WO2020078058A1 - Procédé et dispositif d'identification d'anomalies de données médicales, terminal et support de stockage - Google Patents

Procédé et dispositif d'identification d'anomalies de données médicales, terminal et support de stockage Download PDF

Info

Publication number
WO2020078058A1
WO2020078058A1 PCT/CN2019/096631 CN2019096631W WO2020078058A1 WO 2020078058 A1 WO2020078058 A1 WO 2020078058A1 CN 2019096631 W CN2019096631 W CN 2019096631W WO 2020078058 A1 WO2020078058 A1 WO 2020078058A1
Authority
WO
WIPO (PCT)
Prior art keywords
medical data
preset
disease type
abnormality
days
Prior art date
Application number
PCT/CN2019/096631
Other languages
English (en)
Chinese (zh)
Inventor
周竹凌
汪丽娟
Original Assignee
平安医疗健康管理股份有限公司
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 平安医疗健康管理股份有限公司 filed Critical 平安医疗健康管理股份有限公司
Publication of WO2020078058A1 publication Critical patent/WO2020078058A1/fr

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
    • G06Q40/06—Asset management; Financial planning or analysis
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients

Definitions

  • the present application relates to the technical field of medical data processing, and in particular, to a method, device, terminal, and computer-readable storage medium for identifying abnormal medical data.
  • the medical insurance system is a kind of social compulsory insurance for the state to prevent and share the medical expenses caused by diseases.
  • the expenses are jointly paid by the employer and the individual, and the medical insurance money is paid by the medical insurance institution to solve the laborers ’illness or illness. Medical risks from injury.
  • the medical insurance department will evaluate the hospital.
  • the number of single hospital stays is an important evaluation index of the hospital's medical efficiency.
  • the main purpose of the present application is to provide a method for identifying abnormal medical data, aiming to solve the technical problem that the medical insurance card behavior data cannot be detected abnormally.
  • the present application provides a medical data abnormality identification method, characterized in that the medical data abnormality identification method includes the following steps:
  • the target hospital corresponding to the abnormality detection request and the medical data to be detected of the target hospital are obtained;
  • the target medical data is marked abnormally.
  • the present application also provides a medical data abnormality identification device
  • the medical data abnormality identification device includes:
  • An obtaining module configured to obtain the target hospital corresponding to the abnormal detection request and the medical data to be detected of the target hospital when the abnormal detection request is detected;
  • a grouping module configured to group the medical data to be detected according to a preset dimension to obtain a grouped subset
  • An anomaly analysis module configured to analyze each of the subsets based on a preset rule corresponding to a preset dimension, and determine whether there is target medical data satisfying an abnormal feature in the medical data to be detected;
  • the abnormal marking module is configured to mark the target medical data abnormally if there is target medical data satisfying abnormal characteristics in the medical data to be detected.
  • the present application also provides a medical data abnormality identification terminal
  • the medical data abnormality identification terminal includes a processor, a memory, and a computer stored on the memory and executable by the processor. Read instructions, where the computer-readable instructions are executed by the processor to implement the steps of the medical data abnormality identification method as described above.
  • the present application also provides a computer-readable storage medium having computer-readable instructions stored on the computer-readable storage medium, wherein when the computer-readable instructions are executed by a processor, the implementation is as described above Steps of the method for identifying abnormal medical data.
  • FIG. 1 is a schematic structural diagram of a medical data abnormality recognition terminal in a hardware operating environment involved in an embodiment of the present application
  • FIG. 2 is a schematic flowchart of a first embodiment of a medical data abnormality identification method of the application
  • FIG. 3 is a schematic flowchart of a second embodiment of a medical data abnormality identification method of the present application.
  • FIG. 4 is a schematic flowchart of a fifth embodiment of a medical data abnormality identification method of the present application.
  • FIG. 5 is a schematic diagram of functional modules of the first embodiment of the medical data abnormality recognition device of the present application.
  • FIG. 1 is a schematic diagram of the hardware structure of the medical data abnormality identification terminal provided by the present application.
  • the medical data abnormality identification terminal may be a PC or a device terminal with a display function such as a smart phone, tablet computer, portable computer, desktop computer, etc.
  • the medical data abnormality identification terminal may be a server device that exists A back-end management system for medical data abnormality identification, and a user manages the medical data abnormality identification terminal through the back-end management system.
  • the medical data abnormality identification terminal may include components such as a processor 101 and a memory 201.
  • the processor 101 is connected to the memory 201, and the memory 201 stores computer-readable instructions, and the processor 101 can call the computer-readable instructions stored in the memory 201, and The steps of each embodiment of the medical data abnormality recognition method described below are implemented.
  • the memory 201 can be used to store software computer readable instructions and various data.
  • the memory 201 may mainly include an area for storing computer-readable instructions and an area for storing data.
  • the memory 201 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other volatile solid-state storage devices.
  • the processor 101 is the control center of the medical data abnormality recognition terminal, and uses various interfaces and lines to connect the various parts of the entire medical data abnormality recognition terminal, by running or executing software computer readable instructions and / or modules stored in the memory 201 , And call the data stored in the memory 201 to perform various functions and process data of the medical data abnormality recognition terminal, thereby performing overall monitoring of the medical data abnormality recognition terminal.
  • the processor 101 may include one or more processing units; optionally, the processor 101 may integrate an application processor and a modem processor, where the application processor mainly processes an operating system, a user interface, and application computer-readable instructions, etc.
  • the modem processor mainly handles wireless communication. It can be understood that the foregoing modem processor may not be integrated into the processor 101.
  • the structure of the medical data abnormality identification terminal shown in FIG. 1 does not constitute a limitation on the medical data abnormality identification terminal, and may include more or less components than the illustration, or a combination of certain components, Or different component arrangements.
  • the identification terminal in the following is an abbreviation of a medical data abnormality identification terminal.
  • This application provides a method for identifying abnormal medical data.
  • FIG. 2 is a schematic flowchart of a first embodiment of a medical data abnormality recognition method of this application.
  • the medical data abnormality identification method includes the following steps:
  • Step S10 when an abnormality detection request is detected, the target hospital corresponding to the abnormality detection request and the medical data to be detected of the target hospital are obtained;
  • the anomaly detection request is an instruction that triggers the start of the computer-readable instruction corresponding to the medical data anomaly detection method of the present application.
