KR20180002229A - An agent apparatus for constructing database for dementia information and the operating method by using the same - Google Patents

An agent apparatus for constructing database for dementia information and the operating method by using the same Download PDF

Info

Publication number
KR20180002229A
KR20180002229A KR1020160081445A KR20160081445A KR20180002229A KR 20180002229 A KR20180002229 A KR 20180002229A KR 1020160081445 A KR1020160081445 A KR 1020160081445A KR 20160081445 A KR20160081445 A KR 20160081445A KR 20180002229 A KR20180002229 A KR 20180002229A
Authority
KR
South Korea
Prior art keywords
dementia
information
data
health information
database
Prior art date
Application number
KR1020160081445A
Other languages
Korean (ko)
Inventor
박병술
Original Assignee
원시스템주식회사
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 원시스템주식회사 filed Critical 원시스템주식회사
Priority to KR1020160081445A priority Critical patent/KR20180002229A/en
Publication of KR20180002229A publication Critical patent/KR20180002229A/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • 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
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/22—Social work or social welfare, e.g. community support activities or counselling services
    • 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
    • G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • General Health & Medical Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Public Health (AREA)
  • Primary Health Care (AREA)
  • Software Systems (AREA)
  • Physics & Mathematics (AREA)
  • Tourism & Hospitality (AREA)
  • Epidemiology (AREA)
  • General Physics & Mathematics (AREA)
  • Databases & Information Systems (AREA)
  • Child & Adolescent Psychology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Mathematical Physics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Pathology (AREA)
  • Biomedical Technology (AREA)
  • Artificial Intelligence (AREA)
  • Economics (AREA)
  • Human Resources & Organizations (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Medical Treatment And Welfare Office Work (AREA)

Abstract

The present invention relates to an agent device for building a dementia information database and an operating method thereof, which collect various types or formats of health information that are kept or newly generated from health information providers such as Health Insurance Review and Assessment Service, National Health Insurance Service, hospitals, drugstores, care centers, and individuals, extract information relating to dementia to build the same into a database, and, if a necessary field for early diagnosis and prognosis of dementia is omitted when the dementia-related information is converted into a database, fill the omitted necessary field by estimating (predicting) the same through learning processes using an artificial intelligence, thereby being used in a convergence technology for an early dementia diagnosis and prognosis on the basis of brain images and diverse biometric information built as a database. The agent device comprises: a data collecting unit for collecting health information from health information providers; a data classifying unit which classifies dementia-related information from the health information; and a control unit which converts the dementia-related information into a specific format to save the same in a database.

Description

TECHNICAL FIELD [0001] The present invention relates to an agent apparatus for establishing a dementia information database and an operation method thereof. [0002]

The present invention collects health information in a variety of types or formats existing or newly generated from a health information provider including a health insurance screening evaluation institution, a National Health Insurance Corporation, a hospital, a pharmacy, a medical institution, If some data on essential items for early diagnosis and prediction of dementia are missing when constructing the information related to dementia by extracting related information and building it into a database, data of essential items missing through learning process using artificial intelligence is estimated The present invention relates to an agent apparatus for constructing a dementia information database that can be utilized in fusion technology for early diagnosis and prediction of dementia, based on brain images and various biological information constructed in a database.

In recent years, with the development of medicine, the proportion of the elderly population has been increasing worldwide, and the rate of aging of Korea is progressing more rapidly than other advanced countries.

According to the Ministry of Health, Welfare and Family Affairs, the number of demented patients aged 65 or older is 8.4% of the total elderly population, 420,000, and in 2027, the elderly population with dementia Is expected to exceed 1 million people. Alzheimer's dementia fee for the elderly has increased 78.3% from 89.1 billion won in 2007 to 163.7 billion won in 2008, and it is increasing rapidly.

Dementia is a condition in which a normally developed brain is damaged or destroyed by external factors such as acquired trauma or illness, resulting in decreased cognitive functions such as memory, judgment, calculation ability, language ability, personality change, It is a syndrome that can not even be performed.

As for the diagnosis of dementia, it is not possible to detect early dementia with the dementia test, which is currently being performed in the public health centers of the whole country. There are several biomarkers candidates for cerebrospinal fluid (CSF) The reliability of the study was not verified because of its high rejection. It was not possible to predict early reliable dementia with only a fragmentary index such as brain imaging, blood test, and neuropsychological test. Recently, There is no systematic attempt of

As such conventional conventional dementia diagnosis methods are performed in a regular pattern according to each field, there is a limit, so various biological information and brain image analysis techniques (for example, MRI (magnetic resonance imaging), PET (Positron Emission Tomography, Positron Emission Tomography, DNA, Protein, etc.) are becoming increasingly necessary.

Therefore, it is imperative to analyze various biological information in a convergent manner and to prepare statistically meaningful fusion indexes and standards. In particular, since dementia is an irreversible disease, cure through medication is inevitable. However, if prediction and early detection are possible, it will be possible to delay and alleviate the onset of dementia, which will contribute to reducing socioeconomic burden and improving quality of life It is necessary to shift to a new paradigm based on computer technology for early dementia prediction and diagnosis.

In other words, the amount of information prior to bioinformatics was only about several tens of KB, but the size of brain image data increased from several MB to several hundred MB due to the development of CT and MRI equipment, and the genome information rapidly increased from MB to GB level, Big data era of brain data, genome, and protein has heterogeneous big data information and big data processing technology based on computation is needed to process them.

