KR102544813B1 - Integrated Mediation System And Method For Multidisciplinary Care - Google Patents

Abstract

An integrated mediation method for medical treatment according to an embodiment of the present invention includes the steps of obtaining, by a mediation server, basic data including a personal health record of a user from an original collection group including a plurality of sources; generating converted data by the mediation server by converting the basic data into a target format of input data that can be processed by a target artificial intelligence model; and generating, by the mediation server, diagnostic data based on output data generated by a diagnostic server using the converted data as input data of the target artificial intelligence model, wherein the diagnostic data includes a plurality of artificial intelligence models It is generated based on the output data generated using the target artificial intelligence model selected from among.

Description

Integrated Mediation System And Method For Multidisciplinary Care

The present invention relates to an integrated mediation system for medical treatment and a mediation method using the same, and more particularly, to an integrated mediation system for medical treatment using a plurality of artificial intelligence models and a mediation method using the same.

Recently, artificial intelligence, deep learning-based medical diagnosis/prediction field, personal health record (PHR) or electronic medical record (EMR) acquired from individuals or hospitals, etc. The field of diagnosis/prediction is attracting attention.

Among them, as part of a method for increasing the accuracy of the medical diagnosis, there is a method of performing diagnosis based on a plurality of medical/health data. Linkage and integration between medical/health data from different sources is an important element in order to obtain meaningful diagnostic data in a method of treatment using data obtained from such multiple routes.
[Prior literature]
Republic of Korea Patent Publication No. 10-2017-0081520 (2017.07.12)

Embodiments of the present invention mediate a diagnosis server including an original collection group including a plurality of sources and a plurality of artificial intelligence models as a mediation server, thereby improving the accuracy of prediction/diagnosis for a specific disease based on medical records from different sources. It is intended to provide an integrated mediation system and method for improved medical treatment.

An integrated mediation method for medical treatment according to an embodiment of the present invention includes the steps of obtaining, by a mediation server, basic data including a personal health record of a user from an original collection group including a plurality of sources; generating converted data by the mediation server by converting the basic data into a target format of input data that can be processed by a target artificial intelligence model; and generating, by the mediation server, diagnostic data based on output data generated by a diagnostic server using the converted data as input data of the target artificial intelligence model, wherein the diagnostic data includes a plurality of artificial intelligence models It is generated based on the output data generated using the target artificial intelligence model selected from among.

The generating of the converted data may include analyzing items of the basic data and extracting item data; analyzing minimum parameter information included in the target format; and matching the item data and the minimum parameter information corresponding thereto and generating the converted data based on the matching result.

The item may include at least one of medical images, EMR data, biometric information, health data, and time series data, and the item data may include the item and parameter information constituting each item.

The basic data may include original data including the personal health record and a request signal for selecting a target artificial intelligence model into which the original data is to be input.

Generating the converted data may include identifying the target artificial intelligence model corresponding to the request signal; extracting item data by analyzing items of the basic data; analyzing minimum parameter information included in the target format; and matching the item data and the minimum parameter information corresponding thereto, and generating the conversion data based on at least one of the identification result and the matching result.

The request signal includes an individual request signal of each of the plurality of sources, and the mediation server may independently receive the individual request signal from each of the plurality of sources.

The generating of the converted data may include standardizing a data structure format of the basic data for each of the plurality of sources; and a noise processing step of removing noise including information unrelated to the minimum parameter of the target format with respect to the basic data for each of the plurality of sources.

Generating the diagnostic data may include generating an identification UI capable of identifying a source of the output data; Matching phenotypes for each of the output data; and providing a comprehensive diagnosis certificate by integrating the output data matching the phenotype and the identification UI corresponding thereto.

An integrated mediation system for medical treatment according to an embodiment of the present invention includes a basic data acquisition unit that acquires basic data including a personal health record of a user from an original collection group including a plurality of sources; a conversion data generating unit generating converted data by converting the basic data into a target format of input data that can be processed by a target artificial intelligence model; and a diagnostic data generator configured to generate diagnostic data based on output data generated by a diagnostic server using the converted data as input data of the target artificial intelligence model, wherein the diagnostic data is selected from among a plurality of artificial intelligence models. It is generated based on the output data generated using the selected target artificial intelligence model.

The conversion data generator may include: an item data extractor configured to extract item data by analyzing items of the basic data; a minimum parameter analysis unit analyzing minimum parameter information constituting the target format; and a matching generation unit that matches the item data and the minimum parameter information corresponding thereto and generates the converted data based on the matching result.

