WO2024057839A1 - Acceptance assistance system, acceptance assistance method, and acceptance assistance program - Google Patents

Acceptance assistance system, acceptance assistance method, and acceptance assistance program Download PDF

Info

Publication number
WO2024057839A1
WO2024057839A1 PCT/JP2023/030144 JP2023030144W WO2024057839A1 WO 2024057839 A1 WO2024057839 A1 WO 2024057839A1 JP 2023030144 W JP2023030144 W JP 2023030144W WO 2024057839 A1 WO2024057839 A1 WO 2024057839A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
medical examination
health checkup
abnormal value
underwriting
Prior art date
Application number
PCT/JP2023/030144
Other languages
French (fr)
Japanese (ja)
Inventor
真 上田
浩一 伊藤
Original Assignee
真 上田
浩一 伊藤
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 真 上田, 浩一 伊藤 filed Critical 真 上田
Publication of WO2024057839A1 publication Critical patent/WO2024057839A1/en

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/08—Insurance

Definitions

  • the present invention relates to an underwriting support system, an underwriting support method, and an underwriting support program.
  • the insurance company checks the age, occupation, and health condition of the insured person to determine whether they can underwrite the insurance and what is the appropriate insurance amount. It was necessary to determine whether In particular, when it comes to checking health conditions, it has been a problem that costs are required for operations such as converting the insured person's health certificate into data, confirming the authenticity of the data, and conducting insurance assessments based on such data.
  • Non-Patent Document 1 proposes to automate the data conversion and input of health certificates using atypical OCR (Optical Character Recognition) technology by utilizing a handy scanner and image transmission technology, and to It discloses that it has improved the notification method for diseases, etc., and that an automatic underwriting appraisal engine determines whether or not the company can underwrite the product and contract conditions by comparing the contents of the notifications imported through the notification method with the appraisal rule database.
  • OCR Optical Character Recognition
  • Patent Document 1 discloses a system that can smoothly perform application procedures for group credit life insurance, and an application content receiving means 46 records the application content of an applicant or a receiver, and an application information transmitting means 42 sends the application details and attached document images to the server device 2, the reception confirmer terminal device 6 accesses the server device 2, checks whether there are any deficiencies in the application information, and sends the reception confirmation result to the server device 2.
  • the reception confirmation result receiving means 26 of the server device 2 receives the reception confirmation result and records it in the recording unit 34, and the temporary access setting means 28 transmits the confirmation result if the reception confirmation result is "incomplete".
  • a temporary access key is set so that the applicant can temporarily access the application information in the recording unit 34, and the correction notification sending means 30 notifies the applicant of the correction with the temporary access key.
  • a technique is disclosed in which the notification is sent, the applicant accesses the application information in the recorder 34 by means of a remediation temporary access key, and the remediation information is transmitted.
  • Non-Patent Document 1 states that although it was possible to automatically convert health certificates into data using OCR technology, converting health certificates into data using OCR technology is difficult due to the diverse formats of health certificates and the lack of OCR accuracy. The problem was that it was detected as an incorrect value.
  • Patent Document 1 has room for improvement in that the task of checking for deficiencies in application information is not automated.
  • the present invention has been made in view of the above-mentioned problems, and it is possible to improve business efficiency by accurately checking for errors and false values included in insurance applications during automatic assessment operations or in the preceding stage.
  • the issue to be solved is to provide technology that supports the decision to accept underwriting.
  • the present invention provides first medical examination data generated based on the medical examination image data of the insurance applicant, second medical examination data input by the insurance applicant, an acquisition unit that acquires an item, a detection unit that detects an abnormal value including a mismatch between items of the first health checkup data and the second health checkup data, and a detection unit that reacquires the item in which the abnormal value was detected. , an application data generation unit that generates application data.
  • the present invention also provides a step of acquiring first medical examination data generated based on the medical examination image data of the insurance applicant and second medical examination data input by the insurance applicant;
  • a computer executes the steps of detecting an abnormal value including a mismatch between the item of the data and the second health checkup data, and reacquiring the item for which the abnormal value was detected and generating application data. do.
  • the present invention also provides an acquisition unit that acquires first medical examination data generated based on medical examination image data of the insurance applicant and second medical examination data input by the insurance applicant; a detection unit that detects abnormal values including mismatch between items of the medical examination data and the second medical examination data; and an application data generation unit that reacquires the items for which the abnormal values have been detected and generates application data. , to make the computer function as.
  • the application data generation unit generates the application data based on second medical examination data re-inputted by the insurance applicant regarding the item in which the abnormal value was detected.
  • the invention further includes a registration reception unit that registers a facial image of the insurance applicant, and the application data generation unit authenticates the captured facial image of the insurance applicant.
  • the application data is generated based on the medical examination data or the second medical examination data.
  • the present invention further includes a setting reception unit that receives detection setting information in which a range of normal values or abnormal values is set for each item of health checkup data, and the detection unit is configured to detect the first health check based on the detection setting information. An abnormal value in at least one item of the medical examination data and the second medical examination data is detected.
  • the setting reception unit further receives detection setting information in which an abnormal value is set for a combination of values of a plurality of items of medical examination data.
  • the setting reception unit further receives detection setting information in which abnormal values are set for items that have a correlation with time-series changes included in the user's past health checkup data and current health checkup data. With such a configuration, abnormal values included in the medical examination data can be detected.
  • the application data and clinical data are input to a machine-learned judgment model using the insurance applicant's clinical data and health checkup data as input data and the insurance assessment result as output data, It is equipped with an assessment department that outputs insurance assessment results.
  • insurance assessment operations can be suitably automated by taking clinical data into consideration in addition to medical examination data.
  • FIG. 1 shows a system configuration diagram of an underwriting support system according to the present embodiment.
  • a hardware configuration diagram of each device according to the present embodiment is shown.
  • a flowchart of underwriting processing according to the present embodiment is shown.
  • a screen display example of an input screen according to the present embodiment is shown.
  • a flowchart of detection processing according to the present embodiment is shown.
  • the computer that constitutes the underwriting support system has a calculation device such as a CPU (Central Processing Unit) and a storage device.
  • the computer can function as an underwriting support device by using the arithmetic device to execute an underwriting support program stored in the storage device.
  • This specification describes an example of an underwriting support system that supports underwriting operations related to insurance such as life insurance, health insurance, and medical insurance. As long as a judgment is made, it is not particularly limited to insurance.
  • other underwriting services may include renewal of driver's licenses, etc., immigration/departure services that require medical examination results, and verification of the authenticity of medical examination results at companies, etc.
  • FIG. 1 shows a block diagram of the underwriting support system 1.
  • the underwriting support system 1 includes an underwriting support device 2, a user terminal 3, a medical institution system 4, a business terminal 5 operated by an insurance company, and a storage unit DB as a database, and these components are , are communicably connected via a communication network NW.
  • the user terminal 3 is a terminal device operated by an insured person who applies for insurance, and there may be a plurality of user terminals.
  • the storage unit DB only needs to be configured to be able to communicate data with the underwriting support device 2, and may be installed inside the underwriting support device 2.
  • the underwriting support device 2 includes, as functional components, a setting reception section 21 that receives setting input of various information including criteria for abnormal value detection and judgment models, a registration reception section 22 that receives registration of user information, and medical examination data.
  • a setting reception section 21 that receives setting input of various information including criteria for abnormal value detection and judgment models
  • a registration reception section 22 that receives registration of user information
  • medical examination data an acquisition unit 23 that acquires the data
  • a detection unit 24 that detects abnormal values included in the medical examination data
  • an application data generation unit 25 that generates insurance-related application data based on the medical examination data
  • an insurance-related assessment based on the application data includes an assessment unit 26 that executes processing.
  • a part of the functional configuration of the underwriting support device 2 may be a configuration installed in the user terminal 3, and the various functional configurations described above may be realized by the underwriting support device 2 and the user terminal 3 as a whole. .
  • the medical institution system 4 is installed in a medical institution such as a hospital.
  • the medical institution system 4 carries out health examinations of users including insurance applicants, and stores the medical examination data in a medical institution database in association with user identifiers.
  • the medical institution system 4 stores clinical data including the user's medical history at medical institutions, current medical history, etc. in the medical institution database in association with the user's identifier.
  • the medical institution system 4 provides the requested data in response to the user's request regarding various data from the underwriting support device 2.
  • FIG. 2(a) shows a hardware configuration diagram of the underwriting support device 2.
  • the underwriting support device 2 includes a control unit 201 including a CPU, and a storage unit including a main storage device such as a RAM (Random Access Memory), an auxiliary storage device such as an HDD (Hard Disk Drive), an SSD (Solid State Drive), and a flash memory. 202, a communication unit 203 for controlling communication with the communication network NW, and the like.
  • the storage unit 202 of the underwriting support device 2 is an OS (Operating System). ) and an underwriting support program that implements various functional components in cooperation with the OS, and the control unit 201 can implement various functional configurations by executing the underwriting support program.
  • the underwriting support device 2 is configured as a server device. Note that the server device can be of a cloud type, an on-premise type, or the like as appropriate.
  • FIG. 2(b) shows a hardware configuration diagram of the user terminal 3.
  • the user terminal 3 includes a control unit 301 such as a CPU, a storage unit 302 such as a main memory or auxiliary storage, a communication unit 303, an input unit 304 such as a touch panel, a mouse, and a keyboard, and a display unit such as a display. 305, and an imaging unit 306 such as a camera.
  • a control unit 301 such as a CPU
  • storage unit 302 such as a main memory or auxiliary storage
  • a communication unit 303 such as a main memory or auxiliary storage
  • an input unit 304 such as a touch panel, a mouse, and a keyboard
  • a display unit such as a display.
  • an imaging unit 306 such as a camera.
  • the vendor terminal 5 is also assumed to have a similar hardware configuration.
  • the detection unit 24 of the underwriting support system 1 detects abnormal values for items of medical examination data transmitted via the user terminal 3 operated by the user applying for insurance.
  • the health checkup data consists of the first health checkup data generated based on the health checkup image data obtained by capturing the health checkup form of the health checkup that the user received, and the data that the user refers to the health checkup form and inputs via the input screen. and second medical examination data to be input.
  • medical examination data when two pieces of medical examination data are not distinguished, they are simply referred to as medical examination data.
  • the detection unit 24 detects abnormal values in the items of the medical examination data based on the detection setting information set by the setting reception unit 21.
  • the detection setting information includes settings regarding the range of abnormal values or normal values that serve as detection standards, and a detection model that inputs medical examination data and outputs abnormal value detection results.
  • abnormal values are items of first health checkup data acquired based on health checkup image data obtained by capturing a user's health checkup form, and second health checkup data input via the user terminal 3. This includes values where the items do not match, and outliers that are judged to be inadequate based on the correlation between the items of each health checkup data.
  • the setting reception unit 21 receives and stores detection setting information for detecting abnormal values for each item of medical examination data, abnormal values for combinations of items, and abnormal values for correlations due to time-series changes in items. Store in the section DB.
  • the detection setting information is inputted, for example, via the provider terminal 5 operated by the insurance company or the medical institution system 4.