  • the anomaly detection request can be triggered by a preset external event / page request, such as a click operation for anomaly detection input by a user; it can also be passed Use the timed JOB (task) method as the starting interface for computer-readable instructions, which triggers anomaly detection of medical data at regular intervals. It can also perform anomaly detection immediately after receiving medical data to be detected. Anomaly detection is performed after the amount of medical data to be detected reaches a certain threshold.
  • the anomaly detection of medical data in this application may be one or more process nodes in the medical data review process, and the abnormality detection of medical data may be automatically triggered by the flow of the process nodes.
  • the medical data of the target hospital to be detected is used as the basis for abnormal identification.
  • the anomaly detection request contains the identification of the corresponding target hospital.
  • the medical data to be detected is the specific medical data pointed to by the abnormal detection request and the target hospital requested for detection.
  • the medical data includes but is not limited to: type of treatment, identification of treatment items, time of treatment, diagnosis results, medication situation, amount of treatment, amount of medication Wait.
  • the card swiping device After the insured person swipes the medical insurance card, the card swiping device will upload the relevant medical data to the hospital terminal, and then the hospital terminal sends it to the identification terminal or directly uploads it to the identification terminal, which can be directly stored in the memory of the identification terminal, when performing abnormal detection ,
  • the identification terminal directly obtains the medical data to be detected from the local storage according to the abnormal detection request, or it can be stored in the data storage terminal / cloud of the medical data, so that the identification terminal can detect the abnormal medical data from the data according to the abnormal detection request
  • the storage terminal / cloud obtains medical data to be tested.
  • Step S20 Group the medical data to be detected according to a preset dimension to obtain a grouped subset
  • the preset dimension includes at least one of the hospital dimension and the personal dimension.
  • the hospital dimension is based on the analysis of all medical data of the target hospital to determine whether there is abnormal medical data
  • the personal dimension is based on the medical data analysis of each participant to determine whether there is abnormal medical data .
  • Each preset dimension can be set as a clickable control, and the preset dimension for analysis selected by the user can be determined according to the user's click operation, or the default preset dimension can be directly obtained.
  • the preset dimensions are different, and the medical data to be tested are grouped in different ways.
  • the identification terminal groups the medical data to be tested according to the disease type, and a corresponding subset of each disease type can be obtained;
  • the preset dimension is an individual, the identification terminal groups the medical data to be tested according to the insured person By grouping, you can get a subset of each insured person.
  • the identification terminal may simultaneously and separately analyze the medical data to be detected according to the hospital dimension and the individual dimension to identify abnormal data.
  • Step S30 Analyze each of the subsets based on a preset rule corresponding to a preset dimension to determine whether there is target medical data satisfying abnormal characteristics in the medical data to be detected;
  • Each preset dimension has its own corresponding preset rule, and each subset is the analysis object of the preset rule.
  • the preset dimension is a hospital dimension
  • the medical data to be tested are grouped according to disease types to obtain a first subset corresponding to each disease type, based on the Calculate the average number of single hospitalization days corresponding to each disease type, and / or average number of hospitalizations per person, and / or average amount of single hospitalization, and calculate the average number of single hospitalization days corresponding to each disease type, and / or per person
  • the average number of hospitalizations, and / or the average single hospitalization amount are compared with preset values corresponding to each disease type, and it is determined whether the first subset corresponding to each disease type meets the abnormal feature.
  • Abnormal characteristics refer to the characteristics set in advance to limit abnormal medical data. For example, the average number of days of hospitalization corresponding to each disease type is compared with the preset value corresponding to each type of disease. The number of hospitalization days is greater than the preset value corresponding to each disease type, then the first subset corresponding to the disease type meets the abnormal characteristics.
  • Step S40 if there is target medical data satisfying abnormal characteristics in the medical data to be detected, the target medical data is marked abnormally.
  • the target medical data is marked abnormally, and the abnormal detection result is output to prompt the user to perform offline verification. If there is no abnormal target medical data in the medical data to be detected, it is prompted that there is no abnormality.
  • the abnormality so as to identify the terminal and / or the user to select the corresponding processing method for the target medical data to process, and obtain the abnormal type of the target medical data; according to the abnormal type, obtain the corresponding abnormality identifier to detect medical treatment
  • the data is marked abnormally. Further, when the correction instruction input after the offline verification by the user is detected, the state of the target medical data is modified to be normal.
  • the target hospital corresponding to the abnormality detection request and the medical data to be detected of the target hospital are obtained; the medical data to be detected are grouped according to a preset dimension to obtain the grouped Sub-sets; analyze each of the sub-sets based on preset rules corresponding to preset dimensions to determine whether there is target medical data satisfying abnormal characteristics in the medical data to be tested; For target medical data, the target medical data is abnormally marked.
  • the medical data to be tested of the target hospital can be obtained, and the target hospital can be targeted to analyze and identify whether the target hospital violates the medical insurance regulations; Grouped analysis can effectively detect abnormal medical data; mark abnormal medical target data that meets abnormal characteristics in the medical data to be detected, and distinguish the abnormal medical data for subsequent targeted treatment of abnormal medical data to standardize medical treatment The normal operation of the expense reimbursement system.
  • the step S20 includes:
  • Step S21 Group the medical data to be tested according to the disease type, and obtain a first sub-set corresponding to each disease type;
  • Disease types including primary diagnosis (such as diabetes) and secondary diagnosis / complication (kidney disease, fundus disease, neuropathy, etc.), when grouping, according to the primary diagnosis + secondary diagnosis, that is, the medical data is grouped according to the breakdown .
  • primary diagnosis such as diabetes
  • secondary diagnosis / complication kidney disease, fundus disease, neuropathy, etc.
  • the step S30 includes:
  • Step S310 Obtain the total number of hospitalizations and the total number of hospitalization days of each disease type from the first subset corresponding to each disease type, and calculate and obtain the average number of each disease type based on the total number of hospitalizations and the total number of hospitalization days Number of hospital stays;
  • the average number of hospital days per disease refers to the average number of hospital days for a disease.
  • the average number of hospital stays for different diseases is different. Therefore, the average number of hospital stays for each disease is calculated based on the first subset corresponding to each disease.