The item may include at least one of medical images, EMR data, biometric information, health data, and time series data, and the item data may include the item and parameter information constituting each item.

The basic data may include original data including the personal health record and a request signal for selecting a target artificial intelligence model into which the original data is to be input.

The converted data generation unit may include: a target artificial intelligence model identification unit identifying the target artificial intelligence model corresponding to the request signal; an item data extraction unit configured to extract item data by analyzing items of the basic data; a minimum parameter analysis unit analyzing minimum parameter information constituting the target format; and a matching generating unit that matches the item data and the minimum parameter information corresponding thereto and generates the converted data based on at least one of the identification result and the matching result.

The request signal may include an individual request signal of each of the plurality of sources, and the basic data acquisition unit may independently receive the individual request signal from each of the plurality of sources.

The converted data generation unit may include: a standardization unit standardizing a data format of the basic data for each of the plurality of sources; and a noise processing unit that removes noise including information unrelated to the minimum parameter of the target format with respect to the basic data for each of the plurality of sources.

The diagnostic data generation unit generates an identification UI capable of identifying the source of the output data, matches phenotypes for each of the output data, and synthesizes the output data matched with the phenotype and the identification UI corresponding thereto. A comprehensive medical certificate can be provided.

According to the integrated mediation system and method according to the embodiments of the present invention, personal medical/health from different sources is provided by mediating an original collection group including a plurality of sources and a diagnosis server including a plurality of artificial intelligence models as a mediation server. The accuracy and efficiency of prediction and diagnosis of a specific disease based on records can be improved.

1 is a diagram schematically showing the configuration of an integrated mediation system according to an embodiment of the present invention.
Figure 2 is a block diagram schematically showing the configuration of the integrated mediation system according to an embodiment of the present invention.
Figure 3 is a configuration diagram schematically showing the configuration of the processor of the mediation server according to an embodiment of the present invention.
Figure 4 is a flow chart for explaining the integration mediation method according to an embodiment of the present invention.
Figure 5 is a flow chart for explaining one step of the integration mediation method according to an embodiment of the present invention in more detail.
Figure 6 is a system diagram for explaining in more detail one step of the integrated mediation method according to an embodiment of the present invention.

Since the present invention can apply various transformations and have various embodiments, specific embodiments will be illustrated in the drawings and described in detail in the detailed description. Effects and features of the present invention, and methods for achieving them will become clear with reference to the embodiments described later in detail together with the drawings. However, the present invention is not limited to the embodiments disclosed below and may be implemented in various forms.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings, and when describing with reference to the drawings, the same or corresponding components are assigned the same reference numerals, and overlapping descriptions thereof will be omitted. .

In the following embodiments, terms such as first and second are used for the purpose of distinguishing one component from another component without limiting meaning. In the following examples, expressions in the singular number include plural expressions unless the context clearly dictates otherwise. In the following embodiments, terms such as include or have mean that features or components described in the specification exist, and do not preclude the possibility that one or more other features or components may be added. In the drawings, the size of components may be exaggerated or reduced for convenience of description. For example, since the size and shape of each component shown in the drawings are arbitrarily shown for convenience of description, the present invention is not necessarily limited to those shown.

1 is a diagram schematically showing the configuration of an integrated mediation system according to an embodiment of the present invention. An integrated mediation system according to the present invention is an integrated mediation system for medical treatment with improved accuracy and efficiency. The integrated mediation system 10 may include a mediation server 100, an original collection group 200, and a diagnosis server 300. can

The original collection group 200 includes a plurality of sources 200-1, 200-2, ..., 200-m; 200, and the integrated mediation system 10 transmits personal health records for each user/patient from the plurality of sources. ) can be provided. Each of the plurality of sources 200-1, 200-2, ..., 200-m may mean various institutions such as hospitals that provide medical treatment and diagnosis records to users, and according to an embodiment, personal health records are provided. It may also include an individual group including an individual's vehicle, a wearable device worn by the individual, and the like. The plurality of sources 200 may communicate with other sources 200, the mediation server 100, and the diagnosis server stage 300 through a communication network using a wireless or wired communication method. Phenotypes of personal health records provided by a plurality of sources may be different, and this will be described in detail in related figures to be described later. Hereinafter, a personal health record may be referred to as a concept including both medical data obtained from a hospital and health data obtained from an individual.