  • the setting reception unit 21 receives detection setting information regarding the range of normal values or abnormal values for each item of medical examination data.
  • the BMI is set within a normal value range of 5 as a lower limit and 100 as an upper limit
  • the detection unit 24 detects a value that deviates from the normal value as an abnormal value in an item of medical examination data.
  • the range of normal values or abnormal values is not limited to those described above, and can be set appropriately for each item.
  • the setting reception unit 21 receives, as detection setting information, a combination of a plurality of items of medical examination data that is determined to be an abnormal value.
  • the combination of items includes, for example, user attributes such as gender and age, and one or more items of medical examination data. Examples of combinations (combinations A to C) that are determined to be abnormal values are shown below.
  • Combination A Male, 50 years old, BMI: 30, Waist circumference: 60 cm
  • Combination B Male, 60 years old, BMI: 35
  • systolic blood pressure 80
  • Combination C Female, 25 years old, hemoglobin: 15.0, serum iron: 10
  • Combination A is determined to be potentially defective because the BMI value indicates a tendency towards obesity, but the waist circumference value is below the male standard.
  • Combination B is determined to be potentially defective because the BMI value indicates a tendency toward obesity, whereas the systolic blood pressure value indicates low blood pressure.
  • Combination C is determined to be potentially defective because the hemoglobin value shows the standard value for women, but the serum iron value shows a tendency toward anemia.
  • the detection setting information for combinations that are determined to be abnormal values may be a machine-learned detection model.
  • the detection model is machine-learned using a dataset that takes multiple items of health check data as input and outputs whether the combination of the items is a normal value or an abnormal value.
  • the detection model generated by machine learning can accept input of multiple items of health check data and output whether the combination of those items is a normal value or an abnormal value.
  • the setting reception unit 21 receives detection setting information that sets abnormal values of items due to time-series changes between the user's past health checkup data and current health checkup data.
  • either positive correlation or negative correlation is set for two or more items of medical checkup data.
  • the detection unit 24 detects a change in a second item that is set to have a positive or negative correlation with a change in the first item of the same user's previous health checkup data and current health checkup data. .
  • the detection unit 24 detects that the first item and/or the second item is an abnormal value when the change in the second item is contrary to the set correlation. For example, BMI and waist circumference are set as having a positive correlation.
  • the waist circumference decreases as BMI increases, which contradicts the positive correlation.
  • BMI and/or waist circumference values are detected to be abnormal values. Note that a tolerance value may be further set in the case where the set correlation is violated.
  • the abnormal value detection setting information based on time-series changes may be stored in the storage unit DB as a machine-learned detection model.
  • the detection model uses a dataset whose input is each item of previous health checkup data and each item of the current health checkup data, and whose output is whether the time-series changes in these items are normal or abnormal values.
  • Machine learned A determination model generated by machine learning can accept input of previous and current health checkup data and output whether the time-series changes in those items are normal values or abnormal values.
  • the settings reception unit 21 inputs the user's medical examination data and clinical data, receives settings for a determination model that outputs assessment results, and stores the settings in the storage unit DB.
  • the determination model is a model generated by machine learning. Note that the machine learning process may be executed by a device external to the underwriting support system 1, and the setting reception unit 21 may be configured to receive the learned determination model.
  • Clinical data includes past medical history, current medical history, drug prescription history, treatment history, presence or absence of complications, etc.
  • the clinical data may be data generated based on a notification obtained from the medical institution system 4.
  • the assessment results include whether or not the insurance can be underwritten, the amount that can be underwritten, and proposed insurance products.
  • Data sets used for machine learning of the judgment model include, for example, blood pressure values indicating a tendency toward hypertension included in health checkup data, prescription history of antihypertensive drugs included in clinical data, and other medical history and no complications. and are input data, and an assessment result showing a tendency to relax underwriting conditions (for example, underwriting is acceptable) is set as output data.
  • underwriting conditions for example, underwriting is acceptable
  • the judgment model is a model obtained by machine learning the combination of medical examination data and clinical data, and the tendency of judgment of the assessment result by the evaluator based on the combination.
  • FIG. 3 shows a flowchart regarding processing in the underwriting support device 2.
  • FIG. 3 shows an example of a processing flow when an insurance company guides a user to input various information necessary for insurance underwriting, and the user inputs various information via the user terminal 3.
  • step S11 the registration reception unit 22 receives input of items necessary for insurance underwriting such as the user's user name, gender, age, address, place of work, and annual income via the user terminal 3, and stores it as basic information.
  • the basic information is given a unique user ID to the user.
  • the user ID is associated with the patient number and ID of the user managed by the medical institution system 4.
  • the basic information may be input via the vendor terminal 5.
  • the basic information includes the user's face image. It is preferable that the registration reception unit 22 is configured such that the insurance company confirms that the facial image is the user's own face image, and the facial image is registered upon approval. Note that the face image may be acquired as an image whose identity has been verified via the medical institution system 4.
  • step S12 if the user's face authentication is successful (YES in S22), the acquisition unit 23 permits acquisition of the subsequent medical examination data.
  • the success or failure of face authentication is determined based on the face image captured by the imaging unit 306 of the user terminal 3 and the face image registered as basic information. If the face authentication fails (NO in S22), the acquisition unit 23 does not permit acquisition of the medical examination data and completes the process.
  • the underwriting support device 2 may be configured to transmit a face authentication process request including a face image to an external face authentication server connected to the communication network NW, and obtain the result of the face authentication process.
  • step S13 the acquisition unit 23 acquires medical examination image data via the user terminal 3.
  • FIG. 4A shows a screen display example of the first medical examination data input screen W1 on the user terminal 3.
  • the first medical examination data input screen W1 includes a medical examination image data selection section W11 and a send button W12.
  • the medical examination image data selection unit W11 receives a selection of one or more image data obtained by capturing a medical examination form stored in the storage unit 302 of the user terminal 3, and uploads the image data.
  • the send button W12 is pressed, the selected medical examination image data is sent to the underwriting support device 2.
  • step S14 the acquisition unit 23 performs optical character authentication processing on each medical examination item included in the acquired medical examination image data, and acquires it as first medical examination data. If the optical character recognition process fails, the underwriting support device 2 transmits a request to the user terminal 3 to upload the medical examination image data again. Further, if the first medical examination data does not include the necessary items, the underwriting support device 2 transmits a request to the user terminal 3 to re-upload the medical examination image data including the necessary items.
  • Necessary items include the name of the medical institution that conducted the medical examination, the name of the doctor in charge, the overall assessment, and test items set by the insurance company. Further, the necessary items may include the designation of the health checkup recipient, and may be checked against the user name of the basic information.
  • step S15 the acquisition unit 23 acquires the second medical examination data via the second medical examination data input screen W2 of the user terminal 3.
  • FIG. 4(b) shows a screen display example of the second medical examination data input screen W2.
  • the second medical examination data input screen W2 includes a medical examination data input section W21 and a send button W22.
  • the medical examination data input unit W21 accepts input of each necessary item of the medical examination data as text data.
  • the user inputs each item while referring to the health checkup form sent in step S13.
  • the send button W22 is pressed, the value of each input item is sent to the underwriting support device 2.
  • steps S12 to S15 are not limited.
  • the face authentication in step S12 may be executed before the subsequent application data is processed, and may be executed in parallel when the second medical examination data is input in step S15, for example.
  • the underwriting support device 2 completes acquiring the first health checkup data and the second health checkup data (YES in step S16), it executes the subsequent detection process. If either the first medical examination data or the second medical examination data is insufficient, the underwriting support device 2 generates a display on the user terminal 3 requesting the necessary data.
  • the face authentication in step S12 may be performed multiple times at arbitrary timing or at random timing when various information is input via the user terminal 3 in steps S11 to S15.
  • the face authentication preferably includes authentication regarding at least one facial movement (blinking, tilting the face, changing facial expression, etc.).
  • step S17 the detection unit 24 executes a detection process to detect abnormal values included in the first health checkup data and the second health checkup data. Details of the detection process will be described later with reference to FIG.
  • step S18 the application data generation unit 25 generates application data based on the second medical examination data.
  • the application data generation unit 25 reacquires the item for which the abnormal value was detected and generates application data.
  • the application data generation unit 25 generates application data based on the second medical examination data re-inputted via the user terminal 3 regarding the items for which abnormal values have been detected.
  • the application data is data in which abnormal values detected in the second health checkup data are corrected and is actually used in the assessment process.
  • step S19 the acquisition unit 23 acquires the clinical data of the user corresponding to the user ID from the medical institution system 4. Note that if there is no clinical data, step S19 may be omitted.
  • step S20 the assessment unit 26 inputs the application data and clinical data into the judgment model and outputs the assessment result.
  • the assessment result is notified to the vendor terminal 5.
  • the assessment results include results regarding whether or not insurance can be underwritten. Furthermore, the assessment results may include proposal information on the amount of money that can be underwritten and insurance products and plans that can be underwritten.
  • the assessment result may be output as a numerical value or level indicating the degree of underwriting, or the insurance company may propose insurance according to the numerical value or level.
  • the insurance company makes a final decision on whether or not to accept the underwriting, notifies the user terminal 3, and completes the process. Insurers can carry over existing operations such as external assessment systems and double-check application data when making a final decision on whether or not to underwrite the product.
  • FIG. 5 shows a detailed flowchart of the detection process in step S17 of FIG. 3.
  • the detection unit 24 detects abnormal values in the medical examination data in four steps S21 to S24.
  • the detection unit 24 assigns an abnormal value flag to an item in which an abnormal value has been detected.
  • the detection unit 24 detects items that are inconsistent between the first health checkup data and the second health checkup data as abnormal values.
  • the detection unit 24 refers to the detection setting information set for each item and detects an abnormal value in at least one of the first health checkup data and the second health checkup data.
  • the detection unit 24 refers to the abnormal value detection setting information regarding the combination of items and detects an abnormal value in at least one of the first health checkup data and the second health checkup data.
  • the detection unit 24 refers to detection setting information for determining abnormal values between items of the medical examination data that have a correlation with time-series changes, and detects abnormal values of the first medical examination data and the second medical examination data. At least one of the abnormal values is detected.
  • step S26 the detection unit 24 re-inputs the item detected as an abnormal value via the user terminal 3. and updates the second medical examination data.
  • the detection unit 24 may perform the detection process again using the four steps S21 to S24 on the updated second health checkup data.
  • an abnormal value flag may be added to the item and the process may proceed to S18 in FIG. 3.
  • the abnormal value flag is presented to the vendor terminal 5 and used to determine whether or not to accept the offer. Specifically, if an abnormal value flag is assigned only to a predetermined item in the first medical checkup data, the underwriting support device 2 performs confirmation regarding the item even if all the items in the second medical checkup data are normal. A notification requesting the same is sent to the vendor terminal 5.
  • the number of times the user terminal 3 is requested to re-enter when an abnormal value is detected can be set as appropriate.
  • re-input is requested only once, and thereby it is possible to almost eliminate defects due to typographical errors and the like.
  • Underwriting support system Underwriting support device 21
  • Setting reception unit 22
  • Registration reception unit 23
  • Acquisition unit 24
  • Detection unit 25
  • Application data generation unit 26
  • User terminal 4 Medical institution system