Accordingly, in the present invention, dementia-related information is extracted from health information made up of various types and formats collected or provided from a plurality of health information providers through an agent apparatus, and constructed as a database, and early diagnosis and prediction of dementia If some data on essential items are missing, we can estimate and fill missing essential data through learning processing using artificial intelligence, and apply them to fusion technology for early diagnosis and prediction of dementia. And suggest ways to perform early diagnosis and prediction of dementia while solving the dementia diagnosis problems.

Next, a brief description will be given of the prior arts that exist in the technical field of the present invention, and technical matters which the present invention intends to differentiate from the prior arts will be described.

Korean Patent Laid-Open Publication No. 2014-0022641 (Feb. 25, 2014) discloses a health logger for chronic disease management using intelligent agent technology. More specifically, the present invention relates to personal health devices such as a blood pressure monitor, a blood glucose meter, Health log service that can systematically manage chronic diseases based on the health information registered through the application, social web, and app connected to the device, and a service system and method that can help in the follow-up observation and clinical examination .

The prior art collects information such as biometric information, personal information, past history, etc. of various patients and provides an effect that the intelligent agent can automatically provide reliable comprehensive health status information at any time and anywhere using fuzzy inference do.

On the other hand, the present invention is not a technique of inferring information of various patients collected by the prior art using fuzzy theory in an agent, but rather a technique of providing various types or formats of information collected or provided from a health information provider through an agent apparatus The technology is a technology that extracts information about dementia from health information and constructs it as a database. More particularly, the present invention relates to a technique for estimating and filling missing essential data through learning processing using artificial intelligence when some data on items essential for early diagnosis and prediction of dementia are missing when constructing dementia-related information as a database The technical features are clear from the point of view of the composition.

Korean Published Patent Application No. 2012-0014585 (Feb. 17, 2012) further identifies the source, application or content of traffic in order to enable network operators and ISPs to establish and apply source specific policies within their networks. And to provide a method and a device that make it possible to do so.

The prior art provides the effect of identifying the source of traffic by a scout agent having a network connection and performing awareness of the service provider and the carrier of the application transmitted by these traffic sources over their networks.

On the other hand, the present invention collects health information in various types and formats from a health information provider, not a technology that identifies a source of traffic by an agent and recognizes a service provider and a network carrier by these traffic sources as in the prior art If some data on items essential for early diagnosis and prediction of dementia are missing when building a database, data of missing essential items are estimated through learning processing using artificial intelligence There is a technical difference in terms of filling technology.

In addition, Korean Patent No. 1038337 (June 1, 2011) uses ontology reasoning to construct indexing in various ways on a single web page to enable meaning searching for keywords inputted by the user, Based index method and a search engine using the index method.

The prior art provides an indexing unit configured with a gathering agent, an analysis agent, a vector analysis agent, and a filtering agent to construct indexing in various ways on a web page, thereby providing an effect of performing a quick search according to a user's intention.

On the other hand, the present invention is not a technology for indexing and constructing web pages in various ways as in the prior art, but rather extracting dementia-related information from various types or formats of health information collected or provided by a health information provider If some data on essential items for early diagnosis and prediction of dementia are missing, data can be estimated and filled in missing essential items through learning processing using artificial intelligence, so that it can be utilized in the fusion technology for early diagnosis and prediction of dementia Because of technology, the difference in technical characteristics is clear.

Disclosure of Invention Technical Problem [8] The present invention has been made to solve the above-mentioned problems, and it is an object of the present invention to collect health information of various types or formats that are already possessed or newly generated from various health information providers, The present invention aims to provide an agent apparatus and method for constructing a dementia information database that can be utilized for fusion technology for early diagnosis and prediction of dementia by constructing a database.

The present invention also relates to a method and system for diagnosing dementia, comprising the steps of: if a part of data essential to early diagnosis and prediction of dementia is missing when constructing the dementia-related information based on the health information collected from the health information provider, The data of the missing essential items are estimated and filled in and the estimated parts of the learning process are checked separately so that the dementia information that allows external experts (for example, doctors, researchers, etc.) to refer to the dementia- And an agent apparatus for establishing a database and a method of operating the agent apparatus.

An agent apparatus for constructing a dementia information database according to an embodiment of the present invention includes a data collection unit for collecting health information from a plurality of health information providers; A data classifier for classifying the dementia-related information from the health information using a predefined dementia-related item; And a controller for converting the classified dementia-related information into a specific format and storing the dementia-related information in a database. When the dementia-related information is stored in a database, if data of a specific item necessary for early diagnosis and prediction of dementia are missing, And the missing data is estimated and filled in.

The data classification unit may include: a confirmation unit for verifying details of the health information; A matching unit for matching each of the details of the confirmed health information with each of the predetermined dementia related items; An extracting unit for extracting a matching part based on the matching result; And an arranging unit for sorting the extracted parts according to each individual dementia-related item.

When the dementia-related information of various health information collected from the plurality of health information providers is stored in the database, if the data of some items among the items necessary for early diagnosis and prediction of dementia are missing for each user, And a learning processing unit for estimating and filling out the missing data through machine learning based on the control of the learning processing unit. The learning processing unit classifies a user having similar information from the entire database based on the dementation-related information of the user Extracting an item having missing data from the dementia-related information of the specific user by referring to the dementia-related information of each user clustered through the classification, and distributing the extracted data with respect to the specific item ≪ / RTI > and the variance < RTI ID = The average value of data corresponding to a specific item of the clustered users is calculated and filled in the missing position and if the threshold value is not less than the threshold value, And the like.