  • the medical data of a single hospitalization day above / below the first preset value can be marked in the first subset corresponding to the disease type Come out and further verify, and at the same time, calculate the average number of days of hospitalization for the disease based on the remaining medical data.
  • the number of hospitalization days is mostly between 4-10 days, but once the hospitalization was 2 days and once the hospitalization was 14 days, the two abnormal data were marked for further verification, and the These two abnormal data are eliminated, and the average number of hospital stays is calculated based on the remaining 18 data.
  • Step S311 Compare the average number of hospitalization days of each disease type with the preset days corresponding to each disease type, obtain the difference between the average number of hospitalization days of each disease type and the corresponding preset days, and calculate each disease type The ratio of the difference in the number of days to the corresponding preset number of days to obtain the abnormal factor for the number of days corresponding to each disease type;
  • the preset number of days is calculated based on the average number of hospital stays in the same class and the same type of disease in the same type of medical institutions that have been reviewed.
  • the same level refers to the general hospital grades such as "Third Class A Hospital” and "Second Class A Hospital”, and the category refers to the same specialty. For example, they belong to the specialty of children, and the same disease includes the main diagnosis and the auxiliary diagnosis.
  • the preset number of days can be a range.
  • the abnormal number of days refers to the parameter used to identify the abnormal degree of the average number of days in a single hospital stay.
  • the ratio of the difference between the average number of days of hospitalization for each disease type and the number of days corresponding to the preset number of days and the number of days corresponding to the preset number of days is used as the abnormal factor for the number of days corresponding to each disease type. For example, if the preset number of days for diabetes complicated with lactic acidosis is 15 days, and the average number of single hospital stays in the target hospital is 7 days, the difference is 8 days, then it is calculated based on the preset days, average number of single hospital stays, and the difference in days Days anomaly factor, the days anomaly factor in the above example is (15-7) / 15.
  • the number of days abnormal factor When the number of days abnormal factor is 0, that is, the difference between the average number of days of single hospitalization and the corresponding preset days is 0, indicating that the average number of days of corresponding single hospitalization of the corresponding disease is not abnormal; when the number of days abnormal factor ⁇ 0, it is the average single The number of hospitalization days is greater than the corresponding preset number of days, indicating that there can be no decomposition hospitalization for the corresponding disease, that is, the average number of single hospitalization days is not abnormal; when the abnormality factor of the number of days> 0, the average number of single hospitalization days is less than the corresponding preset number of days, indicating that the corresponding disease The average number of hospitalization days for each species is abnormal, and the value indicates the degree of abnormality.
  • Step S312 when the abnormal number of days corresponding to any disease type is greater than the first preset threshold, the first subset corresponding to the disease type satisfies the abnormal characteristics.
  • the first preset threshold refers to a value preset by the user and can be adjusted according to actual conditions.
  • the abnormal score of the target hospital may be calculated according to the abnormal number of days corresponding to each disease type.
  • the abnormal number of days corresponding to each disease type is integrated to calculate the abnormal score of the target hospital.
  • the sum of all the abnormal factors of the number of days can be used as the abnormal score of the target hospital.
  • the degree of instability of the disease (a disease with a large individual difference) is assigned a reference value, and the sum of the product of the reference value corresponding to each disease and the abnormal number of days is used as the abnormal score of the target hospital.
  • the average number of single hospitalization days of each disease type is obtained by calculation, and the average number of single hospitalization days of each disease type of the target hospital is compared with the average number of single hospitalization days of the same disease type and medical institution of the same class (ie Preset days) to compare, identify the medical data in the target hospital's medical data that does not meet the usual situation, and characterize the degree of abnormality by the number of days abnormal factor, when the number of days abnormal factor is greater than the first preset threshold, the corresponding disease type
  • the first sub-set is determined to satisfy the abnormal features and realize the identification of abnormal medical data.
  • the step S30 includes:
  • Step S313 Obtain the number of insured persons and the total number of hospitalizations of each disease type from the first subset corresponding to each disease type, and calculate the average per capita corresponding to each disease type based on the number of insured persons of each disease type and the total number of hospitalizations Number of hospitalizations;
  • the first subset corresponding to each disease type contains all medical data corresponding to each disease type in the target hospital. From the first subset corresponding to each disease type, the number of insured persons corresponding to each disease type and the total number of hospitalizations are counted. The average number of hospitalizations per person can be obtained by dividing the total number of hospitalizations by the number of insured persons. For example, if diabetes is complicated by lactic acidosis, the total number of hospitalizations is 100, and the number of insured persons is 14, then the average number of hospitalizations per person for this disease is 100/14.
  • Step S314 Compare the average number of hospitalizations per person for each disease type with the corresponding preset number of times, obtain the difference between the number of average hospitalizations per person for each disease type and the corresponding preset number of times, and calculate the number of times for each disease type The ratio of the difference to the corresponding preset number of times to obtain the abnormal factor of the number of times corresponding to each disease type;
  • the preset number of times is calculated based on the average number of hospitalizations of the same diagnosis (primary diagnosis / subdiagnosis) of the same disease in the same level and the same category of medical institutions under review.
  • the number of abnormal factors refers to the parameter used to identify the abnormal degree of the average number of hospitalizations per person.
  • the frequency anomaly factor in the above example is (17-10) / 10.
  • the number of abnormal factors When the number of abnormal factors is 0, that is, the difference between the average number of hospitalizations per person and the corresponding preset number of times is 0, indicating that the average number of hospitalizations per person for the corresponding disease type is not abnormal; when the number of abnormal factors ⁇ 0, the average The number of hospitalizations is less than the corresponding preset number, indicating that there can be no decomposition hospitalization for the corresponding disease, that is, the average number of hospitalizations per person is not abnormal; when the number of abnormal factors is> 0, the average number of hospitalizations per person is greater than the corresponding preset number, indicating the corresponding disease
  • the average number of hospitalizations per person for each species is abnormal, and the value indicates the degree of abnormality. The larger the value, the greater the degree of abnormality.