Abstract

The present invention can provide technology for assisting in determining whether insurance can be accepted in a state where a typo or false value included in the insurance application has been highly accurately checked. The present invention comprises: an acquisition unit which acquires first health diagnosis data generated on the basis of health diagnosis image data of an insurance applicant and second health diagnosis data input by the insurance applicant; a detection unit which detects an outlier including an item in which the first health diagnosis data does not match with second health diagnosis data; and an application data generation unit which reacquires the item in which the outlier is detected and generates the application data.

Description

引受支援システム、引受支援方法、引受支援プログラムUnderwriting support system, underwriting support method, underwriting support program
　本発明は、引受支援システム、引受支援方法および、引受支援プログラムに関する。 The present invention relates to an underwriting support system, an underwriting support method, and an underwriting support program.
　従来、保険契約時の引受査定業務において、保険業者は、被保険者の年齢、職業、更には健康状態をチェックして保険引受が可能であるか否か、また、適切な保険額はどの程度であるかを判断する必要があった。特に、健康状態のチェックでは、被保険者の健康診断書のデータ化やデータ真正性の確認、それらデータに基づく保険査定などの業務にコストを要することが課題であった。 Traditionally, in the underwriting process when signing an insurance contract, the insurance company checks the age, occupation, and health condition of the insured person to determine whether they can underwrite the insurance and what is the appropriate insurance amount. It was necessary to determine whether In particular, when it comes to checking health conditions, it has been a problem that costs are required for operations such as converting the insured person's health certificate into data, confirming the authenticity of the data, and conducting insurance assessments based on such data.
　非特許文献１では、上述した課題に鑑みて、ハンディスキャナとイメージ伝送技術を活用し、健康診断書のデータ変換および入力を非定型ＯＣＲ（Optical Character Recognition）技術により自動化すること、被保険者の疾病等に関する告知方式の改善と、当該告知方式により取り込んだ告知の内容と査定ルールデータベースを照合することで引受可否および契約条件を自動引受査定エンジンが判定すること、を開示している。 In view of the above-mentioned problems, Non-Patent Document 1 proposes to automate the data conversion and input of health certificates using atypical OCR (Optical Character Recognition) technology by utilizing a handy scanner and image transmission technology, and to It discloses that it has improved the notification method for diseases, etc., and that an automatic underwriting appraisal engine determines whether or not the company can underwrite the product and contract conditions by comparing the contents of the notifications imported through the notification method with the appraisal rule database.
　また、特許文献１では、団体信用生命保険の申込手続をスムースに行うことのできるシステムであって、申込内容受付手段４６は、申込者または受付者の申込内容を記録し、申込情報送信手段４２は、申込内容および添付書類画像を、サーバ装置２に送信し、受付確認者端末装置６は、サーバ装置２にアクセスして、申込情報について不備があるか確認し、受付確認結果をサーバ装置２に送信し、サーバ装置２の受付確認結果受付手段２６は、受付確認結果を受信し、記録部３４に記録し、一時的アクセス設定手段２８は、受付確認結果が「不備」であった場合、申込者が記録部３４の申込情報に一時的にアクセス可能となるように、一時的アクセスキーを設定し、是正通知送信手段３０は、申込者に対し、当該一時的アクセスキーを伴って、是正通知を送信し、申込者は是正一時的アクセスキーによって記録部３４の申込情報にアクセスし、是正情報を送信する技術を開示している。 Further, Patent Document 1 discloses a system that can smoothly perform application procedures for group credit life insurance, and an application content receiving means 46 records the application content of an applicant or a receiver, and an application information transmitting means 42 sends the application details and attached document images to the server device 2, the reception confirmer terminal device 6 accesses the server device 2, checks whether there are any deficiencies in the application information, and sends the reception confirmation result to the server device 2. The reception confirmation result receiving means 26 of the server device 2 receives the reception confirmation result and records it in the recording unit 34, and the temporary access setting means 28 transmits the confirmation result if the reception confirmation result is "incomplete". A temporary access key is set so that the applicant can temporarily access the application information in the recording unit 34, and the correction notification sending means 30 notifies the applicant of the correction with the temporary access key. A technique is disclosed in which the notification is sent, the applicant accesses the application information in the recorder 34 by means of a remediation temporary access key, and the remediation information is transmitted.
特開２０１７－１６７８０４号公報Japanese Patent Application Publication No. 2017-167804
　非特許文献１は、健康診断書のＯＣＲ技術により自動的にデータ化することができたものの、健康診断書のＯＣＲ技術によるデータ化は、健康診断書のフォーマットが多様であることやＯＣＲ精度の観点から誤った数値として検出される点に課題があった。 Non-Patent Document 1 states that although it was possible to automatically convert health certificates into data using OCR technology, converting health certificates into data using OCR technology is difficult due to the diverse formats of health certificates and the lack of OCR accuracy. The problem was that it was detected as an incorrect value.
　特許文献１は、申込情報についての不備を確認する業務が自動化されていない点に改善の余地があった。 Patent Document 1 has room for improvement in that the task of checking for deficiencies in application information is not automated.
　本発明は、上述した課題に鑑みてなされたものであって、保険申込時に含まれる誤記や虚偽の値を、自動査定業務またはその前段階において精度よくチェックすることで業務効率を改善でき、保険引受の可否判断を支援する技術を提供することを解決すべき課題とする。 The present invention has been made in view of the above-mentioned problems, and it is possible to improve business efficiency by accurately checking for errors and false values included in insurance applications during automatic assessment operations or in the preceding stage. The issue to be solved is to provide technology that supports the decision to accept underwriting.
　上述したような課題を解決するために、本発明は、前記保険申請者の健診画像データに基づき生成される第１健診データと、保険申請者により入力される第２健診データと、を取得する取得部と、前記第１健診データと前記第２健診データの項目が不一致であることを含む異常値を検出する検出部と、前記異常値が検出された項目を再取得し、申請データを生成する申請データ生成部と、を備える。 In order to solve the above problems, the present invention provides first medical examination data generated based on the medical examination image data of the insurance applicant, second medical examination data input by the insurance applicant, an acquisition unit that acquires an item, a detection unit that detects an abnormal value including a mismatch between items of the first health checkup data and the second health checkup data, and a detection unit that reacquires the item in which the abnormal value was detected. , an application data generation unit that generates application data.
　また、本発明は、前記保険申請者の健診画像データに基づき生成される第１健診データと、保険申請者により入力される第２健診データを取得するステップと、前記第１健診データと前記第２健診データの項目が不一致であることを含む異常値を検出するステップと、前記異常値が検出された項目を再取得し、申請データを生成するステップと、をコンピュータが実行する。 The present invention also provides a step of acquiring first medical examination data generated based on the medical examination image data of the insurance applicant and second medical examination data input by the insurance applicant; A computer executes the steps of detecting an abnormal value including a mismatch between the item of the data and the second health checkup data, and reacquiring the item for which the abnormal value was detected and generating application data. do.
　また、本発明は、前記保険申請者の健診画像データに基づき生成される第１健診データと、保険申請者により入力される第２健診データを取得する取得部と、前記第１健診データと前記第２健診データの項目が不一致であることを含む異常値を検出する検出部と、前記異常値が検出された項目を再取得し、申請データを生成する申請データ生成部と、としてコンピュータを機能させる。 The present invention also provides an acquisition unit that acquires first medical examination data generated based on medical examination image data of the insurance applicant and second medical examination data input by the insurance applicant; a detection unit that detects abnormal values including mismatch between items of the medical examination data and the second medical examination data; and an application data generation unit that reacquires the items for which the abnormal values have been detected and generates application data. , to make the computer function as.
　このような構成とすることで、２つの健診データの項目が一致しない箇所について、正しい値に修正された申請データを生成することができる。 With such a configuration, it is possible to generate application data in which the values are corrected for areas where the items of the two medical checkup data do not match.
　本発明の好ましい形態では、前記申請データ生成部は、前記異常値が検出された項目に関して前記保険申請者により再入力された第２健診データに基づき前記申請データを生成する。
　このような構成とすることで、健診画像データに基づき生成される第１健診データの項目に不備がある場合であっても、当該項目について正しい値に修正された申請データを生成することができる。 In a preferred embodiment of the present invention, the application data generation unit generates the application data based on second medical examination data re-inputted by the insurance applicant regarding the item in which the abnormal value was detected.
With this configuration, even if there is a defect in an item of the first medical examination data generated based on the medical examination image data, application data with correct values for the item can be generated. Can be done.
　本発明の好ましい形態では、前記保険申請者の顔画像を登録する登録受付部を備え、前記申請データ生成部は、撮像された前記保険申請者の顔画像を認証することで、前記第１健診データまたは前記第２健診データに基づく前記申請データを生成する。
　このような構成とすることで、保険申請者が別人に成りすまして保険申請することを防止できる。 In a preferred embodiment of the present invention, the invention further includes a registration reception unit that registers a facial image of the insurance applicant, and the application data generation unit authenticates the captured facial image of the insurance applicant. The application data is generated based on the medical examination data or the second medical examination data.
With such a configuration, it is possible to prevent an insurance applicant from impersonating another person and applying for insurance.
　本発明の好ましい形態では、健診データの項目別に正常値または異常値の範囲を設定した検出設定情報を受け付ける設定受付部を備え、前記検出部は、前記検出設定情報に基づいて前記第１健診データおよび前記第２健診データの少なくとも一方の項目の異常値を検出する。
　本発明の好ましい形態では、前記設定受付部は、健診データの複数の項目の値の組合せについて、異常値を設定した検出設定情報を更に受け付ける。
　前記設定受付部は、利用者の過去の健診データと現在の健診データに含まれる時系列変化に相関関係を有する項目に関する異常値を設定した検出設定情報を更に受け付ける。
　このような構成とすることで、健診データに含まれる値の異常値を検出することができる。 In a preferred embodiment of the present invention, the present invention further includes a setting reception unit that receives detection setting information in which a range of normal values or abnormal values is set for each item of health checkup data, and the detection unit is configured to detect the first health check based on the detection setting information. An abnormal value in at least one item of the medical examination data and the second medical examination data is detected.
In a preferred embodiment of the present invention, the setting reception unit further receives detection setting information in which an abnormal value is set for a combination of values of a plurality of items of medical examination data.
The setting reception unit further receives detection setting information in which abnormal values are set for items that have a correlation with time-series changes included in the user's past health checkup data and current health checkup data.
With such a configuration, abnormal values included in the medical examination data can be detected.
　本発明の好ましい形態では、前記保険申請者の臨床データおよび健診データを入力データとし、保険査定結果を出力データとして機械学習された判定モデルに対して、前記申請データおよび臨床データを入力し、保険査定結果を出力する査定部を備える。
　このような構成とすることで、健診データに加えて臨床データを加味することで、保険査定業務を好適に自動化することができる。 In a preferred embodiment of the present invention, the application data and clinical data are input to a machine-learned judgment model using the insurance applicant's clinical data and health checkup data as input data and the insurance assessment result as output data, It is equipped with an assessment department that outputs insurance assessment results.
With such a configuration, insurance assessment operations can be suitably automated by taking clinical data into consideration in addition to medical examination data.