At this time, when estimating and applying some data of the missing essential items through the learning process, a check signal indicating that the data has been generated through the learning process is generated and applied together, and the data estimated by the learning process is applied to the dementia And a predetermined weight is applied for error correction when used for early diagnosis and prediction.

The agent apparatus includes a storage unit storing a control program used in the agent apparatus and storing dementia related items for extracting dementia related information from the health information collected from the respective health information providers; And a format conversion unit for converting the dementia related information in various formats classified by the data classification unit into a specific format used in the database through the control of the control unit.

And the health information is collected from the respective health information providers who are requested by the agent apparatus every predetermined period, or collected through direct input of each health information provider.

The control unit controls to extract the individual dementia-related information stored in the database based on a request from the communication terminal held by an external expert performing the dementia early diagnosis and prediction, and transmits the extracted individual dementia- Converted into a received format, and then output to the communication terminal through a data providing unit.

Further, an agent operating method for constructing a dementia information database according to an embodiment of the present invention includes: a data collecting step of collecting health information from a plurality of health information providers in a data collecting unit; A data classifying step of classifying the dementia-related information from the health information using a predefined dementia-related item in the data classification unit; And a database storing step of converting the classified dementia-related information into a specific format and storing the dementia-related information in a database in the control unit. In the database storing step, when the dementia-related information is stored in the database, And if the data of a specific item necessary is missing, the missing data is estimated and filled in.

The data classification step may include: a health information checking step of checking the details of the health information in the data classification unit; A matching step of matching each of the details of the health information checked in the health information checking step with each of the predetermined dementia related items; An extracting step of extracting a matching part based on a matching result performed in the matching step; And sorting the extracted part by classifying the extracted part according to each individual dementia-related item.

The agent operating method may further include a step of, when storing the dementia-related information among various health information collected from the plurality of health information providers at the control unit after the database storing step, Further comprising a learning processing step of estimating and filling out the missing data through machine learning based on the control of the control unit when data of the item is missing, Extracts items having missing data from the dementia-related information of the specific user by referring to the dementia-related information for each user clustered through the classification, extracts items with similar data from the entire database, Extraction for specific items Calculating a variance for the data and calculating an average value of data corresponding to a specific item of the clustered users if the variance is less than a preset threshold value and filling the missing position into the missing position; And calculates an intermediate value of the data corresponding to the missing data.

If a part of the missing essential items is estimated and applied through the learning process in the learning process step, a check signal indicating that the missing data is generated through the learning process is generated and applied to the learning process unit, A predetermined weight is applied for error correction when the data is utilized for early diagnosis and prediction of dementia.

And storing the dementia-related items for extracting the dementia-related information from the health information collected from the respective health information providers before the data collection step; And a format conversion step of converting the dementia related information in various formats classified in the data classification step to a specific format used in the database before the database storing step.

The health information collected from the plurality of health information providers in the data collection step may be collected from each of the health information providers requested by the control unit at predetermined intervals, Is collected.

The agent operating method may further include a step of extracting individual dementia-related information stored in the database based on a request of a communication terminal held by an external expert performing early diagnosis and prediction of dementia at the control unit after storing the database, Converting the dementia-related information into a format requested by the communication terminal, and outputting the converted dementia-related information to the communication terminal through the data providing unit.

As described above, according to the agent device and the method for operating the dementia information database of the present invention, after collecting the existing data and the health information of various types or formats newly generated from the health information provider, It is possible to collect, store and reuse data by using a database specialized for dementia-related information, thereby making it possible to maximize the synergy of early diagnosis of dementia and research on prediction fusion It is effective.

In addition, if some data on items essential for early diagnosis and prediction of dementia are missing when constructing information on each individual's dementia by a database, agent device estimates and fills missing essential data items by learning processing using artificial intelligence Since the estimated portion is separately checked, the physician or the researcher can refer to it to make the early diagnosis and prediction of dementia more accurately.

In addition, it is possible to analyze new and various dementia-specific information using database information in the early stage diagnosis and prediction technology of dementia, which is deviated from the stereotypes of medicine, biotechnology, and biology. Therefore, pattern normalization and fusion of brain image and multi- There is a possible effect.

In addition, based on the database construction of dementia-related information, it is possible to reduce national and social costs related to dementia prevention and dementia through development of a predictive dementia repair model, integration of computerized systems, and development of medical services.

1 is a conceptual diagram for explaining a dementia information database construction process to which the present invention is applied.
FIG. 2 is a diagram schematically showing the configuration of an agent apparatus for constructing a dementia information database according to an embodiment of the present invention.
FIG. 3 is a detailed view showing the configuration of the agent apparatus of FIG. 2. FIG.
4 is a diagram for explaining a data set of dementia-related information for each individual constructed in the database.
5 is a flowchart illustrating an operation procedure of an agent operating method for constructing a dementia information database according to an embodiment of the present invention in detail.
FIG. 6 is a flowchart illustrating an operation of classifying dementia-related information from the health information of FIG. 5 in more detail.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an agent apparatus and method for constructing a dementia information database according to the present invention will be described in detail with reference to the accompanying drawings. The present invention may be embodied in many different forms and is not limited to the embodiments described herein. Like parts are designated with like reference numerals throughout the specification.

