and stores it in the memory 13.

According to the medical information processing apparatus 1 of the first application example according to the above-described embodiment, past learning data having past test reliability for the same test purpose as the test purpose is acquired, and the treatment result and the past By adding or subtracting a predetermined value to the past test reliability based on the diagnosis result in the training data of, the past test reliability is recalculated and the past test reliability is calculated. , Update to recalculated inspection reliability. For example, if the correlation between the treatment result and the past diagnosis result is low and the reliability of the past examination is high, the credibility of the interpretation result in the interpretation report is low. At this time, according to the information processing apparatus 1 for medical use, it is possible to reduce the inspection reliability indicating the credibility of the image interpretation result. In addition, when the correlation between the treatment result and the past diagnosis result is high and the reliability of the past examination is high, the credibility of the interpretation result in the interpretation report becomes high. At this time, according to the information processing apparatus 1 for medical use, it is possible to increase the reliability of the examination, which indicates the credibility of the interpretation result.

From the above, according to the medical information processing apparatus 1 according to the first application example, it is possible to adjust the past test reliability according to the effectiveness of the treatment result and the accuracy of the diagnosis name in the past diagnosis result. can. In addition, according to the information processing apparatus 1 for medical use, the accuracy of the learning data can be improved, and more accurate learning can be performed on the machine learning model.

(Second application example)
A second application example is to update the past diagnostic reliability (hereinafter referred to as the past diagnostic reliability) for the same test purpose as the test purpose to the calculated diagnostic reliability based on the treatment result. The second application example corresponds to a situation in which a set of examination, diagnosis, and treatment for a subject is performed, and then a re-examination or follow-up (follow-up, follow-up) is performed on the same subject.

The processing circuit 19 acquires the past learning data from the learning data set by the acquisition function 193. The processing circuit 19 acquires an electronic medical record of the same subject as the subject related to the treatment result from the electronic medical record server device. The processing circuit 19 uses the extraction function 197 to extract from the acquired electronic medical record whether or not there is a change in the state such as the diagnosis name, stage, and severity of the disease, and whether or not there is treatment.

The processing circuit 19 recalculates the past diagnosis reliability, which indicates the credibility of the past diagnosis, based on the extracted information by the calculation function 199. When performing the recalculation for the past diagnosis reliability, the display 15 may indicate the necessity of reviewing the diagnosis reliability, that is, the recalculation. At this time, the user selects whether or not the review is necessary via the input interface 17. The processing circuit 19 adds or subtracts a predetermined value to the past diagnosis reliability based on the treatment result and the diagnosis result (hereinafter referred to as the past diagnosis result) in the past learning data. Recalculate the reliability of past diagnosis. The predetermined value may be stored in the memory 13 in advance, or may be set by a user such as a doctor or an administrator via the input interface 17. Further, the predetermined value may be different between addition and subtraction.

Specifically, when the treatment results in reducing the stage and severity of the disease in comparison between the treatment result and the past diagnosis result, that is, when the past diagnosis result is determined to be valid, the processing circuit 19 determines. The calculation function 199 adds a predetermined value to the past diagnosis reliability. Further, when the treatment does not reduce or worsen the severity of the stage and the disease in the comparison between the treatment result and the past diagnosis result, the processing circuit 19 subtracts a predetermined value from the past diagnosis reliability. Further, when the diagnosis name is changed in the comparison between the treatment result and the past diagnosis result, the processing circuit 19 subtracts a predetermined value from the past diagnosis reliability.

The display 15 may display a comparison between the treatment result and the past diagnosis result and the reliability of the past diagnosis. At this time, the past diagnosis reliability may be changed according to the instruction of the user via the input interface 17.

The processing circuit 19 updates the past diagnostic reliability with the recalculated diagnostic reliability by the integrated function 201. The processing circuit 19 updates the past learning data by integrating the updated past diagnosis reliability into the past learning data. The processing circuit 19 incorporates the updated past learning data into the learning data set and stores it in the memory 13.

According to the medical information processing apparatus 1 of the second application example according to the above-described embodiment, past learning data having past diagnostic reliability for the same examination purpose as the examination purpose is acquired, and the treatment result and the past By adding or subtracting a predetermined value to the past diagnostic reliability based on the diagnostic result in the training data of, the past diagnostic reliability is recalculated and the past diagnostic reliability is calculated. , Update to recalculated diagnostic reliability. For example, if the correlation between the treatment result and the past diagnosis result is low, the credibility of the diagnosis result is low. At this time, according to the information processing apparatus 1 for medical use, it is possible to reduce the reliability of the diagnosis, which indicates the credibility of the diagnosis result. Moreover, when the correlation between the treatment result and the past diagnosis result is high, the credibility of the diagnosis result is high. At this time, according to the information processing apparatus 1 for medical use, it is possible to increase the reliability of the test indicating the credibility of the diagnosis result.

From the above, according to the medical information processing apparatus 1 according to the second application example, the past diagnosis reliability can be adjusted according to the effectiveness of the treatment result and the accuracy of the diagnosis name in the past diagnosis result. can. In addition, according to the information processing apparatus 1 for medical use, the accuracy of the learning data can be improved, and more accurate learning can be performed on the machine learning model.

When the technical idea in the present embodiment is realized by the learning data generation program, the inspection purpose of the first examination is specified on the computer based on the first examination result regarding the medical image acquired by the first examination on the subject. , The second test result in the second test performed on the subject for the purpose of the test is acquired from the medical data server based on the order information about the first test result, and the first test result is used for learning the machine learning model. The available first test data is extracted, the second test data that can be used for the learning is extracted from the second test result, the learning image that can be used for the learning is extracted from the medical image, and the first test data is extracted. And the second inspection data and the learning image are integrated as learning data that can be used for the learning.