　本発明によれば、保険申込時に含まれる誤記や虚偽の値を精度よくチェックしたうえで、保険引受の可否判断を支援する技術を提供することができる。 According to the present invention, it is possible to provide a technology that accurately checks for errors or false values included in an insurance application, and then supports a decision on whether or not to accept insurance.
本実施形態にかかる引受支援システムのシステム構成図を示す。1 shows a system configuration diagram of an underwriting support system according to the present embodiment. 本実施形態にかかる各装置のハードウェア構成図を示す。A hardware configuration diagram of each device according to the present embodiment is shown. 本実施形態にかかる引受処理のフローチャートを示す。A flowchart of underwriting processing according to the present embodiment is shown. 本実施形態にかかる入力画面の画面表示例を示す。A screen display example of an input screen according to the present embodiment is shown. 本実施形態にかかる検出処理のフローチャートを示す。A flowchart of detection processing according to the present embodiment is shown.
　以下、図面を用いて、本発明の引受支援システムについて説明する。なお、以下に示す実施形態は本発明の一例であり、本発明を以下の実施形態に限定するものではなく、様々な構成を採用することもできる。 Hereinafter, the underwriting support system of the present invention will be explained using the drawings. Note that the embodiment shown below is an example of the present invention, and the present invention is not limited to the following embodiment, and various configurations can be adopted.
　本実施形態では引受支援システムの構成、動作等について説明するが、同様の構成の方法、装置、コンピュータのプログラムおよび当該プログラムを格納したプログラム記録媒体なども、同様の作用効果を奏することができる。以下で説明する本実施形態にかかる一連の処理は、コンピュータで実行可能なプログラムとして提供され、ＣＤ－ＲＯＭやフレキシブルディスクなどの非一過性コンピュータ可読記録媒体、更には通信回線を経て提供可能である。 In this embodiment, the configuration, operation, etc. of the underwriting support system will be described, but methods, devices, computer programs, and program recording media storing the programs with similar configurations can also produce similar effects. A series of processes according to the present embodiment described below are provided as a computer-executable program, and can be provided via a non-transitory computer-readable recording medium such as a CD-ROM or a flexible disk, or even via a communication line. be.
　引受支援システムの各機能構成部と、引受支援方法の各ステップと、は同様の作用効果を実現する。引受支援システムを構成するコンピュータは、ＣＰＵ（Central Processing Unit）などの演算装置および記憶装置を有する。当該コンピュータは、記憶装置に格納される引受支援プログラムを、演算装置により実行することで、当該コンピュータを引受支援装置として機能させることができる。 Each functional component of the underwriting support system and each step of the underwriting support method achieve similar effects. The computer that constitutes the underwriting support system has a calculation device such as a CPU (Central Processing Unit) and a storage device. The computer can function as an underwriting support device by using the arithmetic device to execute an underwriting support program stored in the storage device.
　本明細書では、生命保険や健康保険、医療保険などの保険に関する引受業務を支援する引受支援システムの実施例を説明するが、引受対象は、例えば、ローン契約、カード契約など健康情報に基づき査定判断がなされるものであれば、特に保険に限定されるものではない。また、その他の引受業務の対象として、免許証等の更新業務、健康診断結果を必要とする入出国時業務、企業等における健康診断結果の真贋確認業務などが含まれてもよい。 This specification describes an example of an underwriting support system that supports underwriting operations related to insurance such as life insurance, health insurance, and medical insurance. As long as a judgment is made, it is not particularly limited to insurance. In addition, other underwriting services may include renewal of driver's licenses, etc., immigration/departure services that require medical examination results, and verification of the authenticity of medical examination results at companies, etc.
　図１は、引受支援システム１のブロック図を示す。引受支援システム１は、引受支援装置２と、利用者端末３と、医療機関システム４と、保険業者により操作される業者端末５と、データベースとしての記憶部ＤＢと、を備え、これら構成部は、通信ネットワークＮＷを介して通信可能に接続されている。利用者端末３は、保険申請を行う被保険者により操作される端末装置であって、複数存在してもよい。また、記憶部ＤＢは、引受支援装置２とデータ通信可能に構成されていればよく、引受支援装置２の内部に搭載される態様であってもよい。 FIG. 1 shows a block diagram of the underwriting support system 1. The underwriting support system 1 includes an underwriting support device 2, a user terminal 3, a medical institution system 4, a business terminal 5 operated by an insurance company, and a storage unit DB as a database, and these components are , are communicably connected via a communication network NW. The user terminal 3 is a terminal device operated by an insured person who applies for insurance, and there may be a plurality of user terminals. Further, the storage unit DB only needs to be configured to be able to communicate data with the underwriting support device 2, and may be installed inside the underwriting support device 2.
　引受支援装置２は、機能構成要素として、異常値検出の基準や判定モデルを含む各種情報の設定入力を受け付ける設定受付部２１と、利用者情報の登録を受け付ける登録受付部２２と、健診データを取得する取得部２３と、健診データに含まれる異常値を検出する検出部２４と、健診データに基づき保険に関する申請データを生成する申請データ生成部２５と、申請データに基づき保険に関する査定処理を実行する査定部２６と、を備える。引受支援装置２の一部の機能構成は、利用者端末３に搭載される構成であってもよく、引受支援装置２および利用者端末３の全体で上述した各種機能構成が実現されればよい。 The underwriting support device 2 includes, as functional components, a setting reception section 21 that receives setting input of various information including criteria for abnormal value detection and judgment models, a registration reception section 22 that receives registration of user information, and medical examination data. an acquisition unit 23 that acquires the data, a detection unit 24 that detects abnormal values included in the medical examination data, an application data generation unit 25 that generates insurance-related application data based on the medical examination data, and an insurance-related assessment based on the application data. The evaluation unit 26 includes an assessment unit 26 that executes processing. A part of the functional configuration of the underwriting support device 2 may be a configuration installed in the user terminal 3, and the various functional configurations described above may be realized by the underwriting support device 2 and the user terminal 3 as a whole. .
　医療機関システム４は、病院などの医療機関に設置される。医療機関システム４は、保険申請者を含む利用者の健康診断を実施し、健診データを利用者の識別子に対応付けて医療機関データベースに格納している。また、医療機関システム４は、利用者の医療機関への既往歴、現病歴などを含む臨床データを利用者の識別子に対応付けて医療機関データベースに格納している。医療機関システム４は、引受支援装置２より利用者の各種データに関する要求に応じて、要求されたデータを提供する。 The medical institution system 4 is installed in a medical institution such as a hospital. The medical institution system 4 carries out health examinations of users including insurance applicants, and stores the medical examination data in a medical institution database in association with user identifiers. Furthermore, the medical institution system 4 stores clinical data including the user's medical history at medical institutions, current medical history, etc. in the medical institution database in association with the user's identifier. The medical institution system 4 provides the requested data in response to the user's request regarding various data from the underwriting support device 2.
　図２（ａ）は、引受支援装置２のハードウェア構成図を示す。引受支援装置２は、ＣＰＵなどによる制御部２０１と、ＲＡＭ（Random Access Memory）などの主記憶装置、ＨＤＤ（Hard Disk Drive）やＳＳＤ（Solid State Drive）、フラッシュメモリなどの補助記憶装置による記憶部２０２と、通信ネットワークＮＷとの通信制御を行うための通信部２０３などとを備える。引受支援装置２の記憶部２０２は、ＯＳ（Operating System
）と、ＯＳと協働して各種機能構成要素を実現する引受支援プログラムと、を格納し、制御部２０１は、当該引受支援プログラムを実行することで各種機能構成を実現することができる。本実施形態において、引受支援装置２は、サーバ装置として構成される。なお、サーバ装置は、クラウド型、オンプレミス型などの態様を適宜採用できる。 FIG. 2(a) shows a hardware configuration diagram of the underwriting support device 2. As shown in FIG. The underwriting support device 2 includes a control unit 201 including a CPU, and a storage unit including a main storage device such as a RAM (Random Access Memory), an auxiliary storage device such as an HDD (Hard Disk Drive), an SSD (Solid State Drive), and a flash memory. 202, a communication unit 203 for controlling communication with the communication network NW, and the like. The storage unit 202 of the underwriting support device 2 is an OS (Operating System).
) and an underwriting support program that implements various functional components in cooperation with the OS, and the control unit 201 can implement various functional configurations by executing the underwriting support program. In this embodiment, the underwriting support device 2 is configured as a server device. Note that the server device can be of a cloud type, an on-premise type, or the like as appropriate.
　図２（ｂ）は、利用者端末３のハードウェア構成図を示す。利用者端末３は、ＣＰＵなどによる制御部３０１と、主記憶装置や補助記憶装置などによる記憶部３０２と、通信部３０３と、タッチパネル、マウス、キーボードなどによる入力部３０４と、ディスプレイなどによる表示部３０５と、カメラなどによる撮像部３０６と、を備える。なお、業者端末５も同様のハードウェア構成を備えるものとする。 FIG. 2(b) shows a hardware configuration diagram of the user terminal 3. The user terminal 3 includes a control unit 301 such as a CPU, a storage unit 302 such as a main memory or auxiliary storage, a communication unit 303, an input unit 304 such as a touch panel, a mouse, and a keyboard, and a display unit such as a display. 305, and an imaging unit 306 such as a camera. Note that the vendor terminal 5 is also assumed to have a similar hardware configuration.
　本実施形態において、引受支援システム１の検出部２４は、保険申請する利用者の操作する利用者端末３を介して送信される健診データの項目について、異常値を検出する。健診データは、利用者が受診した健康診断の健康診断票を撮像した健診画像データに基づき生成される第１健診データと、利用者が当該健康診断票を参照し、入力画面を介して入力する第２健診データと、を含む。以下の説明において、２つの健診データを区別しない場合、単に健診データと記載する。 In the present embodiment, the detection unit 24 of the underwriting support system 1 detects abnormal values for items of medical examination data transmitted via the user terminal 3 operated by the user applying for insurance. The health checkup data consists of the first health checkup data generated based on the health checkup image data obtained by capturing the health checkup form of the health checkup that the user received, and the data that the user refers to the health checkup form and inputs via the input screen. and second medical examination data to be input. In the following description, when two pieces of medical examination data are not distinguished, they are simply referred to as medical examination data.
　検出部２４は、設定受付部２１により設定される検出設定情報に基づき健診データの項目の異常値を検出する。検出設定情報は、検出の基準となる異常値または正常値の範囲に関する設定や、健診データを入力し異常値の検出結果を出力する検出モデルを含む。また、異常値とは、利用者の健康診断票を撮像した健診画像データに基づいて取得される第１健診データの項目と、利用者端末３を介して入力される第２健診データの項目が一致しない値や、それぞれの健診データの項目間の相関から不備と判断される外れ値などを含む。 The detection unit 24 detects abnormal values in the items of the medical examination data based on the detection setting information set by the setting reception unit 21. The detection setting information includes settings regarding the range of abnormal values or normal values that serve as detection standards, and a detection model that inputs medical examination data and outputs abnormal value detection results. In addition, abnormal values are items of first health checkup data acquired based on health checkup image data obtained by capturing a user's health checkup form, and second health checkup data input via the user terminal 3. This includes values where the items do not match, and outliers that are judged to be inadequate based on the correlation between the items of each health checkup data.
　設定受付部２１は、健診データの項目別の異常値と、項目間の組合せの異常値と、項目の時系列変化による相関の異常値と、を検出するための検出設定情報を受け付け、記憶部ＤＢに格納する。検出設定情報は、例えば、保険業者により操作される業者端末５や、医療機関システム４を介して入力される。 The setting reception unit 21 receives and stores detection setting information for detecting abnormal values for each item of medical examination data, abnormal values for combinations of items, and abnormal values for correlations due to time-series changes in items. Store in the section DB. The detection setting information is inputted, for example, via the provider terminal 5 operated by the insurance company or the medical institution system 4.