The diagnosis server unit 300 may include a plurality of artificial intelligence models, including a plurality of servers 300-1, 300-2, ..., 300-n having artificial intelligence models. The diagnosis server 300 may predict/diagnose a prognosis for a specific disease of a user based on a plurality of personal health records acquired from the original collection group 200 through the mediation server 100 . At this time, each of the plurality of servers 300-1, 300-2, ..., 300-n may include at least one artificial intelligence model, and the plurality of artificial intelligence models included in the diagnosis server unit 300 are of the same type. disease can be predicted and diagnosed. Depending on the embodiment, a plurality of artificial intelligence models may diagnose different types of diseases.

The mediation server 100 may be implemented on a network as a server that mediates between the original collection group 200 and the diagnosis server stage 300. The mediation server 100 may be implemented as a computer device or a plurality of computer devices that provide commands, codes, files, contents, services, etc. by communicating with the original collection group 100 and the diagnosis server stage 300 through a communication network. .

The mediation server 100 generates conversion data based on basic data including personal health records and request signals obtained from a plurality of sources of the original collection group 200, and transmits the transformed data to the diagnosis server stage 300 again. Output data generated by the diagnosis server unit 300 may be received. Thereafter, comprehensive diagnosis data may be generated based on the output data and transmitted to a necessary source or other source among the original collection group 100, thereby providing a system and method for treating treatment using a plurality of artificial intelligence models with improved accuracy.

In a method of predicting/diagnosing medical care based on a personal health record (PHR) or an electronic medical record (EMR) from multiple sources, a plurality of artificial intelligence models may be utilized. At this time, in order to secure data quality, it is very important not only the process of collecting data from each institution, but also the work of standardizing data. For example, it is necessary to integrate and analyze data from various sources to identify the association between different phenotypes, and to unify and transform basic data to have a specific format.

Accordingly, according to the integrated mediation system and method according to an embodiment of the present invention, medical records from different sources are recorded by mediating an original collection group including a plurality of sources and a diagnosis server including a plurality of artificial intelligence models as a mediation server. Based on this, the accuracy and efficiency of prediction/diagnosis for a specific disease can be improved.

Figure 2 is a block diagram schematically showing the internal configuration of the integrated mediation system 10 according to an embodiment of the present invention. The mediation server 100 , the original collection group 200 , and the diagnosis server 300 may communicate with each other while exchanging data through the communication network 400 .

In this figure, one source (200-1) among a plurality of sources of the original collection group (200), the inside of one diagnosis server (300-1) among a plurality of artificial intelligence models of the diagnosis server stage (300) The configuration is shown as an example. Hereinafter, for convenience of description, the original collection group 200 including a plurality of sources may be referred to as the source 200, and the diagnosis server unit 300 including a plurality of diagnosis servers may be referred to as a diagnosis server. (300). The diagnosis server 300 may refer to an artificial intelligence model included in the diagnosis server 300 . Other sources 200 and other diagnostic servers 300 not shown in this figure may also have the same or similar internal configuration.

The mediation server 100 , the source 200 and the diagnosis server 300 may include memories 110 , 210 , and 310 . The memories 110, 210, and 310 are computer-readable recording media, and may include random access memory (RAM), read only memory (ROM), and a permanent mass storage device such as a disk drive. there is. The memory 110, 210, 310 may temporarily or permanently store program codes for controlling each component 100, 200, 300 and data generated and processed on the integrated mediation system 10 of the present invention.

The mediation server 100 , the source 200 and the diagnosis server 300 may include processors 120 , 220 , and 320 . The processors 120, 220, and 320 may control overall operations of the components 100, 200, and 300. The processors 120, 220, and 320 may be configured to process commands of a computer program by performing basic arithmetic, logic, and input/output operations. Instructions may be provided to the processor 120 , 220 , 320 by the memory 110 , 210 , 310 or the communication module 130 , 230 , 330 . For example, the processors 120 , 220 , and 320 may be configured to execute received instructions according to program codes stored in a recording device such as the memory 110 , 210 , and 310 .

The detailed configuration and operation of the processor 120 of the mediation server 100 will be described in more detail in FIG. 3 and related drawings to be described later in association with the respective processors 220 and 320 of the source 200 and the diagnosis server 300. do.

The processor 220 of the source 200 transmits basic data including personal health records of users stored in the memory 210 to the mediation server 100, and again from the mediation server 100 by the diagnosis server 300. Diagnostic data generated based on the generated output data may be received.