1 is a conceptual diagram for explaining a dementia information database construction process to which the present invention is applied.

As shown in FIG. 1, in an agent apparatus for performing a database construction service for dementia-related information for use in early diagnosis and prediction of dementia, a health insurance screening evaluation center, a National Health Insurance Corporation, a hospital, a pharmacy, (1) to provide health information providers with health information including the existing data or newly generated health information.

(For example, a desktop PC, a notebook PC, a tablet, a smart phone, or the like) connected to a communication network through an agent apparatus, For example, patient history information related to various diseases such as dementia, various cancers, and heart diseases, examination information, medication information, image information, etc.).

The agent device collects existing health information data or health information in various types or formats that are newly generated from each health information provider requested to provide various health information (2).

At this time, in addition to the method of requesting and collecting the health information every predetermined period, the agent device provides health information from each health information provider that has previously provided a predetermined program for inputting health information Can receive. That is, the agent device is not collecting the health information provider after requesting the health information provider, and each health information provider can directly provide various health information to the agent device.

After collecting various kinds of health information from a plurality of health information providers (or after receiving health information directly from a plurality of health information providers), the agent apparatus transmits health information collected or provided from the health information provider to a predetermined dementia The information related to dementia is categorized in the health information by comparing the related items (for example, the keywords used in the symptoms and the dementia-related symptoms, the related words, etc.) (③). That is, after collecting or providing various health information from a plurality of health information providers, it is compared with predefined dementia related items to find information related to dementia among health information.

For example, the health information collected from each health information provider includes a variety of information not related to dementia, such as information on various diseases such as cancer, heart disease, Therefore, the agent device performs a task of separately extracting data related to dementia from the large-sized health information big data.

After classification of the dementia-related information, the agent device converts the selected dementia-related information from various health information collected or provided by the health information provider into a specific format used in the database, ④). That is, since each type of application program used by each health information provider is various and file format is different, the agent device converts the dementia-related information into a file form and stores it in a database .

At this time, the specific dementia related information stored in the database may include the member information (ID, authority / role, name, resident registration number, telephone number, address, affiliation, registration date, , Diagnosis information (ID, subject ID, etc.), subject information (ID, name, date of diagnosis, resident registration number, phone number, address, affiliation, registration date, (Eg ID, height, weight, chest, blood pressure, hearing, vision, etc.), family history information (ID, family name, family relationship, SNSB information (ID, attention ability, language ability, time and space function, memory ability, comprehensive evaluation, etc.), clinical diagnosis information (ID, intelligence test, developmental test, general psychological examination , Learning ability assessment, attention test, intelligence test, , Parietal lobe, parietal lobe, temporal lobe, occipital lobe, cerebellum, brainstem, ganglia, thalamus, hypothalamus, pituitary gland, pineal gland, midbrain, (File code, description, original file name, storage file name, file path, file size, file type, file registration date, registrant, etc.) ID, date of registration, etc.), result information (ID, file code, file order, dementia type, etc.).

On the other hand, when constructing the dementia-related information as a database, the agent device estimates the missing data through machine learning when some items of the items necessary for early diagnosis and prediction of dementia are missing for each user Fill it up (⑤). In this process, users who have similar information are classified and extracted from the entire database based on the dementation-related information of the user. The above classification refers to a process of clustering users having similar conditions based on information related to dementia of each individual composed of time series (age) through machine learning.

With reference to the dementia-related information for each user clustered through the classification, items with missing data are extracted from the dementation-related information of the specific user. Calculating a variance of the extracted data with respect to the specific item and calculating an average value of data corresponding to a specific item of the clustered users if the variance is less than a preset threshold value, If the threshold value is not less than the threshold value, a median of data corresponding to a specific item of the clustered users is calculated and filled in the missing position.

For example, missing data for a particular item of a particular user may be used to clustering similar user groups from all users using the missing information of a particular user, And estimates the data.

More specifically, in the process of building dementia-related information into a database, when a family member, which is one of the essential items of the information of a specific individual, is missing, the agent device performs a learning process using artificial intelligence, And then fill in the missing item data.

At this time, when the agent device estimates and applies some data of the missing essential items through the learning process, it is preferable to separately display a check signal indicating that the missing data is generated through the learning process, When used for diagnosis and prediction, a method of applying a predetermined weight to the estimated data may be used to prevent diagnosis and prediction errors due to the estimated information.

After collecting or providing various health information from a plurality of health information providers, only the dementia-related information is separately extracted and constructed as a database. Then, the agent apparatus acquires a specific person (doctor or researcher performing dementia diagnosis and prediction) (6), the information related to dementia related to the individual is extracted from the database and converted into the format requested by the external expert.

Accordingly, it is very easy to collect, store and reuse data using a database specialized for dementia-related information, and can perform early diagnosis and prediction of dementia more precisely, as well as early diagnosis and prediction of dementia in medicine, biotechnology, biology , It is possible to analyze new and various dementia-specific information using database information, and it is possible to alleviate national and social costs related to dementia prevention and dementia based on the database construction of dementia-related information.

FIG. 2 is a view schematically showing the configuration of an agent apparatus for constructing a dementia information database according to an embodiment of the present invention. FIG. 3 is a detailed view of the configuration of the agent apparatus of FIG. 2, And a data set of dementia-related information for each individual to be constructed.