　設定受付部２１は、健診データの項目別の正常値または異常値の範囲に関する検出設定情報を受け付ける。例えば、ＢＭＩは、下限値：５、上限値：１００の正常値の範囲を設定され、検出部２４は、健診データの項目において、正常値から外れた値を異常値として検出する。正常値または異常値の範囲は、上述したものに限定されず、項目別に適宜設定可能に構成されている。 The setting reception unit 21 receives detection setting information regarding the range of normal values or abnormal values for each item of medical examination data. For example, the BMI is set within a normal value range of 5 as a lower limit and 100 as an upper limit, and the detection unit 24 detects a value that deviates from the normal value as an abnormal value in an item of medical examination data. The range of normal values or abnormal values is not limited to those described above, and can be set appropriately for each item.
　設定受付部２１は、健診データの複数の項目の組合せについて、異常値と判定される組合せを検出設定情報として受け付ける。項目の組合せは、例えば、性別、年齢などの利用者の属性と、１つ以上の健診データの項目が含まれる。異常値と判定される組合せの例（組合せＡ～Ｃ）を以下に示す。
　組合せＡ：男性、５０歳、ＢＭＩ：３０、腹囲：６０ｃｍ
　組合せＢ：男性、６０歳、ＢＭＩ：３５、収縮期血圧：８０
　組合せＣ：女性、２５歳、ヘモグロビン：１５．０、血清鉄：１０ The setting reception unit 21 receives, as detection setting information, a combination of a plurality of items of medical examination data that is determined to be an abnormal value. The combination of items includes, for example, user attributes such as gender and age, and one or more items of medical examination data. Examples of combinations (combinations A to C) that are determined to be abnormal values are shown below.
Combination A: Male, 50 years old, BMI: 30, Waist circumference: 60 cm
Combination B: Male, 60 years old, BMI: 35, systolic blood pressure: 80
Combination C: Female, 25 years old, hemoglobin: 15.0, serum iron: 10
　組合せＡは、ＢＭＩの値が肥満傾向を示すのに対して、腹囲の値が男性の標準以下であることから不備の可能性があると判定される。組合せＢは、ＢＭＩの値が肥満傾向を示すのに対して、収縮期血圧の値が低血圧を示すことから不備の可能性があると判定される。組合せＣは、ヘモグロビンの値が女性の基準値を示すのに対して、血清鉄の値が貧血の傾向を示すことから不備の可能性があると判定される。これらの組合せは一例であり、検査項目間に相関が認められる値の外れ値が適宜設定可能に構成されている。 Combination A is determined to be potentially defective because the BMI value indicates a tendency towards obesity, but the waist circumference value is below the male standard. Combination B is determined to be potentially defective because the BMI value indicates a tendency toward obesity, whereas the systolic blood pressure value indicates low blood pressure. Combination C is determined to be potentially defective because the hemoglobin value shows the standard value for women, but the serum iron value shows a tendency toward anemia. These combinations are just examples, and the configuration is such that outliers of values where correlation is recognized between inspection items can be set as appropriate.
　異常値と判定される組合せの検出設定情報は、機械学習された検出モデルであってもよい。検出モデルは、健診データの複数の項目を入力とし、当該項目の組合せが正常値または異常値であるかを出力とするデータセットを用いて機械学習される。機械学習により生成された検出モデルは、健診データの複数の項目の入力を受け付け、それらの組合せが正常値であるか異常値であるかを出力することができる。 The detection setting information for combinations that are determined to be abnormal values may be a machine-learned detection model. The detection model is machine-learned using a dataset that takes multiple items of health check data as input and outputs whether the combination of the items is a normal value or an abnormal value. The detection model generated by machine learning can accept input of multiple items of health check data and output whether the combination of those items is a normal value or an abnormal value.
　設定受付部２１は、利用者の過去の健診データと現在の健診データの間における時系列変化による項目の異常値を設定した検出設定情報を受け付ける。本実施形態では、健診データの２つ以上の項目に対して、正の相関または負の相関の何れかが設定される。検出部２４は、同一の利用者の以前の健診データと現在の健診データの第１の項目の変化に対して、正または負の相関を設定された第２の項目の変化を検出する。検出部２４は、第２の項目の変化が設定された相関に反する場合、第１の項目および／または第２の項目が異常値であることを検出する。例えば、ＢＭＩと腹囲は、正の相関として設定される。以前の健診データがＢＭＩ：２５、腹囲：８５ｃｍであり、現在の健診データがＢＭＩ：３０、腹囲：７５ｃｍである場合、ＢＭＩの増加に対して腹囲が減少し、正の相関に反するため、ＢＭＩおよび／または腹囲の値が異常値であると検出される。なお、設定された相関に反する場合の許容値が更に設定されてもよい。 The setting reception unit 21 receives detection setting information that sets abnormal values of items due to time-series changes between the user's past health checkup data and current health checkup data. In this embodiment, either positive correlation or negative correlation is set for two or more items of medical checkup data. The detection unit 24 detects a change in a second item that is set to have a positive or negative correlation with a change in the first item of the same user's previous health checkup data and current health checkup data. . The detection unit 24 detects that the first item and/or the second item is an abnormal value when the change in the second item is contrary to the set correlation. For example, BMI and waist circumference are set as having a positive correlation. If the previous health checkup data is BMI: 25 and waist circumference: 85cm, and the current health checkup data is BMI: 30 and waist circumference: 75cm, the waist circumference decreases as BMI increases, which contradicts the positive correlation. , BMI and/or waist circumference values are detected to be abnormal values. Note that a tolerance value may be further set in the case where the set correlation is violated.
　時系列変化による異常値の検出設定情報は、機械学習された検出モデルとして記憶部ＤＢに格納されてもよい。検出モデルは、以前の健診データの各項目と、現在の健診データの各項目を入力とし、それら項目の時系列変化が正常値または異常値であるかを出力とするデータセットを用いて機械学習される。機械学習により生成された判定モデルは、以前および現在の健診データの入力を受け付け、それらの項目の時系列変化が正常値であるか異常値であるかを出力することができる。 The abnormal value detection setting information based on time-series changes may be stored in the storage unit DB as a machine-learned detection model. The detection model uses a dataset whose input is each item of previous health checkup data and each item of the current health checkup data, and whose output is whether the time-series changes in these items are normal or abnormal values. Machine learned. A determination model generated by machine learning can accept input of previous and current health checkup data and output whether the time-series changes in those items are normal values or abnormal values.
　設定受付部２１は、利用者の健診データと臨床データを入力とし、査定結果を出力する判定モデルの設定を受け付け、記憶部ＤＢに格納する。本実施形態において、判定モデルは、機械学習により生成されたモデルである。なお、機械学習処理は、引受支援システム１の外部の装置で実行され、設定受付部２１は、学習された判定モデルを受け付ける構成であってよい。 The settings reception unit 21 inputs the user's medical examination data and clinical data, receives settings for a determination model that outputs assessment results, and stores the settings in the storage unit DB. In this embodiment, the determination model is a model generated by machine learning. Note that the machine learning process may be executed by a device external to the underwriting support system 1, and the setting reception unit 21 may be configured to receive the learned determination model.
　臨床データは、既往歴、現病歴、薬処方歴、治療歴、合併症の有無、などを含む。臨床データは、医療機関システム４より取得される告知書に基づき生成されるデータであってもよい。査定結果は、引受の可否、引受可能な金額、提案する保険商品などを含む。 Clinical data includes past medical history, current medical history, drug prescription history, treatment history, presence or absence of complications, etc. The clinical data may be data generated based on a notification obtained from the medical institution system 4. The assessment results include whether or not the insurance can be underwritten, the amount that can be underwritten, and proposed insurance products.
　判定モデルの機械学習に用いるデータセットは、例えば、健診データに含まれる高血圧の傾向を示す血圧の値と、臨床データに含まれる降圧剤の薬処方歴と、その他の既往歴や合併症なしと、を入力データとし、引受の条件を緩和する傾向の査定結果（例えば、引受可）を出力データとする。このデータセット例では、利用者の健診データから高血圧であると判断されるものの、臨床データから処方薬により血圧が正常にコントロールされ、他の疾患の既往歴や合併症もないことから、高血圧の重症化リスクが低いと判断され、引受可能と査定する傾向を機械学習させることができる。すなわち、判定モデルは、健診データと臨床データの組合せと、その組合せによる査定医の査定結果の判断の傾向を機械学習されたモデルとなる。 Data sets used for machine learning of the judgment model include, for example, blood pressure values indicating a tendency toward hypertension included in health checkup data, prescription history of antihypertensive drugs included in clinical data, and other medical history and no complications. and are input data, and an assessment result showing a tendency to relax underwriting conditions (for example, underwriting is acceptable) is set as output data. In this example dataset, it is determined that the user has high blood pressure based on the health checkup data, but the clinical data shows that the user's blood pressure is normally controlled with prescription medication, and there is no history of other diseases or complications. It is possible to use machine learning to determine the tendency for patients to be assessed as having a low risk of becoming seriously ill and therefore underwritten. That is, the judgment model is a model obtained by machine learning the combination of medical examination data and clinical data, and the tendency of judgment of the assessment result by the evaluator based on the combination.
　図３は、引受支援装置２における処理に関するフローチャートを示す。図３は、保険業者が保険引受に必要な各種情報の入力を利用者に案内し、利用者が利用者端末３を介して各種情報の入力する際の処理フローの例を示す。 FIG. 3 shows a flowchart regarding processing in the underwriting support device 2. FIG. 3 shows an example of a processing flow when an insurance company guides a user to input various information necessary for insurance underwriting, and the user inputs various information via the user terminal 3.
　ステップＳ１１において、登録受付部２２は、利用者端末３を介して利用者の利用者名、性別、年齢、住所、勤務先、年収など保険引受に必要な項目の入力を受け付け、基本情報として記憶部ＤＢに格納する。基本情報は、利用者に一意な利用者ＩＤを付与される。利用者ＩＤは、医療機関システム４で管理される当該利用者の患者番号やＩＤと対応付けされる。本実施形態において、基本情報は、業者端末５を介して入力されてもよい。 In step S11, the registration reception unit 22 receives input of items necessary for insurance underwriting such as the user's user name, gender, age, address, place of work, and annual income via the user terminal 3, and stores it as basic information. Store in the section DB. The basic information is given a unique user ID to the user. The user ID is associated with the patient number and ID of the user managed by the medical institution system 4. In this embodiment, the basic information may be input via the vendor terminal 5.
　基本情報は、利用者の顔画像を含む。登録受付部２２は、保険業者が利用者本人の顔画像であることを確認し、承認されることで登録される構成とすることが好ましい。なお、顔画像は、医療機関システム４を介して本人確認済の画像として取得されてもよい。 The basic information includes the user's face image. It is preferable that the registration reception unit 22 is configured such that the insurance company confirms that the facial image is the user's own face image, and the facial image is registered upon approval. Note that the face image may be acquired as an image whose identity has been verified via the medical institution system 4.
　ステップＳ１２において、取得部２３は、利用者の顔認証に成功した場合（Ｓ２２でＹＥＳ）、続く健診データの取得を許可する。顔認証は、健診データの入力に進む際に、利用者端末３の撮像部３０６により撮像された顔画像と、基本情報として登録された顔画像と、に基づき成否を判定される。取得部２３は、顔認証に失敗した場合（Ｓ２２でＮＯ）、健診データの取得を許可せず、処理を完了する。なお、引受支援装置２は、通信ネットワークＮＷに接続された外部の顔認証サーバに顔画像を含む顔認証処理要求を送信し、その顔認証処理の結果を取得する構成であってもよい。 In step S12, if the user's face authentication is successful (YES in S22), the acquisition unit 23 permits acquisition of the subsequent medical examination data. When proceeding to input the medical examination data, the success or failure of face authentication is determined based on the face image captured by the imaging unit 306 of the user terminal 3 and the face image registered as basic information. If the face authentication fails (NO in S22), the acquisition unit 23 does not permit acquisition of the medical examination data and completes the process. Note that the underwriting support device 2 may be configured to transmit a face authentication process request including a face image to an external face authentication server connected to the communication network NW, and obtain the result of the face authentication process.