  • Step S315 when the abnormality factor of the number of times corresponding to any disease type is greater than the second preset threshold, the first subset corresponding to the disease type satisfies the abnormal characteristics.
  • the second preset threshold refers to a value preset by the user and can be adjusted according to actual conditions.
  • the abnormal score of the target hospital may be calculated according to the abnormal factor of the number of times corresponding to each disease type.
  • the abnormal factors of the number of times corresponding to each disease are integrated to calculate the abnormal score of the target hospital.
  • the sum of all the abnormal factors of the number of times can be used as the abnormal score of the target hospital;
  • the degree of instability of the disease type (the disease type with large individual difference) is assigned a reference value, and the sum of the product of the reference value corresponding to each disease type and the number of abnormal factors is used as the abnormal score of the target hospital;
  • the abnormal number of days and the abnormal number of times are combined to calculate the overall abnormal score of the target hospital, with more participation dimensions and more accurate abnormal scores.
  • the average number of hospitalizations per person of each disease type is obtained by calculation, and the average number of hospitalizations per person of each disease type in the target hospital is compared with the average number of hospitalizations per person of the same disease type and medical institution of the same class (ie Preset times) to compare, identify the medical data in the target hospital's medical data that does not meet the usual situation, and characterize the degree of abnormality by the number of times abnormal factor, when the number of times abnormal factor is greater than the second preset threshold, the corresponding disease type
  • the first sub-set is determined to satisfy the abnormal features and realize the identification of abnormal medical data.
  • the step S30 includes:
  • Step S316 Obtain the total number of hospitalizations and total hospitalization amount of each disease type from the first subset corresponding to each disease type, and calculate and obtain the average bill of each disease type based on the total number of hospitalizations and total hospitalization amount of each disease type Amount of second hospitalization;
  • the average amount of single hospitalization for each disease refers to the average amount of single hospitalization for a certain disease.
  • the average single hospitalization amount for different diseases is different. Therefore, the average single hospitalization amount for each disease type is calculated based on the first subset corresponding to each disease type.
  • the medical data of a single hospitalization amount higher / lower than the first preset value may be marked in the first subset corresponding to the disease type Come out and further verify, at the same time, calculate the average single hospitalization amount of the disease based on the remaining medical data.
  • Step S317 compare the average single hospitalization amount of each disease type with the corresponding preset amount, obtain the difference between the average single hospitalization amount of each disease type and the corresponding preset amount, and calculate the The ratio of the amount difference to the corresponding preset amount to obtain the amount abnormality factor corresponding to each disease type;
  • the preset amount is calculated based on the average single hospitalization amount of the same class and the same disease in the medical institutions of the same class that have been reviewed. For example, they belong to the specialty of children, and the same disease includes the main diagnosis and the auxiliary diagnosis.
  • the preset amount can be a range of values.
  • the amount abnormality factor refers to the parameter used to identify the abnormal degree of the average single hospitalization amount.
  • the ratio of the difference between the average single hospitalization amount of each disease type and the corresponding preset amount and the corresponding preset amount is used as the amount abnormality factor corresponding to each disease type. For example, if the preset amount of diabetes complicated with lactic acidosis is 10,000 yuan / time, and the average single hospitalization amount of the target hospital is 5,000 yuan / time, the difference is 5000 yuan / time, according to the preset amount, the average single The hospitalization amount and the amount difference are calculated as the amount abnormality factor.
  • the amount abnormality factor in the above example is (10000-5000) / 10000. If the average amount of a single hospitalization is relatively small, it may be shared with another hospitalization, that is, there may be decomposition of hospitalization.
  • the amount abnormality factor ⁇ 1 it means there is an abnormality. The smaller the value, the greater the degree of abnormality.
  • the third preset threshold refers to a value preset by the user and can be adjusted according to actual conditions.
  • the abnormal score of the target hospital may be calculated according to the amount abnormality factor corresponding to each disease type.
  • the abnormal amount of money corresponding to each disease type is integrated to calculate the abnormal score of the target hospital.
  • the sum of all abnormal amounts of money can be used as the abnormal score of the target hospital;
  • the degree of instability of the disease type (the disease with large individual difference) is assigned a reference value, and the sum of the product of the reference value corresponding to each disease type and the amount abnormality factor is used as the abnormal score of the target hospital; it can also be based on the amount abnormality factor and the number of days
  • the abnormal factor and the frequency abnormal factor calculate the abnormal score of the target hospital.
  • the average single hospitalization amount of each disease type is obtained by calculation, and the average single hospitalization amount of each disease type of the target hospital is compared with the average single hospitalization amount of the same type and type of medical institution (ie. Preset amount) to compare, identify the medical data in the target hospital's medical data that does not meet the usual situation, and characterize the degree of abnormality by the amount abnormality factor, when the amount abnormality factor is greater than the third preset threshold, the corresponding disease type
  • the first sub-set is determined to satisfy the abnormal features and realize the identification of abnormal medical data.
  • the step S20 includes:
  • Step S22 Group the medical data to be tested according to the insured persons to obtain a second subset corresponding to each insured person;
  • the medical data set to be tested corresponding to each insured person is the second subset.
  • the step S30 includes:
  • Step S320 extracting the consultation time corresponding to each piece of medical data in the second sub-collection, and calculating the interval between adjacent consultations to obtain the adjacent medical data in the second sub-collection;
  • the identification terminal can obtain the visit time of each visit from the medical data, wherein each visit generates a corresponding piece of medical data, which is arranged in order of the visit time, and then calculates the adjacent visit interval for arranging the adjacent medical data.
  • the identification terminal obtains the adjacent visit intervals of all adjacent medical data in the second subset; in another embodiment, the second subset is regrouped according to the disease type to obtain the third subset, and the third Each piece of medical data in the three sub-collections is arranged according to the order of visit time, and the adjacent visit interval of the adjacent hospital visits of the same disease type is calculated.
  • Step S321 Obtain a preset time interval threshold, calculate a time difference between the adjacent consultation interval and the time interval threshold, calculate a ratio of the time difference to the time interval threshold, and obtain each item in the second subset The interval anomaly factor corresponding to medical data;