2, the apparatus of the present invention comprises a network 100, a plurality of health information providers 200, an agent apparatus 300, a database 400, and the like.

The network 100 is a variety of communication networks including a wired / wireless Internet and the like, and is connected to a communication line between a plurality of health information providers 200 and an agent apparatus 300 to process mutual data transmission and reception related to health information.

The health information provider 200 may be a health insurance screening and assessment institution, a national health insurance corporation, a hospital, a pharmacy, a medical institution, an individual user, etc. that holds and manages various kinds of health information, After providing the communication connection, the agent device 300 provides various health information, which is held or newly generated at the request of the agent device 300, to the agent device 300.

At this time, the health information provider 200 can provide various health information that he / she holds after proceeding to establish communication connection to the agent apparatus 300 directly via the network 100 without a separate request of the agent apparatus 300 have.

The agent apparatus 300 collects various health information that is held or newly generated from a plurality of health information providers 200, extracts dementia-related information from the health information, and stores the information in the database 400. If some data on the items necessary for early diagnosis and prediction of dementia are missing when storing the dementia-related information in the database 400, the function of estimating and filling the missing essential items through the learning process using the artificial intelligence is added . The dementia-related information of a specific individual stored in the database 400 is extracted based on a request from an external expert such as a doctor or a researcher performing diagnosis and prediction of dementia, and converted into a format requested by an external expert.

3, the agent apparatus 300 includes a storage unit 310, a data collection unit 320, a data classification unit 330, a format conversion unit 340, a learning processing unit 350, A studying unit 360, a control unit 370, and the like.

The storage unit 310 stores a control program used in the agent apparatus 300 and stores dementia related items for extracting dementia related information from the health information collected from each health information provider 200. [

At this time, the dementia-related item is defined in advance to cull the dementia-related information from various health information collected from a plurality of health information providers in the agent apparatus 300, and is preferably in a general text form, but is not limited thereto, It is needless to say that an image form may be included.

The data collection unit 320 collects health information from a plurality of health information providers 200. At this time, the health information is collected from each health information provider 200 requested by the agent apparatus 300 every predetermined period, or collected through the direct input of each health information provider 200.

The data classification unit 330 performs classification of dementia related information from various health information collected by the data collection unit 320 using a predefined dementia related item.

The data classification unit 330 includes a verification unit 332 that reads various health information collected from a plurality of health information providers 200 and confirms the details of the health information, A matching unit 334 for matching each of the predefined dementia related items stored in the storage unit 310 with each other and a matching unit 334 for matching the predefined dementia- An extracting unit 336 for extracting a matching part and an arranging unit 338 for sorting and sorting the parts extracted by the extracting unit 336 according to each individual dementia related item.

The format converting unit 340 converts the dementia related information in various formats classified by the data classifying unit 330 into a specific format used in the database 400 through the control of the controller 370. [

In addition, when the dementia-related information of a specific individual is requested from an external expert (that is, a doctor or a researcher performing dementia diagnosis and prediction) through the data provider 360, the format conversion unit 340 converts the dementia- And then converts the extracted individual dementation-related information into a specific format requested by an external expert.

That is, the format conversion unit 340 converts the dementia-related information extracted from the various health information collected from the plurality of health information providers into the form of N: 1, stores the information in the database 400, The related information is converted into 1: N format and provided to external experts.

The learning processing unit 350 may be configured such that when storing the dementia related information among the various health information collected from the plurality of health information providers 200 in the database 400, The controller 370 performs a function of estimating and filling the missing data through machine learning based on the control of the controller 370.

That is, the learning processing unit 350 first classifies a user having similar information from the entire database based on the dementation-related information of the user (that is, the dementia-related information of each individual constituted by the time series Extracting items with missing data from the dementation-related information of the specific user by referring to the dementia-related information for each user clustered through the classification. If the variance is less than a preset threshold value, an average value of data corresponding to a specific item of the clustered users is calculated and filled in the missing position, A middle value of data corresponding to a specific item of the clustered users is calculated and filled in the missing position.

The process of estimating and filling out the missing item data will be described in more detail with reference to FIG. 4. In the process of building the information on the dementia of the user A in the seventies of age as a database, a specific item? (For example, Data, etc.) of the 40th, 50th, and 70th data of the user are listed and if the data of 60th data is missing, the learning processing unit 350 extracts the average value or the intermediate value The calculation can simply fill in the missing 60 pieces of data of a specific item α.

If the 50 pieces of data of a specific item? (For example, a brain image item, a cognitive ability test item, and the like) are missing in the process of building the dementia related information of the user B in the database, the learning processing unit 350, The user A and the user N having similar conditions to the user B and extracts the data portion of the 50 items of the missing item? From the dementation-related information of the user B. Then, 50 missing data of the specific item? Of the user B is estimated and filled in from the 50 items data extracted from the items related to the dementia of the users A and N by calculating the average value or the median value.

In this case, when the learning processing unit 350 estimates and applies some data of the missing essential items through the learning process, it can generate and generate a check signal indicating that it is generated through the learning process. When data estimated through the learning process is used for early diagnosis and prediction of dementia, it is preferable to implement a predetermined weight for preventing errors. For example, assuming that there are 20 required items, if there are five items missing, the learning processing unit 350 performs a learning process based on machine learning and artificial intelligence to estimate data of five missing items, And filling in the data of the estimated essential items.