　ステップＳ１３において、取得部２３は、利用者端末３を介して健診画像データを取得する。図４（ａ）は、利用者端末３における第１健診データ入力画面Ｗ１の画面表示例を示す。第１健診データ入力画面Ｗ１は、健診画像データ選択部Ｗ１１と、送信ボタンＷ１２と、を備える。健診画像データ選択部Ｗ１１は、利用者端末３の記憶部３０２に格納される健康診断票を撮像した１つ以上の画像データの選択を受け付け、当該画像データをアップロードする。送信ボタンＷ１２は、押下されることで、選択された健診画像データを引受支援装置２に送信する。 In step S13, the acquisition unit 23 acquires medical examination image data via the user terminal 3. FIG. 4A shows a screen display example of the first medical examination data input screen W1 on the user terminal 3. The first medical examination data input screen W1 includes a medical examination image data selection section W11 and a send button W12. The medical examination image data selection unit W11 receives a selection of one or more image data obtained by capturing a medical examination form stored in the storage unit 302 of the user terminal 3, and uploads the image data. When the send button W12 is pressed, the selected medical examination image data is sent to the underwriting support device 2.
　ステップＳ１４において、取得部２３は、取得した健診画像データに含まれる各健診項目について光学文字認証処理し、第１健診データとして取得する。引受支援装置２は、光学文字認識処理に失敗した場合、利用者端末３に対して健診画像データを再度アップロードする要求を送信する。また、引受支援装置２は、第１健診データに必要項目が含まれない場合、利用者端末３に対して当該必要項目を含む健診画像データを再度アップロードする要求を送信する。必要項目は、健康診断を実施した医療機関名、担当医師名、総合判定および、保険業者が設定する検査項目を含む。また、必要項目は、健康診断受診者の指名を含み、基本情報の利用者名と照合されてもよい。 In step S14, the acquisition unit 23 performs optical character authentication processing on each medical examination item included in the acquired medical examination image data, and acquires it as first medical examination data. If the optical character recognition process fails, the underwriting support device 2 transmits a request to the user terminal 3 to upload the medical examination image data again. Further, if the first medical examination data does not include the necessary items, the underwriting support device 2 transmits a request to the user terminal 3 to re-upload the medical examination image data including the necessary items. Necessary items include the name of the medical institution that conducted the medical examination, the name of the doctor in charge, the overall assessment, and test items set by the insurance company. Further, the necessary items may include the designation of the health checkup recipient, and may be checked against the user name of the basic information.
　ステップＳ１５において、取得部２３は、利用者端末３の第２健診データ入力画面Ｗ２を介して第２健診データを取得する。図４（ｂ）は、第２健診データ入力画面Ｗ２の画面表示例を示す。第２健診データ入力画面Ｗ２は、健診データ入力部Ｗ２１と、送信ボタンＷ２２と、を備える。健診データ入力部Ｗ２１は、健診データの各必要項目についてテキストデータとして入力を受け付ける。このとき、利用者は、ステップＳ１３において送信した健康診断票を参照しながら、各項目を入力する。送信ボタンＷ２２は、押下されることで、入力された各項目の値を引受支援装置２に送信する。 In step S15, the acquisition unit 23 acquires the second medical examination data via the second medical examination data input screen W2 of the user terminal 3. FIG. 4(b) shows a screen display example of the second medical examination data input screen W2. The second medical examination data input screen W2 includes a medical examination data input section W21 and a send button W22. The medical examination data input unit W21 accepts input of each necessary item of the medical examination data as text data. At this time, the user inputs each item while referring to the health checkup form sent in step S13. When the send button W22 is pressed, the value of each input item is sent to the underwriting support device 2.
　ステップＳ１２～Ｓ１５は、処理の順序は限定されない。ステップＳ１２の顔認証は、後の申請データを処理するまでに実行されればよく、例えば、ステップＳ１５において第２健診データの入力時に並行して実行されてもよい。引受支援装置２は、第１健診データおよび第２健診データを取得完了すると（ステップＳ１６でＹＥＳ）、続く検出処理を実行する。第１健診データと第２健診データの何れかが不足する場合、引受支援装置２は、利用者端末３に必要なデータを要求する表示を生成する。 The order of steps S12 to S15 is not limited. The face authentication in step S12 may be executed before the subsequent application data is processed, and may be executed in parallel when the second medical examination data is input in step S15, for example. When the underwriting support device 2 completes acquiring the first health checkup data and the second health checkup data (YES in step S16), it executes the subsequent detection process. If either the first medical examination data or the second medical examination data is insufficient, the underwriting support device 2 generates a display on the user terminal 3 requesting the necessary data.
　ステップＳ１２における顔認証は、ステップＳ１１～Ｓ１５の利用者端末３を介した各種情報入力の際に、任意のタイミングで、または、ランダムなタイミングで複数回実行されてもよい。なお、顔認証は、少なくとも１回以上の顔の動作（瞬き、顔の傾け、表情変化など）に関する認証を含むことが好ましい。 The face authentication in step S12 may be performed multiple times at arbitrary timing or at random timing when various information is input via the user terminal 3 in steps S11 to S15. Note that the face authentication preferably includes authentication regarding at least one facial movement (blinking, tilting the face, changing facial expression, etc.).
　ステップＳ１７において、検出部２４は、第１健診データと第２健診データに含まれる異常値を検出する検出処理を実行する。検出処理の詳細は、後に図５を参酌しながら詳述する。 In step S17, the detection unit 24 executes a detection process to detect abnormal values included in the first health checkup data and the second health checkup data. Details of the detection process will be described later with reference to FIG.
　ステップＳ１８において、申請データ生成部２５は、第２健診データに基づいて、申請データを生成する。申請データ生成部２５は、検出処理により異常値が検出された場合、異常値が検出された項目を再取得し、申請データを生成する。ここで、申請データ生成部２５は、異常値が検出された項目に関して利用者端末３を介して再入力された第２健診データに基づき申請データを生成する。申請データは、第２健診データにおいて検出された異常値が修正され、査定処理で実際に用いられるデータを示す。 In step S18, the application data generation unit 25 generates application data based on the second medical examination data. When an abnormal value is detected by the detection process, the application data generation unit 25 reacquires the item for which the abnormal value was detected and generates application data. Here, the application data generation unit 25 generates application data based on the second medical examination data re-inputted via the user terminal 3 regarding the items for which abnormal values have been detected. The application data is data in which abnormal values detected in the second health checkup data are corrected and is actually used in the assessment process.
　ステップＳ１９において、取得部２３は、利用者ＩＤに対応する利用者の臨床データを医療機関システム４より取得する。なお、臨床データが存在しない場合、ステップＳ１９は省略されてもよい。 In step S19, the acquisition unit 23 acquires the clinical data of the user corresponding to the user ID from the medical institution system 4. Note that if there is no clinical data, step S19 may be omitted.
　ステップＳ２０において、査定部２６は、申請データおよび臨床データを、判定モデルに入力し、査定結果を出力する。査定結果は、業者端末５に通知される。 In step S20, the assessment unit 26 inputs the application data and clinical data into the judgment model and outputs the assessment result. The assessment result is notified to the vendor terminal 5.
　査定結果は、保険引受の可否に関する結果を含む。また、査定結果は、引受可能な金額や引受可能な保険商品やプランの提案情報を含んでもよい。査定結果は、引受可否の度合いを示す数値またはレベルとして出力されてもよく、それら数値やレベルに応じた保険の提案を保険業者が提案する態様であってもよい。保険業者は、最終的な引受の可否を決定し、利用者端末３に対して通知し、処理を完了する。保険業者は、最終的な引受の可否を決定するうえで、外部の査定システムなど既存業務への引継ぎや申請データのダブルチェックなどを実施することができる。 The assessment results include results regarding whether or not insurance can be underwritten. Furthermore, the assessment results may include proposal information on the amount of money that can be underwritten and insurance products and plans that can be underwritten. The assessment result may be output as a numerical value or level indicating the degree of underwriting, or the insurance company may propose insurance according to the numerical value or level. The insurance company makes a final decision on whether or not to accept the underwriting, notifies the user terminal 3, and completes the process. Insurers can carry over existing operations such as external assessment systems and double-check application data when making a final decision on whether or not to underwrite the product.
　図５は、図３のステップＳ１７における検出処理の詳細なフローチャートを示す。本実施形態において、検出部２４は、Ｓ２１～Ｓ２４の４ステップで健診データの異常値をそれぞれ検出する。検出部２４は、異常値が検出された項目に異常値フラグを付与する。 FIG. 5 shows a detailed flowchart of the detection process in step S17 of FIG. 3. In this embodiment, the detection unit 24 detects abnormal values in the medical examination data in four steps S21 to S24. The detection unit 24 assigns an abnormal value flag to an item in which an abnormal value has been detected.
　検出部２４は、ステップＳ２１において、第１健診データと第２健診データの項目について不一致となる項目を異常値として検出する。検出部２４は、ステップＳ２２において、項目別に設定された検出設定情報を参照し、第１健診データおよび第２健診データの少なくとも何れか一方の異常値を検出する。検出部２４は、ステップＳ２３において、項目間の組合せに関する異常値の検出設定情報を参照し、第１健診データおよび第２健診データの少なくとも何れか一方の異常値を検出する。検出部２４は、ステップＳ２４において、時系列変化に相関関係を有する健診データの項目間の異常値を判定するための検出設定情報を参照し、第１健診データおよび第２健診データの少なくとも何れか一方の異常値を検出する。 In step S21, the detection unit 24 detects items that are inconsistent between the first health checkup data and the second health checkup data as abnormal values. In step S22, the detection unit 24 refers to the detection setting information set for each item and detects an abnormal value in at least one of the first health checkup data and the second health checkup data. In step S23, the detection unit 24 refers to the abnormal value detection setting information regarding the combination of items and detects an abnormal value in at least one of the first health checkup data and the second health checkup data. In step S24, the detection unit 24 refers to detection setting information for determining abnormal values between items of the medical examination data that have a correlation with time-series changes, and detects abnormal values of the first medical examination data and the second medical examination data. At least one of the abnormal values is detected.
　検出部２４は、Ｓ２１～Ｓ２４の少なくとも何れか１つのステップで異常値を検出した場合（ステップＳ２５でＹＥＳ）、ステップＳ２６において、異常値として検出された項目について利用者端末３を介して再入力を受け付け、第２健診データを更新する。検出部２４は、更新された第２健診データについて、Ｓ２１～Ｓ２４の４ステップにより検出処理を再度実行してもよい。 When the detection unit 24 detects an abnormal value in at least one of steps S21 to S24 (YES in step S25), in step S26, the detection unit 24 re-inputs the item detected as an abnormal value via the user terminal 3. and updates the second medical examination data. The detection unit 24 may perform the detection process again using the four steps S21 to S24 on the updated second health checkup data.
　検出処理により２回目以降に異常値が検出された場合、当該項目に異常値フラグを付与し、図３のＳ１８に進んでもよい。異常値フラグは、業者端末５に対して提示され、引受可否の判断に利用される。具体的には、引受支援装置２は、第１健診データの所定の項目のみに異常値フラグが付与される場合、第２健診データの項目が全て正常であっても、当該項目に関する確認を要請する通知を業者端末５に対して送信する。 If an abnormal value is detected for the second time or later in the detection process, an abnormal value flag may be added to the item and the process may proceed to S18 in FIG. 3. The abnormal value flag is presented to the vendor terminal 5 and used to determine whether or not to accept the offer. Specifically, if an abnormal value flag is assigned only to a predetermined item in the first medical checkup data, the underwriting support device 2 performs confirmation regarding the item even if all the items in the second medical checkup data are normal. A notification requesting the same is sent to the vendor terminal 5.
　異常値が検出された場合に利用者端末３に対して再入力を要求する回数は、適宜設定可能である。本実施形態では、再入力の要求は１回であり、これにより誤記などによる不備は概ね解消することができる。 The number of times the user terminal 3 is requested to re-enter when an abnormal value is detected can be set as appropriate. In the present embodiment, re-input is requested only once, and thereby it is possible to almost eliminate defects due to typographical errors and the like.
　以上に示したように、本発明によれば、保険申込時に含まれる誤記や虚偽の値を自動的に検出し、それらの値が修正された申請データを生成できる。これにより、引受査定における人的チェックの業務コストを削減するとともに、査定医による専門的な保険査定業務を支援することができる。 As described above, according to the present invention, it is possible to automatically detect errors or false values included in an insurance application, and generate application data with those values corrected. As a result, it is possible to reduce the operational cost of human checks in underwriting appraisals, and to support specialized insurance appraisal duties by appraisers.
１　引受支援システム
２　引受支援装置
２１　設定受付部
２２　登録受付部
２３　取得部
２４　検出部
２５　申請データ生成部
２６　査定部
３　利用者端末
４　医療機関システム
  1 Underwriting support system 2 Underwriting support device 21 Setting reception unit 22 Registration reception unit 23 Acquisition unit 24 Detection unit 25 Application data generation unit 26 Assessment unit 3 User terminal 4 Medical institution system