The processor 320 of the diagnosis server 300 includes a plurality of artificial intelligence models that generate output data based on basic data obtained from a plurality of sources 200 of the user, and a request signal to be described later includes the plurality of artificial intelligence models. Among the intelligence models, a target artificial intelligence model into which the basic data is input may be selected. At this time, the request signal may select a plurality of target artificial intelligence models.

The mediation server 100 , the source 200 and the diagnosis server 300 may include communication modules 130 , 230 , and 330 . The communication modules 130, 230, and 330 may communicate with each other between the respective components 100, 200, and 300 of the integrated mediation system 10 through the communication network 400, or between different sources and different diagnostic servers. function.

The communication method of the system 10 of the present invention is not limited, and the communication network 400 (eg, a mobile communication network, wired Internet, wireless Internet, and a communication method utilizing a broadcasting network) as well as short-range wireless communication between devices may also be included. The communication network 500 includes a personal area network (PAN), a local area network (LAN), a campus area network (CAN), a metropolitan area network (MAN), a wide area network (WAN), a broadband network (BBN), and the Internet. The network 500 may include one or more arbitrary networks, and the communication network 500 may include a bus network, a star network, a ring network, a mesh network, a star-bus network, a tree or a hierarchical network, and the like. It may include any one or more of the topologies, but is not limited thereto.

Depending on the embodiment, of course, the mediation server 100, the source 200, and the diagnosis server 300 may further include components other than those shown in the drawings.

Figure 3 is a configuration diagram schematically showing the configuration of the processor 120 of the mediation server 100 according to an embodiment of the present invention.

The processor 120 may include a basic data acquisition unit 121 , a conversion data generation unit 122 and a diagnostic data generation unit 123 . The conversion data generator 122 may include a target artificial intelligence model identification unit 1221, an item data extraction unit 1222, a minimum parameter analysis unit 1223, and a matching generator 1224. Depending on embodiments, it goes without saying that the processor 120 may include components other than those shown in this figure. Hereinafter, the operation of the integrated mediation system 10 according to an embodiment of the present invention with a focus on the internal configuration of the processor 120 will be described in more detail.

Figure 4 is a flow chart for explaining the integration mediation method according to an embodiment of the present invention. The integration mediation method of the present invention includes the steps described below.

The basic data acquisition unit 121 of the mediation server 100 includes original data from the original collection group 200 including a plurality of sources and a request signal for selecting a target artificial intelligence model to which the original data is to be input. Acquire data (S100).

Here, with reference to FIG. 6, the basic data will be described in more detail. 6 is a system diagram for explaining an example of an integrated mediation method according to an embodiment of the present invention in more detail. The original data may be personal health records (S1, S2, ..., Sm) transmitted from each source (eg, hospital). The request signal may be a signal for selecting a target artificial intelligence model capable of acquiring desired diagnostic data based on the original data. For example, the basic data acquisition unit 121 inputs the original data S1 from a user who has acquired the original data S1 as a medical treatment result from the first source 200-1 or from the first source 200-1. and transmits to the diagnosis server 300 equipped with the artificial intelligence model to obtain prediction and diagnosis contents, in particular, the target artificial intelligence model, and the converted data can be input to the target artificial intelligence model (S21). Thereafter, the diagnosis server 300 may generate output data based on the converted data using the target artificial intelligence model (S22). The output data may be data including information obtained by predicting and diagnosing the occurrence probability, elapsed time, prognosis, and surgical progress of a specific disease based on the patient's personal health record. Thereafter, the processor 320 of the diagnosis server 300 may transmit the output data to the mediation server 100 through the communication module 330 (S23).

Thereafter, the diagnostic data generator 123 of the mediation server 100 generates diagnostic data based on the output data (S300). The output data represents data generated by the diagnostic server 300 using the converted data as input data of the target artificial intelligence model. In other words, the diagnostic data may be generated based on output data generated using a target artificial intelligence model selected from among a plurality of artificial intelligence models, and at least two target artificial intelligence models may also be included.

The diagnostic data generating unit 123 may perform operations described below. An identification UI capable of identifying a source of the output data, an artificial intelligence model, may be generated, and phenotypes may be matched for each of the output data. Finally, a comprehensive medical certificate may be provided by integrating the output data matched with the phenotype and the identification UI corresponding thereto.

Multiple target artificial intelligence models can generate output data related to the same disease or diagnostic field.

Figure 5 is a flow chart for explaining in more detail the step (S200) of generating converted data, which is one step of the integrated mediation method according to an embodiment of the present invention. It will be described using the above-described FIG. 3 together. Step S200 may include steps described below.