The data providing unit 360 transmits a request signal of a dementia related information of a specific individual input from an external expert performing early diagnosis and prediction of dementia to the control unit 370. Based on the control signal from the control unit 370, And provides the stored dementia-related information to a corresponding external expert.

The control unit 370 controls the collection of various types of health information from a plurality of health information providers through the data collection unit 320 and receives various health information collected from the data collection unit 320 through the data classification unit 330, And controls the classification unit 330 to convert the classified dementia-related information into a specific format through the format conversion unit 340 and store it in the database 400.

The control unit 370 controls to extract the individual dementia-related information stored in the database 400 based on the request of the communication terminal held by the external expert performing the dementia early diagnosis and prediction, and outputs the extracted individual dementia- Converts the data into a format requested by an external expert through the data providing unit 340, and controls the data providing unit 360 to output the data to a communication terminal held by an external expert.

Next, an agent operating method for constructing the dementia information database according to the present invention will be described in detail with reference to FIGS. 5 and 6. FIG. At this time, each step according to the method of the present invention may be changed in the use environment or the order by a person skilled in the art.

FIG. 5 is a flowchart illustrating in detail an operation process of an agent operating method for building a dementia information database according to an embodiment of the present invention. FIG. 6 is a flowchart illustrating an operation of classifying dementia- Fig.

First, various kinds of information collected from a health information provider such as the National Health Insurance Corporation, a hospital, a pharmacy, a medical institution, an individual, etc. in an agent apparatus 300 for building dementia-related information for use in early diagnosis and prediction of dementia in the database 400 Related items used for extracting dementia-related information for use in early diagnosis and prediction of dementia from health information are set and stored (S100).

The agent apparatus 300 that has set the dementia-related item through step S100 is provided to the health information provider 200 including the health insurance screening evaluation institution, the National Health Insurance Corporation, the hospital, the pharmacy, the medical institution, Data or various health information to be newly generated (S200). That is, the agent device 300 includes the patient history information, the examination information, the medication information, the image information, and the like related to various diseases such as dementia, various cancers and heart diseases, etc. to each of the health information providers 200 through the network 100 To provide health information.

The data collecting unit 310 of the agent apparatus 300 receives the health information data or the health information data from the respective health information providers 200 requested to provide various health information, Health information in a variety of types and formats that are newly generated is collected (S300).

At this time, the agent apparatus 300 requests the health information providers 200 to provide health information every predetermined period, collects health information, or receives health information directly from each health information provider 200 have.

When the health information is collected from each health information provider 200 through the step S300, the data classification unit 320 of the agent apparatus 300 uses the dementia-related items defined in advance through the step S100, 200) from the health information (S400).

5, the data classifier 310 reads the health information collected from the plurality of health information providers 200 and confirms the details of the health information (S410). In operation S410, (Step S420), and the details of the health information identified in step S100 are matched with the dementia-related items defined in advance in step S100.

For example, read health information in a file format such as * .dat, * .html, read it in a temporary storage space, and check whether the content of each part matches a predefined dementia related item .

If there is a matching part based on the matching result in step S420, the data classifying unit 310 extracts a portion matching the dementia-related item (S430), and classifies and arranges the extracted portion according to each individual dementia-related item (S440). That is, the dementia-related information is collected from the health information collected from a plurality of health information providers, and the database is prepared for storage.

After classifying the dementia-related information from the health information through step S400, the controller 370 of the agent apparatus 300 converts the dementia-related information classified in step S400 into a specific format and stores it in the database 400 (step S500) . That is, format conversion is performed to convert the dementia-related information in various formats classified in step S400 into a specific format used in the database 400, and stores the format conversion.

After the dementia-related information is classified from the health information collected from the plurality of health information providers through step S500 and stored in the database 400, the control unit 370 of the agent apparatus 300 determines whether the dementia- And if the data of some items are missing, the learning processing unit 350 performs a learning process using the artificial intelligence (i.e., a machine learning process) (S600). ≪ / RTI >

That is, the learning processing unit 350 classifies and extracts users having similar information from the entire database based on the dementation-related information of the user, and refers to the dementia-related information for each user clustered through the classification, If the variance is less than a preset threshold value, the average value of data corresponding to a specific item of the clustered users is calculated If the threshold value is not less than the threshold value, an intermediate value of data corresponding to a specific item of the clustered users is calculated and filled in the missing position.

In this case, when some data of the missing essential items are estimated and applied through the learning process in step S600, the control unit 370 separately transmits a check signal for informing that data of the item is generated through the learning process of the learning process unit 350 And apply it. In order to prevent errors that may occur when the data estimated through the learning process is used for early diagnosis and prediction of dementia, it is preferable to implement weighting to the estimated information. In other words, when the data estimated through the learning process is used for early diagnosis and prediction of dementia, weights are used to suppress the occurrence of errors as much as possible.

In addition, after the dementia-related information is classified from the health information collected from the plurality of health information providers through step S500 and stored in the database 400, the control unit 370 of the agent apparatus 300 performs diagnosis and prediction of dementia When the dementia-related information of a specific individual is requested from a doctor or a researcher, the extracted dementia-related information is converted into a format requested by an external expert, and is provided to the corresponding communication terminal through the data providing unit 360 (S700 ). In other words, according to the request of an external expert, information on the individual's dementia can be provided so that early diagnosis and prediction of dementia can be performed.