Claims (8)

  1. 　申請者の健診画像データに基づき生成される第１健診データと、前記申請者により入力される第２健診データとを取得する取得部と、
    　取得された前記第１健診データおよび前記第２健診データにおける項目が不一致であることを含む、前記第１健診データおよび前記第２健診データの少なくとも一方の異常値を検出する検出部と、
    　前記異常値が検出されないときは、取得済の前記第２健診データに基づいて申請データを生成し、前記異常値が検出されたときは、取得済の前記第１健診データに拘わりなく、前記異常値が検出された項目に関して前記申請者により再入力された再取得の第２健診データに基づいて修正した申請データを生成する申請データ生成部と、
    　を備える引受支援システム。     an acquisition unit that acquires first medical examination data generated based on the applicant's medical examination image data and second medical examination data input by the applicant;
    A detection unit that detects an abnormal value in at least one of the first health checkup data and the second health checkup data, including items that do not match in the acquired first health checkup data and the second health checkup data. and,
    When the abnormal value is not detected, application data is generated based on the acquired second medical examination data, and when the abnormal value is detected, regardless of the acquired first medical examination data, an application data generation unit that generates modified application data based on re-acquired second medical examination data re-inputted by the applicant regarding the item in which the abnormal value was detected;
    An underwriting support system equipped with
  2. 　前記申請者の顔画像を登録する登録受付部を更に備え、
    　前記取得部は、撮像された前記申請者の顔画像を認証することで、前記第１健診データおよび前記第２健診データを取得する、請求項１に記載の引受支援システム。     further comprising a registration reception unit that registers a facial image of the applicant,
    The underwriting support system according to claim 1, wherein the acquisition unit acquires the first medical examination data and the second medical examination data by authenticating a captured facial image of the applicant.
  3. 　健診データの項目別に正常値または異常値の範囲を設定した検出設定情報を受け付ける設定受付部を更に備え、
    　前記検出部は、前記検出設定情報に基づいて前記第１健診データおよび前記第２健診データの少なくとも一方の項目の異常値を検出する、請求項１に記載の引受支援システム。     The apparatus further includes a setting reception unit that receives detection setting information in which a range of normal values or abnormal values is set for each item of medical examination data,
    The underwriting support system according to claim 1, wherein the detection unit detects an abnormal value in at least one item of the first health checkup data and the second health checkup data based on the detection setting information.
  4. 　前記設定受付部は、健診データの複数の項目の値の組合せについて、異常値を設定した検出設定情報を更に受け付ける、請求項３に記載の引受支援システム。 The underwriting support system according to claim 3, wherein the setting reception unit further receives detection setting information in which an abnormal value is set for a combination of values of a plurality of items of medical examination data.
  5. 　前記設定受付部は、前記申請者の過去の健診データと現在の健診データの間における時系列変化に相関関係を有する項目に関する異常値を設定した検出設定情報を更に受け付ける、請求項４に記載の引受支援システム。 5. The setting reception unit further receives detection setting information in which an abnormal value is set for an item that has a correlation in time-series changes between past health checkup data and current health checkup data of the applicant. Underwriting support system listed.
  6. 　前記申請者の臨床データおよび健診データを入力データとし、査定結果を出力データとして機械学習された判定モデルに対して、前記申請データおよび臨床データを入力し、査定結果を出力する査定部を更に備える、請求項１に記載の引受支援システム。 An assessment unit that inputs the application data and clinical data to a machine-learned judgment model using the applicant's clinical data and health checkup data as input data and the assessment result as output data, and outputs the assessment result. The underwriting support system according to claim 1, comprising:
  7. 　申請者の健診画像データに基づき生成される第１健診データと、前記申請者により入力される第２健診データとを取得するステップと、
    　取得された前記第１健診データおよび前記第２健診データにおける項目が不一致であることを含む、前記第１健診データおよび前記第２健診データの少なくとも一方の異常値を検出するステップと、
    　前記異常値が検出されないときは、取得済の前記第２健診データに基づいて申請データを生成し、前記異常値が検出されたときは、取得済の前記第１健診データに拘わりなく、前記異常値が検出された項目に関して前記申請者により再入力された再取得の第２健診データに基づいて修正した申請データを生成するステップと、
    　をコンピュータが実行する引受支援方法。     acquiring first medical examination data generated based on the applicant's medical examination image data and second medical examination data input by the applicant;
    detecting an abnormal value in at least one of the first health checkup data and the second health checkup data, including a mismatch between items in the acquired first health checkup data and the second health checkup data; ,
    When the abnormal value is not detected, application data is generated based on the acquired second medical examination data, and when the abnormal value is detected, regardless of the acquired first medical examination data, generating corrected application data based on re-acquired second medical examination data re-entered by the applicant regarding the item in which the abnormal value was detected;
    An underwriting support method performed by a computer.
  8. 　申請者の健診画像データに基づき生成される第１健診データと、前記申請者により入力される第２健診データとを取得する取得部と、
    　取得された前記第１健診データおよび前記第２健診データにおける項目が不一致であることを含む、前記第１健診データおよび前記第２健診データの少なくとも一方の異常値を検出する検出部と、
    　前記異常値が検出されないときは、取得済の前記第２健診データに基づいて申請データを生成し、前記異常値が検出されたときは、取得済の前記第１健診データに拘わりなく、前記異常値が検出された項目に関して前記申請者により再入力された再取得の第２健診データに基づいて修正した申請データを生成する申請データ生成部と、
    　としてコンピュータを機能させる引受支援プログラム。
          an acquisition unit that acquires first medical examination data generated based on the applicant's medical examination image data and second medical examination data input by the applicant;
    A detection unit that detects an abnormal value in at least one of the first health checkup data and the second health checkup data, including items that do not match in the acquired first health checkup data and the second health checkup data. and,
    When the abnormal value is not detected, application data is generated based on the acquired second medical examination data, and when the abnormal value is detected, regardless of the acquired first medical examination data, an application data generation unit that generates modified application data based on re-acquired second medical examination data re-inputted by the applicant regarding the item in which the abnormal value was detected;
    An underwriting support program that allows a computer to function as a computer.
PCT/JP2023/030144 2022-09-13 2023-08-22 Acceptance assistance system, acceptance assistance method, and acceptance assistance program WO2024057839A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2022145617A JP7300123B1 (en) 2022-09-13 2022-09-13 Underwriting support system, underwriting support method, underwriting support program
JP2022-145617 2022-09-13