The target artificial intelligence model identification unit 1221 identifies the target artificial intelligence model corresponding to the request signal received by the mediation server 100 (S210).

Thereafter, the item data extractor 1222 analyzes the items of the basic data to extract item data (S220), and the minimum parameter analyzer 1223 analyzes minimum parameter information constituting the target format (S230). . Steps S220 and S230 may be performed in parallel as shown in this figure, or steps S220 or S230 may be performed earlier according to embodiments.

For example, the item may include at least one of medical images, EMR data, biometric information, health data, and time series data, and the item data may include an item and parameter information constituting each item. The medical images may include various types of images used in the medical field, such as CT images and MRI images, and the EMR data may include family history. The biometric information may include a variety of information related to the human body, such as genome information and information derived from a human body, and the health data may include various information related to a user's lifestyle, such as a smoking history. The time series data may include repeatedly measured values obtained from an operating room in a hospital, wearable equipment worn by an individual, and the like, and the repeatedly measured values may include various biosignals such as an individual's blood pressure, electrocardiogram, and brain wave. .

For example, parameter information may include cases to be described later according to items. If the item is a CT image, the parameter information may be a parameter compatible with minimum parameters constituting the target format. When the item is genome information, parameter information may be scale information (whether whole genome or modified gene only). When the item is smoking history, the parameter information is a calculation unit of smoking amount, but the calculation units may be mutually convertible.

Thereafter, the matching generation unit 1224 matches the item data and the minimum parameter information corresponding thereto (S240), and generates the converted data based on the identification result by the identification unit 1221 and the matching result (S240). S250). At this time, the matching generation unit 1224 may generate converted data by converting the basic data into a target format of input data that can be processed by the target artificial intelligence model. The target format is information used in an algorithm for generating output data by a target artificial intelligence model, and may include at least a part of the item data.

6 is a system diagram for explaining an example of an integrated mediation method according to an embodiment of the present invention in more detail. The system diagram of FIG. 6 is the same system as that of FIG. 1, but instead of the diagnosis server 300, artificial intelligence models (M1, M2, ..., Mn) are shown for convenience of description. Unlike this drawing, the number of diagnostic servers 300 and the number of artificial intelligence models M1, M2, ..., Mn may be different, of course.

The basic data acquisition unit 121 of the mediation server 100 may acquire basic data from the original collection group 200 (S100). The basic data includes original data (eg, a personal health record) and a request signal from the plurality of sources 200-1, 200-2, ..., 200-m, and the request signal includes the plurality of sources 200-m. ) contains each individual request signal. At this time, the mediation server 100 may independently receive the individual request signal from each of the plurality of sources 200-1, 200-2, ..., 200-m, and the individual request signal is different from each other. It may be a signal for selecting a model, and one individual request signal may select a plurality of target artificial intelligence models according to an embodiment.

Depending on the embodiment, sharing request signals for requesting original data between different institutions may be exchanged to supplement diagnostic data. For example, the second source 200-2 receives a replenishment request signal from the first source 200-1 and sends its own original data (the second personal health record S2) to the first source 200-1. In addition, the first source 200-1 may transmit basic data including both the first and second personal health records S1 and S2 to the mediation server 100.

The original data may be personal health records (S1, S2, ..., Sm) transmitted from each source (eg, hospital). Personal health records (S1, S2, ..., Sm) obtained from different sources or possessed by different sources may have different phenotypes, such as including different item data (d1, d2, ..., dm).

Hereinafter, it is assumed that lung cancer-related information is diagnosed using the integrated mediation system 10 of the present invention. Items necessary for lung cancer diagnosis include, but are not limited to, lung computed tomography (CT) images (A), family history (B), genomic information (C), and smoking history (D). At this time, the first personal health record S1 includes items A, B, and C (d1), the second personal health record S2 includes items A, B, C, and D (d2), and the third personal health record S2 includes items A, B, C, and D (d2). A case in which the health record S3 includes items A, B, and D (d3) will be described as an example.

Thereafter, the conversion data generation unit 122 of the mediation server 100 may generate conversion data having a target format from the basic data (S200) (see FIG. 4).

The target artificial intelligence model identification unit 1221 may identify a target artificial intelligence model corresponding to a request signal included in the basic data. The request signal may be a signal for specifying and selecting a target artificial intelligence model, or may be a signal requesting that the target artificial intelligence model of the mediation server 100 be determined by a conversion data generation algorithm according to an embodiment.