  • the recognition terminal may calculate the difference between the adjacent visit interval and the time interval threshold, and determine the interval anomaly factor corresponding to each piece of medical data according to the difference.
  • the adjacent visit interval is less than the time interval threshold, the two The greater the difference is, the greater the interval anomaly factor.
  • the adjacent visit interval is greater than or equal to the time interval threshold, the greater the difference between the two, the smaller the interval anomaly factor.
  • each piece of medical data corresponds to at least one interval anomaly factor, for example, [1, 2], [2, 3], etc., where the number refers to a single parameter
  • the serial number of the insured ’s hospitalization times is arranged in chronological order.
  • interval anomaly factors which are the interval anomaly factor and the second hospitalization for the second hospitalization and the first hospitalization, respectively.
  • Step S322 when the interval abnormality factor corresponding to the medical data in the second subset is greater than the fourth preset threshold, the medical data meets the abnormality feature.
  • the fourth preset threshold refers to a value preset by the user and can be adjusted according to actual conditions.
  • the neighboring visit interval of the neighboring medical data in the second subset is calculated; the preset time interval threshold is obtained, and the The time difference between the adjacent visit interval and the time interval threshold, calculating the ratio of the time difference to the time interval threshold to obtain the interval anomaly factor corresponding to each piece of medical data in the second subset; in the second
  • the interval anomaly factor corresponding to the medical data in the subset is greater than the fourth preset threshold, the medical data meets the anomaly characteristics, because for the same insured person, if the interval between adjacent hospitalizations is too short, there may be decomposition hospitalization Suspected, by calculating the adjacent medical interval of adjacent medical data, and screening out the medical data with a large degree of suspiciousness based on the interval abnormality factor corresponding to each medical data, it can accurately determine whether the medical data contains suspicious decomposition of hospitalization behavior, effective Prevent fraudulent behavior and reduce loss of public interest.
  • step after the step S321 includes:
  • Step S323 Extract the diagnosis result corresponding to each piece of medical data in the second subset, and determine whether there is adjacent medical data corresponding to a similar diagnosis result in the second subset;
  • the diagnosis result refers to the doctor's judgment on the cause of the insured person's illness.
  • the diagnosis is diabetic complicated with lactic acidosis.
  • the diagnosis result includes the main diagnosis and the auxiliary diagnosis.
  • the similar diagnosis results in this embodiment include the same diagnosis results and similar diagnosis results.
  • a diagnosis result database can be established in advance, and the same diagnosis results and similar diagnosis results with different names can be stored in the diagnosis result database to determine the diagnosis of adjacent medical data. When the result is similar to the diagnosis result, it can be judged by querying the diagnosis result database.
  • Step S323 if there is adjacent medical data corresponding to a similar diagnosis result in the second subset, the time difference between the adjacent visit interval of the adjacent medical data and the time interval threshold is obtained, and the time difference and The ratio of the time interval threshold to obtain the interval anomaly factor corresponding to the adjacent medical data;
  • the interval anomaly factor determines whether the two or more adjacent medical data have an abnormal time (if there is an abnormality, the size of the abnormality).
  • Step S324 when the interval anomaly factor corresponding to the adjacent medical data is greater than the fifth preset threshold, the medical data meets the anomaly characteristics.
  • the interval anomaly factor corresponding to the adjacent medical data is greater than the fifth preset threshold, the possibility that the adjacent medical data is decomposed into inpatient medical data is further increased. Because if the medical treatment interval corresponding to the adjacent medical data is small and the diagnosis results are the same or similar, the adjacent medical data is likely to be the data for decomposing hospitalization.
  • the step S30 includes:
  • Step S325 Extract the diagnosis and treatment items corresponding to each piece of medical data in the second sub-collection, and compare the diagnosis and treatment items corresponding to each adjacent consultation with the reference diagnosis and treatment item set respectively to obtain the corresponding diagnosis and treatment items and reference diagnosis and treatment for each adjacent treatment The coincidence ratio of the item set, and calculate the average coincidence ratio of all adjacent visits to obtain the continuity of the adjacent treatment items;
  • the identification terminal After the identification terminal arranges each inpatient consultation according to the time sequence of the consultation, it can obtain the identification of the treatment items included in the adjacent inpatient consultations, and compare the corresponding treatment items of the adjacent consultations with the reference set of treatment items, Obtain the coincidence ratio of the corresponding treatment items of the adjacent consultations and the set of reference treatment items, and calculate the average coincidence ratio of all the adjacent consultations to obtain the continuity of the treatment items of the adjacent consultations.
  • a database of diagnosis and treatment specifications corresponding to various diagnosis results can be established in advance, including the diagnosis and treatment items (identification) corresponding to each diagnosis result. Because the treatment of a disease often follows a certain clinical / treatment path, therefore, the corresponding diagnosis results The diagnosis and treatment items are arranged in sequence based on a certain clinical / treatment path.
  • the diagnosis and treatment items are in the reference diagnosis and treatment item set, specifically: the diagnosis and treatment items in the reference diagnosis and treatment item set account for the proportion of the adjacent diagnosis and treatment items, and the average coincidence ratio of all adjacent visits is calculated, and the average coincidence ratio is that of the adjacent consultations Continuity of diagnosis and treatment items.
  • the nth visit is the n + 1 visit as a neighboring visit
  • the nth visit is A, B, C
  • the n + 1 visit is D, E, F, G, H
  • the reference diagnosis and treatment item set is [A, B, C, D, E, F, G]
  • the coincidence ratio of the diagnosis and treatment items corresponding to the nth visit and the reference diagnosis and treatment item set is 100%, corresponding to the n + 1th visit
  • Step S326 when the continuity is greater than the sixth preset threshold, the corresponding medical data meets abnormal characteristics.
  • the sixth preset threshold refers to a value preset by the user and can be adjusted according to actual conditions.
  • the sixth preset threshold may be 45%.
  • step S325 in this embodiment it may further include a judgment based on the similarity between the adjacent medical intervals of the adjacent medical data and the diagnosis results, specifically including:
  • the interval between adjacent visits based on adjacent medical data, the similarity of diagnosis results, and the continuity of the diagnosis and treatment items of adjacent visits can improve the accuracy of identifying abnormal medical data, reduce false positives, and increase the flexibility of abnormal detection.
  • the corresponding medical data satisfies the abnormal characteristics, and the judgment on the decomposition of the hospitalization behavior can be realized based on the clinical treatment path, which is effective Prevent fraudulent behavior and reduce loss of public interest.
  • the present application also provides a medical data abnormality recognition device corresponding to each step of the foregoing medical data abnormality recognition method.