For example, the learning data generation process can be realized by installing the learning data generation program on the computer in the hospital information system 3 and expanding these on the memory. At this time, a program that allows a computer to execute the method can be stored and distributed in a storage medium such as a magnetic disk (hard disk, etc.), an optical disk (CD-ROM, DVD, etc.), or a semiconductor memory. .. Since the processing procedure and the effect in the learning data generation program are the same as those in the embodiment, the description thereof will be omitted.

Further, when the technical idea in the present embodiment is realized by the learning data generation method, the learning data generation method is based on the first examination result regarding the medical image acquired by the first examination on the subject. The second test result in the second test performed on the subject for the test purpose is obtained from the medical data server based on the order information regarding the first test result, and from the first test result. The first test data that can be used for learning the machine learning model is extracted, the second test data that can be used for the learning is extracted from the second test result, and the learning image that can be used for the learning is extracted from the medical image. Then, the first inspection data, the second inspection data, and the learning image are integrated as learning data that can be used for the learning. Since the processing procedure and the effect in the learning data generation method are the same as those in the embodiment, the description thereof will be omitted.

According to the embodiments and application examples described above, learning data can be easily generated.

Although some embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the gist of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, as well as in the scope of the invention described in the claims and the equivalent scope thereof.

1 Medical information system 3 Hospital information system 11 Communication interface 13 Memory 15 Display 17 Input interface 19 Processing circuit 31 Communication network 33 Electronic chart system 35 Radiation department information system 36 Report creation device 37 Medical diagnostic imaging system 39 Medical image management system 41 Clinical Inspection department information system 43 Other department system 191 System control function 193 Acquisition function 195 Specific function 197 Extraction function 199 Calculation function 201 Integration function 203 Selection function

Claims (11)

A specific part that specifies the inspection purpose of the first examination based on the first examination result regarding the medical image obtained by the first examination for the subject, and
An acquisition unit that acquires the second test result in the second test performed on the subject for the purpose of the test from the medical data server based on the order information regarding the first test result.
The first test data that can be used for learning the machine learning model is extracted from the first test result, the second test data that can be used for the learning is extracted from the second test result, and the medical image is used for the learning. An extraction unit that extracts available learning images, and
An integration unit that integrates the first inspection data, the second inspection data, the learning image, and the inspection purpose as learning data that can be used for the learning.
Medical information processing device equipped with. Based on the order information, the acquisition unit further acquires the electronic medical record of the subject from the medical data server.
The extraction unit extracts the diagnostic result of the subject based on the first test result and the second test result from the electronic medical record.
The integration unit further integrates the diagnosis result into the learning data.
The medical information processing device according to claim 1. A calculation unit for calculating the test reliability indicating the reliability of the first test result based on the first test result and the diagnosis result is further provided.
The integration unit further integrates the inspection reliability into the learning data.
The medical information processing device according to claim 2. The extraction unit extracts the treatment result for the subject corresponding to the diagnosis result from the electronic medical record.
The integration unit further integrates the treatment results into the learning data.
The medical information processing device according to claim 3. The acquisition unit acquires past learning data having past inspection reliability for the same inspection purpose as the inspection purpose.
The calculation unit adds or subtracts a predetermined value to the past test reliability based on the treatment result and the diagnosis result in the past learning data, thereby performing the past test. Recalculate the reliability and
The integration unit updates the past inspection reliability with the recalculated inspection reliability.
The medical information processing device according to claim 4. The calculation unit calculates the diagnostic reliability indicating the degree of reliability of the diagnostic result based on the diagnostic result and the treatment result.
The integration unit further integrates the diagnostic reliability into the learning data.
The medical information processing device according to claim 4 or 5. The acquisition unit acquires past learning data having past diagnostic reliability for the same inspection purpose as the inspection purpose.
The calculation unit performs the addition or subtraction of a predetermined value with respect to the past diagnosis reliability based on the treatment result and the diagnosis result in the past learning data, thereby performing the past diagnosis. Recalculate the reliability and
The integration unit updates the past diagnostic reliability with the recalculated diagnostic reliability.
The medical information processing device according to any one of claims 4 to 6. A memory for storing a plurality of the learning data and
Based on at least one of the inspection reliability, the diagnostic reliability, and the learning content for the machine learning model, the learning data used for the learning is obtained from the group of the learning data stored in the memory. Further equipped with a selection unit to select
The medical information processing device according to claim 6. The extraction unit extracts patient information regarding the subject from at least one of the electronic medical record, the questionnaire, and the health examination information.
The integration unit further integrates the patient information in the learning data.
The medical information processing device according to any one of claims 2 to 8. On the computer
The purpose of the first test is specified based on the results of the first test on the medical image obtained by the first test on the subject.
The second test result in the second test performed on the subject for the test purpose is acquired from the medical data server based on the order information regarding the first test result.
The first test data that can be used for learning the machine learning model is extracted from the first test result, the second test data that can be used for the learning is extracted from the second test result, and the medical image is used for the learning. Extract available learning images
Integrating the first inspection data, the second inspection data, and the learning image as learning data that can be used for the learning.
A data generation program for learning that realizes. The purpose of the first test is specified based on the results of the first test on the medical image obtained by the first test on the subject.
The second test result in the second test performed on the subject for the test purpose is acquired from the medical data server based on the order information regarding the first test result.
The first test data that can be used for learning the machine learning model is extracted from the first test result, the second test data that can be used for the learning is extracted from the second test result, and the medical image is used for the learning. Extract available learning images
A learning data generation method that integrates the first inspection data, the second inspection data, and the learning image as learning data that can be used for the learning.

Images