Publications (1)

Publication Number Publication Date
WO2024057839A1 true WO2024057839A1 (en) 2024-03-21

Family

ID=86900552

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2023/030144 WO2024057839A1 (en) 2022-09-13 2023-08-22 Acceptance assistance system, acceptance assistance method, and acceptance assistance program

Country Status (2)

Country Link
JP (1) JP7300123B1 (en)
WO (1) WO2024057839A1 (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2003331127A (en) * 2002-05-10 2003-11-21 Tokio Marine Life Insurance Co Ltd System, device, and method for assisting life insurance sale
WO2015145727A1 (en) * 2014-03-28 2015-10-01 株式会社 日立製作所 Personal information acquisition system and personal information acquisition method
JP2017167804A (en) * 2016-03-16 2017-09-21 ニッセイ情報テクノロジー株式会社 Application reception system
JP2021060629A (en) * 2019-10-02 2021-04-15 大日本印刷株式会社 Question generation apparatus, question generation method, program for question generation apparatus, and document management system
JP7044929B1 (en) * 2021-06-14 2022-03-30 チューリッヒ生命保険株式会社 Programs, information processing methods and information processing equipment

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2003331127A (en) * 2002-05-10 2003-11-21 Tokio Marine Life Insurance Co Ltd System, device, and method for assisting life insurance sale
WO2015145727A1 (en) * 2014-03-28 2015-10-01 株式会社 日立製作所 Personal information acquisition system and personal information acquisition method
JP2017167804A (en) * 2016-03-16 2017-09-21 ニッセイ情報テクノロジー株式会社 Application reception system
JP2021060629A (en) * 2019-10-02 2021-04-15 大日本印刷株式会社 Question generation apparatus, question generation method, program for question generation apparatus, and document management system
JP7044929B1 (en) * 2021-06-14 2022-03-30 チューリッヒ生命保険株式会社 Programs, information processing methods and information processing equipment

Also Published As

Publication number Publication date
JP7300123B1 (en) 2023-06-29
JP2024040944A (en) 2024-03-26

Similar Documents

Publication Publication Date Title
JP6925439B2 (en) Methods and devices for acquiring and recording tracking information on the blockchain
US10467468B2 (en) System and method for identity proofing and knowledge based authentication
US20160103963A1 (en) Method and system for smart healthcare management
US9391986B2 (en) Method and apparatus for providing multi-sensor multi-factor identity verification
US20170169168A1 (en) Patient identification systems and methods
EP2946323B1 (en) Secure real-time health record exchange
WO2020037454A1 (en) Smart auxiliary diagnosis and treatment system and method
US20170068785A1 (en) Secure real-time health record exchange
US20020194131A1 (en) Method and system for electronically transmitting authorization to release medical information
US20130124223A1 (en) Systems and Methods for Reducing Medical Claims Fraud
WO2016144632A2 (en) Method and apparatus for remote identity proofing service issuing trusted identities
US20160180048A1 (en) Cloud-based medical information retrieval method and system thereof
WO2024057839A1 (en) Acceptance assistance system, acceptance assistance method, and acceptance assistance program
US20200380163A1 (en) Process for collecting electronic protected health information without a login
US20080107308A1 (en) Medical biometric identification security system
KR102028844B1 (en) Insurance Benefit Automatic Claim System capable of Automatically Submitting Medical Record and Proof Documents
US20220114675A1 (en) Audit trail and auto reimbursement
KR102451457B1 (en) Payment system of non-faced-to-face treatment through selectively matching doctors and patients
JP6841467B1 (en) Survival insurance system, information processing device, information processing method, and program
US20190035495A1 (en) Method for configuring diabetes management device by healthcare provider
US20160070924A1 (en) Virtual-Account-Initiated Communication of Protected Information
JP2007328616A (en) Contract system
JP7448965B2 (en) Application support system, benefit application system, application support method and program
KR102429183B1 (en) System of non-faced-to-face treatment through selectively accepting of treatment by the doctor
KR20200067636A (en) Method for processing and verificating clinical trial data

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23865197

Country of ref document: EP

Kind code of ref document: A1