  • FIG. 5 is a schematic diagram of functional modules of the first embodiment of the medical data abnormality recognition device of the present application.
  • the medical data abnormality identification device of the present application includes:
  • the obtaining module 10 is configured to obtain the target hospital corresponding to the abnormality detection request and the medical data to be detected of the target hospital when the abnormality detection request is detected;
  • a grouping module 20 configured to group the medical data to be detected according to a preset dimension to obtain a grouped subset
  • An anomaly analysis module 30 configured to analyze each of the subsets based on a preset rule corresponding to a preset dimension, and determine whether there is target medical data satisfying anomalous characteristics in the medical data to be detected;
  • the abnormal marking module 40 is configured to mark the target medical data abnormally if there is target medical data satisfying abnormal characteristics in the medical data to be detected.
  • the medical data abnormality identification device includes:
  • the grouping module 20 is further configured to group the medical data to be tested according to the disease type when the preset dimension is the hospital dimension, to obtain a first subset corresponding to each disease type;
  • the first calculation module is used to obtain the total number of hospitalizations and the total number of hospitalization days of each disease type from the first subset corresponding to each disease type, and calculate and obtain each disease based on the total number of hospitalizations and the total number of hospitalization days of each disease type The average number of hospital stays per species; compare the average number of hospital stays for each disease with the preset days corresponding to each disease to obtain the difference between the average number of hospital stays for each disease and the number of days corresponding to the preset days , Calculate the ratio of the difference between the number of days of each disease type and the corresponding preset number of days to obtain the abnormal factor of the number of days corresponding to each disease type;
  • the first abnormality determination module is configured to: when the abnormality factor of the number of days corresponding to any disease type is greater than the first preset threshold, the first subset corresponding to the disease type meets the abnormality feature.
  • the medical data abnormality identification device includes:
  • the second calculation module is used to obtain the number of insured persons and the total number of hospitalizations of each disease type from the first sub-collection corresponding to each disease type, and calculate and obtain the correspondence of each disease type based on the number of insured persons of each disease type and the total number of hospitalizations
  • the average number of hospitalizations per person for each patient compare the average number of hospitalizations per person for each disease type with the corresponding preset number of times to obtain the difference between the average number of hospitalizations per person for each disease type and the corresponding preset number of times to calculate each disease type
  • the second abnormality determination module is configured to: when the abnormality factor of the number of times corresponding to any disease type is greater than the second preset threshold, the first subset corresponding to the disease type satisfies abnormal characteristics.
  • the medical data abnormality identification device includes:
  • the third calculation module is used to obtain the total number of hospitalizations and total hospitalization amount of each disease type from the first subset corresponding to each disease type, and calculate and obtain each disease based on the total number of hospitalizations and total hospitalization amount of each disease type
  • the average single hospitalization amount of each kind compare the average single hospitalization amount of each disease with the corresponding preset amount to obtain the difference between the average single hospitalization amount of each disease and the corresponding preset amount, and calculate each The ratio of the amount difference of the disease type to the corresponding preset amount to obtain the amount abnormality factor corresponding to each disease type;
  • the third abnormality determination module is configured to: when the amount abnormality factor corresponding to any disease type is greater than the third preset threshold, the first subset corresponding to the disease type meets abnormal characteristics.
  • the medical data abnormality identification device includes:
  • the grouping module 20 is further configured to group the medical data to be tested according to the insured persons when the preset dimension is an individual dimension to obtain a second subset corresponding to each insured person;
  • the fourth calculation module is used for extracting the consultation time corresponding to each piece of medical data in the second subset, calculating the adjacent consultation interval for obtaining the adjacent medical data in the second subset; obtaining the preset time interval threshold Calculate the time difference between the adjacent consultation interval and the time interval threshold, calculate the ratio of the time difference and the time interval threshold to obtain the interval abnormality factor corresponding to each piece of medical data in the second subset;
  • the fourth abnormality determination module is configured to: when the interval abnormality factor corresponding to the medical data in the second subset is greater than the fourth preset threshold, the medical data meets the abnormality feature.
  • the medical data abnormality identification device includes:
  • the similarity judgment module is used to extract the diagnosis result corresponding to each piece of medical data in the second subset and determine whether there is adjacent medical data corresponding to the similar diagnosis result in the second subset;
  • the first obtaining module is used to obtain the time difference between the adjacent medical visit interval of the adjacent medical data and the time interval threshold if there is adjacent medical data corresponding to similar diagnosis results in the second subset The ratio of the time difference to the time interval threshold to obtain an interval anomaly factor corresponding to the adjacent medical data;
  • the fifth abnormality determination module is configured to: when the interval abnormality factor corresponding to the adjacent medical data is greater than a fifth preset threshold, the medical data meets abnormal characteristics.
  • the medical data abnormality identification device includes:
  • the fifth calculation module is used for extracting the diagnosis and treatment items corresponding to each piece of medical data in the second sub-collection, respectively comparing the diagnosis and treatment items corresponding to the adjacent treatments with the reference diagnosis and treatment item set, and obtaining the corresponding treatments for the adjacent treatments.
  • the sixth abnormality determination module is configured to: when the continuity is greater than the sixth preset threshold, the corresponding medical data meets abnormal characteristics.
  • the present application also proposes a computer-readable storage medium on which computer-readable instructions are stored.
  • the computer-readable storage medium may be a non-volatile readable storage medium.
  • the computer-readable storage medium may be the memory 201 in the medical data abnormality recognition terminal of FIG. 1, or may be a ROM (Read-Only Memory, read-only memory) / RAM (Random Access Memory (random access memory), at least one of magnetic disks and optical disks, the computer-readable storage medium includes several instructions to make a terminal device (which may be a mobile phone, computer, server, network device) with a processor Or the medical data abnormality identification terminal in the embodiments of the present application, etc.) execute the methods described in the embodiments of the present application.
  • a terminal device which may be a mobile phone, computer, server, network device