As described above, since the dementia-related information is extracted from the health information of various types or formats collected or provided from the health information provider and constructed as a database, it is easy to collect, store and reuse data specialized for the dementation-related information , The synergy of early diagnosis of dementia and prediction fusion research can be maximized.

In addition, even if some data on items necessary for early diagnosis and prediction of dementia are missing during the database construction of each individual's dementia-related information, the data of missing items are estimated and filled through learning processing using artificial intelligence, The diagnosis and prediction of dementia can be performed more accurately.

In addition, since it is possible to analyze new and various dementia-specific information using database information, it is possible to normalize and converge patterns of brain image and multi-species biological information, away from stereotyped methods such as existing medicine, biotechnology and biology.

In addition, through the database-based information on dementia, it is possible to continuously develop the dementia forecasting repair model, the integration of the computer system, and the development of the medical service, thereby alleviating the national and social costs of dementia prevention and dementia.

It will be apparent to those skilled in the art that various modifications may be made to the invention without departing from the spirit and scope of the invention as defined in the following claims And changes may be made without departing from the spirit and scope of the invention.

100: network 200: health information provider
300: agent apparatus 310: storage unit
320: data collecting unit 330:
332: Confirmation unit 334:
336: Extracting unit 338:
340: format conversion unit 350:
360: data providing unit 370:
400: Database

Claims (14)

A data collection unit for collecting health information from a plurality of health information providers;
A data classifier for classifying the dementia-related information from the health information using a predefined dementia-related item; And
And a controller for converting the classified dementia related information into a specific format and storing the converted dementia related information in a database,
Wherein when the dementia-related information is stored in the database, the missing data is estimated and filled in if the data of a specific item necessary for early diagnosis and prediction of dementia are missing. The method according to claim 1,
Wherein the data classification unit comprises:
A confirmation unit for confirming details of the health information;
A matching unit for matching each of the details of the confirmed health information with each of the predetermined dementia related items;
An extracting unit for extracting a matching part based on the matching result; And
And an arranging unit for sorting the extracted parts according to the individual dementia-related items and arranging the sorted parts. The method according to claim 1,
The agent apparatus comprising:
When storing data on dementia among various kinds of health information collected from the plurality of health information providers in a database, if data of some items among items necessary for early diagnosis and prediction of dementia are missing for each user, machine learning and a learning processor for estimating and filling the missing data through machine learning,
The learning processing unit classifies and extracts users having similar information from the entire database based on the dementation-related information of the user, refers to the dementation-related information for each user clustered through the classification, The average value of the data corresponding to the specific item of the clustered users is calculated if the variance is less than a preset threshold value, And if the threshold value is not less than the threshold, calculating an intermediate value of data corresponding to a specific item of the clustered users and filling the missing position into the missing position. The method of claim 3,
The learning processing unit,
When a part of the missing essential items is estimated and applied through a learning process, a check signal indicating that the data is generated through a learning process is generated and applied together,
And when the data estimated through the learning process is used for early diagnosis and prediction of dementia, the predetermined weight is applied for error correction. The method according to claim 1,
The agent apparatus comprising:
A storage unit storing a control program used in the agent apparatus and storing dementia related items for extracting dementia related information from the health information collected from each health information provider; And
And a format conversion unit for converting the dementia-related information in various formats classified by the data classification unit into a specific format used in the database through the control of the control unit. Device. The method according to claim 1,
The health information includes:
Collected from each health information provider who has been requested by the agent apparatus every predetermined period,
Wherein the agent information is collected through direct input of each health information provider. The method according to claim 1,
Wherein,
Related information stored in the database on the basis of a request of a communication terminal held by an external expert performing the dementia early diagnosis and prediction,
Wherein the control unit controls the communication terminal to convert the extracted individual dementia-related information into a format requested by the communication terminal, and then output the converted dementia-related information to the communication terminal through a data providing unit. A data collection step of collecting health information from a plurality of health information providers in a data collection unit;
A data classifying step of classifying the dementia-related information from the health information using a predefined dementia-related item in the data classification unit; And
And a database storing step of converting the classified dementia related information into a specific format and storing the converted dementia related information in a database,
Wherein when the dementia-related information is stored in the database in the database storing step, the missing data is estimated and filled in if the data of a specific item necessary for early diagnosis and prediction of dementia are missing. Agent operating method. The method of claim 8,
Wherein the data classification step comprises:
A health information checking step of checking the details of the health information in the data classification unit;
A matching step of matching each of the details of the health information checked in the health information checking step with each of the predetermined dementia related items;
An extracting step of extracting a matching part based on a matching result performed in the matching step; And
And sorting the extracted part according to each individual dementia-related item and arranging the divided part. The method of claim 8,
The agent operating method includes:
When the control unit stores the dementia-related information among the various health information collected from the plurality of health information providers in the database, if the data of some items among the items necessary for early diagnosis and prediction of dementia for each user are missing, And a learning processing step of estimating and filling the missing data through machine learning based on the control of the controller,
In the learning processing step, users having similar information are classified and extracted from the entire database based on the user's dementia-related information, and referring to the dementia-related information for each user clustered through the classification, Calculating an average value of data corresponding to a specific item of the clustered users when the variance is less than a preset threshold value and calculating a variance of the extracted data with respect to the missing data, And calculating an intermediate value of data corresponding to a specific item of the clustered users if the threshold value is not less than a threshold value, and filling the missing position with the intermediate value. The method of claim 10,
In the learning processing step,
When a part of the missing essential items is estimated and applied through a learning process, a check signal indicating that the data has been generated through a learning process is generated and applied to the learning processing unit,
Wherein when the estimated information is used for early diagnosis and prediction of dementia, the predetermined weight is applied for error correction. The method of claim 8,
The agent operating method includes:
Storing the dementia-related items for extracting the dementia-related information from the health information collected from the respective health information providers before the data collection step; And
Further comprising a format conversion step of converting the dementia related information in various formats classified in the data classification step to a specific format used in the database before the database storing step Agent operating method. The method of claim 8,
Wherein the health information collected from the plurality of health information providers in the data collection step includes:
Collected from each of the health information providers requested by the control unit every predetermined period,
Wherein the agent is collected through direct input of each health information provider. The method of claim 8,
The agent operating method includes:
The control unit extracts the individual dementia-related information stored in the database on the basis of the request of the communication terminal held by the external expert performing the dementia early diagnosis and prediction, and transmits the extracted individual dementia- And outputting the converted format to the communication terminal through the data providing unit. The method of claim 1, further comprising:
KR1020160081445A 2016-06-29 2016-06-29 An agent apparatus for constructing database for dementia information and the operating method by using the same KR20180002229A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
KR1020160081445A KR20180002229A (en) 2016-06-29 2016-06-29 An agent apparatus for constructing database for dementia information and the operating method by using the same

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
KR1020160081445A KR20180002229A (en) 2016-06-29 2016-06-29 An agent apparatus for constructing database for dementia information and the operating method by using the same

Publications (1)

Publication Number Publication Date
KR20180002229A true KR20180002229A (en) 2018-01-08

Family

ID=61003650

Family Applications (1)

Application Number Title Priority Date Filing Date
KR1020160081445A KR20180002229A (en) 2016-06-29 2016-06-29 An agent apparatus for constructing database for dementia information and the operating method by using the same

Country Status (1)

Country Link
KR (1) KR20180002229A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101881731B1 (en) * 2018-02-27 2018-07-25 한국과학기술정보연구원 Apparatus and Method for Dementia Prediction using Machine Learning, and Storage Medium Storing the Same
CN111352966A (en) * 2020-02-24 2020-06-30 交通运输部水运科学研究所 Data tag calibration method in autonomous navigation
WO2020179950A1 (en) * 2019-03-06 2020-09-10 주식회사 인포메디텍 Deep learning-based method and device for prediction of progression of brain disease

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101881731B1 (en) * 2018-02-27 2018-07-25 한국과학기술정보연구원 Apparatus and Method for Dementia Prediction using Machine Learning, and Storage Medium Storing the Same
WO2020179950A1 (en) * 2019-03-06 2020-09-10 주식회사 인포메디텍 Deep learning-based method and device for prediction of progression of brain disease
CN111352966A (en) * 2020-02-24 2020-06-30 交通运输部水运科学研究所 Data tag calibration method in autonomous navigation

Similar Documents

Publication Publication Date Title
KR101873926B1 (en) Method for providing medical counseling service between insurance organization and specialist based on bigdata
CN112700838B (en) Big data-based medication scheme recommendation method and device and related equipment
KR102558021B1 (en) A clinical decision support ensemble system and the clinical decision support method by using the same
KR20180002234A (en) A smart examination apparatus for dementia early diagnosis and the method by using the same
CN107491630A (en) Clinical decision support integrated system and use its clinical decision support method
WO2019070200A1 (en) A healthcare information management system
Shenas et al. Identifying high-cost patients using data mining techniques and a small set of non-trivial attributes
US7505867B2 (en) System and method for predicting medical condition
US20220172841A1 (en) Methods of identifying individuals at risk of developing a specific chronic disease
US20040172287A1 (en) Method and apparatus for obtaining and distributing healthcare information
KR20200128241A (en) Method for providing osteoporosis diagnosis and treatment service based on bigdata and artificial intelligence
WO2022211385A1 (en) Health care consultation system using distribution of disease prediction values
CN112216361A (en) Follow-up plan list generation method, device, terminal and medium based on artificial intelligence
CN113066562A (en) Medical image transmission method and system based on 5g
US20230170069A1 (en) Integrated, ai-enabled value-based care measurement and objective risk assessment clinical and financial management system
Moon et al. Artificial neural networks in neurorehabilitation: A scoping review
KR20180002229A (en) An agent apparatus for constructing database for dementia information and the operating method by using the same
CN112908452A (en) Event data modeling
JP2004185192A (en) Comprehensive support system for analysis/research of health information and health maintenance/longevity realization
Kataria et al. Harnessing of real-world data and real-world evidence using digital tools: utility and potential models in rheumatology practice
Gujral et al. Artificial intelligence (AI) and data science for developing intelligent health informatics systems
CN117316392A (en) Medical image information management system and method based on intelligent medical treatment
WO2023240837A1 (en) Service package generation method, apparatus and device based on patient data, and storage medium
KR101744800B1 (en) System for providing medical information
JP2000099605A (en) Health examination information analysis server and network health information system

Legal Events

Date Code Title Description
A201 Request for examination
E902 Notification of reason for refusal
E601 Decision to refuse application