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Data Mining & Analysis (AREA)
  • Health & Medical Sciences (AREA)
  • Finance (AREA)
  • Medical Informatics (AREA)
  • Development Economics (AREA)
  • Public Health (AREA)
  • Accounting & Taxation (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Operations Research (AREA)
  • General Health & Medical Sciences (AREA)
  • Epidemiology (AREA)
  • Pathology (AREA)
  • Databases & Information Systems (AREA)
  • Biomedical Technology (AREA)
  • Game Theory and Decision Science (AREA)
  • Human Resources & Organizations (AREA)
  • Primary Health Care (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • Technology Law (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Medical Treatment And Welfare Office Work (AREA)

Abstract

L'invention concerne un procédé et un dispositif d'identification d'anomalies de données médicales basée sur des mégadonnées, un terminal, et un support de stockage lisible par ordinateur. Le procédé comprend les étapes consistant à : lors de la détection d'une requête de détection d'anomalie, obtenir d'un hôpital cible correspondant à la requête de détection d'anomalie et des données médicales à détecter de l'hôpital cible (S10) ; regrouper lesdites données médicales selon une dimension prédéfinie pour obtenir des sous-ensembles groupés (S20) ; analyser chaque sous-ensemble sur la base d'une règle prédéfinie correspondant à la dimension prédéfinie, et déterminer s'il y a des données médicales cibles répondant à des caractéristiques anormales dans lesdites données médicales (S30) ; et s'il y a des données médicales cibles répondant aux caractéristiques d'anomalie dans lesdites données médicales, marquer les données médicales cibles avec une anomalie (S40). Au moyen du procédé, des comportements d'hospitalisation sectoriels peuvent être identifiés avec précision, ce qui permet d'empêcher efficacement des comportements d'assurance frauduleux.
PCT/CN2019/096631 2018-10-19 2019-07-19 Procédé et dispositif d'identification d'anomalies de données médicales, terminal et support de stockage WO2020078058A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811226809.0A CN109636613A (zh) 2018-10-19 2018-10-19 医疗数据异常识别方法、装置、终端及存储介质
CN201811226809.0 2018-10-19

Publications (1)

Publication Number Publication Date
WO2020078058A1 true WO2020078058A1 (fr) 2020-04-23

Family

ID=66066479

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/096631 WO2020078058A1 (fr) 2018-10-19 2019-07-19 Procédé et dispositif d'identification d'anomalies de données médicales, terminal et support de stockage

Country Status (2)

Country Link
CN (1) CN109636613A (fr)
WO (1) WO2020078058A1 (fr)

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109636613A (zh) * 2018-10-19 2019-04-16 平安医疗健康管理股份有限公司 医疗数据异常识别方法、装置、终端及存储介质
CN112305290B (zh) * 2019-07-25 2023-11-24 比亚迪半导体股份有限公司 电流检测装置
CN112447286B (zh) * 2019-09-05 2023-02-03 天津新开心生活科技有限公司 一种自动判别治疗线的方法以及装置
CN111242793B (zh) * 2020-01-16 2024-02-06 上海金仕达卫宁软件科技有限公司 医保数据异常的检测方法和装置
CN111445343A (zh) * 2020-03-23 2020-07-24 平安医疗健康管理股份有限公司 风险医疗数据的识别方法、装置、计算机设备和存储介质
CN111340641B (zh) * 2020-05-22 2020-11-13 浙江工业大学 一种异常就医行为检测方法
CN112445846A (zh) * 2020-12-01 2021-03-05 平安医疗健康管理股份有限公司 医疗项目识别方法、装置、设备及计算机可读存储介质
CN113130053A (zh) * 2021-03-24 2021-07-16 大光惠世医疗科技有限公司 一种基于用户活动数据的自助挂号方法及装置
CN113822365B (zh) * 2021-09-28 2023-09-05 北京恒生芸泰网络科技有限公司 基于区块链技术的医疗数据存储及大数据挖掘方法及系统
CN113838579A (zh) * 2021-09-29 2021-12-24 平安医疗健康管理股份有限公司 一种医疗数据的异常检测方法、装置、设备及存储介质
CN114912804A (zh) * 2022-05-17 2022-08-16 四川大学华西医院 一种科研数据相关性质控方法和系统
CN115267094A (zh) * 2022-09-29 2022-11-01 南通迈伦机电科技有限公司 一种废气排放监控检测方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103514576A (zh) * 2013-09-06 2014-01-15 深圳民太安信息技术有限公司 一种社保就诊违规套现的筛查方法
CN106408141A (zh) * 2015-07-28 2017-02-15 平安科技（深圳）有限公司 异常费用自动提取系统及方法
CN107194184A (zh) * 2017-05-31 2017-09-22 成都数联易康科技有限公司 基于时间序列相似性分析的在院人次异常检测方法及系统
CN107609980A (zh) * 2017-09-07 2018-01-19 平安医疗健康管理股份有限公司 医疗数据处理方法、装置、计算机设备及存储介质
CN109636613A (zh) * 2018-10-19 2019-04-16 平安医疗健康管理股份有限公司 医疗数据异常识别方法、装置、终端及存储介质

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5508912A (en) * 1989-01-23 1996-04-16 Barry Schneiderman Clinical database of classified out-patients for tracking primary care outcome
JP2006018532A (ja) * 2004-06-30 2006-01-19 Toshiba Corp 病院経営情報分析システム、病院経営情報分析方法、コンピュータ読み取り可能な記憶媒体及びコンピュータプログラム
CN101986333A (zh) * 2010-12-01 2011-03-16 福州维胜信息技术有限公司 一种医院辅助决策支持系统
CN106021967A (zh) * 2016-07-12 2016-10-12 上海市浦东新区周浦医院 用于临床路径质量评价的数据处理方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103514576A (zh) * 2013-09-06 2014-01-15 深圳民太安信息技术有限公司 一种社保就诊违规套现的筛查方法
CN106408141A (zh) * 2015-07-28 2017-02-15 平安科技（深圳）有限公司 异常费用自动提取系统及方法
CN107194184A (zh) * 2017-05-31 2017-09-22 成都数联易康科技有限公司 基于时间序列相似性分析的在院人次异常检测方法及系统
CN107609980A (zh) * 2017-09-07 2018-01-19 平安医疗健康管理股份有限公司 医疗数据处理方法、装置、计算机设备及存储介质
CN109636613A (zh) * 2018-10-19 2019-04-16 平安医疗健康管理股份有限公司 医疗数据异常识别方法、装置、终端及存储介质

Also Published As

Publication number Publication date
CN109636613A (zh) 2019-04-16

Similar Documents

Publication Publication Date Title
WO2020078058A1 (fr) Procédé et dispositif d'identification d'anomalies de données médicales, terminal et support de stockage
WO2020078053A1 (fr) Procédé, appareil et dispositif de détection d'anomalies de données médicales et support d'informations
WO2020119384A1 (fr) Procédé, appareil et dispositif de détection d'anomalie d'assurance médicale sur la base d'une analyse de mégadonnées, et support
WO2020113960A1 (fr) Procédé, appareil et dispositif de délivrance de médicaments sur ordonnance partagée, et support de stockage
WO2020113959A1 (fr) Procédé, appareil et dispositif d'évaluation de niveau de dialyse d'institution médicale et support d'informations
WO2020107756A1 (fr) Procédé anti-fraude sur crédit, système, dispositif et support d'informations lisible par ordinateur
WO2020233077A1 (fr) Procédé, dispositif et appareil de surveillance de service de système et support d'informations
WO2020087981A1 (fr) Procédé et appareil de génération de modèle d'audit de contrôle de risque, dispositif, et support de stockage lisible
WO2020119402A1 (fr) Procédé et appareil d'identification de médicaments inappropriés, terminal et support de stockage lisible par ordinateur
WO2020119403A1 (fr) Procédé, appareil et dispositif de détection d'anomalie de données d'hospitalisation, et support d'informations lisible
WO2020215681A1 (fr) Procédé et appareil de génération d'informations d'indication, terminal et support de stockage
WO2020119383A1 (fr) Procédé de supervision d'assurance médicale, dispositif, appareil et support de stockage lisible par ordinateur
WO2021246772A1 (fr) Procédé de distribution de tâches unitaires d'éléments fonctionnels d'un projet basé sur une externalisation ouverte pour générer des données d'apprentissage d'intelligence artificielle, appareil associé et programme informatique associé
WO2020108111A1 (fr) Procédé d'identification de comportement de fraude d'assurance médicale, appareil, dispositif et support de mémoire lisible
WO2018004236A1 (fr) Procédé et appareil de dépersonnalisation d'informations personnelles
WO2020087983A1 (fr) Procédé et appareil d'attribution de tâche, dispositif et support d'enregistrement
WO2021101105A2 (fr) Système et procédé de classification de sujets de matériaux de spécialité médicale
WO2020119115A1 (fr) Procédé, dispositif, appareil de vérification de données, et support de stockage
WO2020062641A1 (fr) Procédé d'identification d'un rôle d'utilisateur, équipement utilisateur, support d'informations et appareil d'identification d'un rôle d'utilisateur
WO2016165177A1 (fr) Procédé et système de paiement en libre-service d'hôpital en réseau reposant sur la reconnaissance faciale
WO2020119118A1 (fr) Procédé, appareil et dispositif de traitement de données anormales, et support de stockage lisible par ordinateur
WO2023106572A1 (fr) Procédé d'entraînement d'un modèle de détection permettant de détecter une négociation frauduleuse d'actifs virtuels, procédé de détection d'une négociation frauduleuse d'actifs virtuels utilisant le modèle de détection, dispositif et programme informatique conçus pour exécuter ces procédés
WO2021027143A1 (fr) Procédé, appareil et dispositif de poussée d'informations et support d'informations lisible par ordinateur
WO2018088585A1 (fr) Procédé de gestion de prise de médicament et dispositif associé
WO2023163248A1 (fr) Serveur, procédé et programme qui met en correspondance des ingrédients alimentaires sur la base de l'intelligence artificielle

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: 19873133

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205 DATED 11/06/2021)

122 Ep: pct application non-entry in european phase

Ref document number: 19873133

Country of ref document: EP

Kind code of ref document: A1