The item data extractor 1222 may extract item data included in the basic data by analyzing items of the basic data. If the item is a lung CT image (A), the measurement parameters may be different for each environment in which the CT image is taken, and the lung CT image (A) or an image preprocessed/processed image for this image is processed in an artificial intelligence model that is input as input data. The minimum possible parameters may also differ. In other words, analysis/diagnosis of lung cancer is possible using the AI model (target AI model) only when the measurement parameters of the lung CT image (A) satisfy the minimum parameters. The minimum parameter may constitute a target format of converted data obtained by converting from basic data.

When the item is genome information (C), the scale of genes included may be different. For example, genome information C may include whole genome information or only modified genes, and gene scale information included in such genome information may be used as parameter information. On the other hand, when the item is smoking history (D), the calculation unit of smoking amount is different, such as weekly smoking amount, daily smoking amount, etc. for each calculation institution, but it can be calculated in such a way that it can be converted to each other. In this case, the calculation unit of the amount of smoking may be parameter information.

Meanwhile, the minimum parameter analyzer 1223 may analyze minimum parameter information constituting the target format. For example, in this figure, the minimum parameter (P1) required for generating output data (eg, lung cancer-related information diagnosis) of the first artificial intelligence model (M1) is a lung CT image (A), and the second artificial intelligence model (M2) The minimum parameter (P2) of is the lung CT image (A), family history (B), and genomic information (C), and the minimum parameter (P3) of the third artificial intelligence model (M3) is the family history (B) and smoking history (D). The case of is described as an example.

The matching generation unit 1224 matches the item data (d1, d2, ..., dm) and the minimum parameter information (P1, P2, ..., Pn) extracted as described above, and based on the identification result and the matching result Conversion data can be created.

For example, it is assumed that the mediation server 100 extracts the source of the basic data from the first source 200-1 and the item data d1 of the original data S2 among the basic data from A, B, and C do. Since the extracted item data d1 satisfies the minimum parameters of the first and second artificial intelligence models M1 and M2 in this drawing, the two artificial intelligence models M1 and M2 become target artificial intelligence models and are Output data can be generated by

As another example, it is assumed that the source of the basic data is an mth source 200-m and item data dm of the original data Sm among the basic data from the mth source 200-m is extracted as A, B, and D. Since the extracted item data dm satisfies the minimum parameters of the first and nth artificial intelligence models M1 and Mn in this figure, the two artificial intelligence models M1 and Mn become target artificial intelligence models and are Output data can be generated by

At this time, the conversion data generating unit 122 may perform data conversion processing to enable compatibility between the item data d1 and the minimum parameter information P1 or P2. For example, among the item data, genome information (C) must be converted to include minimum scale information enabling an output operation of an artificial intelligence model using C as a parameter of input data. Smoking history (D) should be converted to have a calculation unit that enables the output operation of the artificial intelligence model that uses D as a parameter of input data.

In the above, it has been described as an example that the target artificial intelligence model is two for each individual request signal, but it can be more than that.

The conversion data generator 122 and the diagnostic data generator 124 may operate in parallel with respect to each of the individual request signals. In other words, output data and diagnostic data may be generated for all of the target artificial intelligence models corresponding to each of the individual request signals.

The converted data generator 122 according to another embodiment of the present invention may further include a normalization unit and a noise processing unit. The standardization unit may standardize a data structure format for basic data for each of a plurality of sources 200 . For example, standardized data structure formats may include Fast Health Interoperability Resources (FHIR). The noise processing unit may remove noise including information unrelated to the minimum parameter of the target format with respect to the basic data for each of the plurality of sources 200 . For example, if the item is a lung CT image (A), irrelevant images including CT images of other organs or images other than CT images of the lungs may be removed. Alternatively, by analyzing the image parameters implementing the lung CT image, noise, such as an image portion unrelated to the extracted reference data or a broken portion due to a foreign substance or equipment failure in the lung CT image, may be removed.

Although the preferred embodiments of the present invention have been shown and described above, the present invention is not limited to the specific embodiments described above, and is common in the art to which the present invention pertains without departing from the gist of the present invention claimed in the claims. Of course, various modifications are possible by those with knowledge of, and these modifications should not be individually understood from the technical spirit or prospect of the present invention.

Therefore, the spirit of the present invention should not be limited to the above-described embodiments, and not only the claims to be described later, but also all scopes equivalent to or equivalently changed from these claims fall within the scope of the spirit of the present invention. would be considered to be in the category.

100: mediation server
200: original collection
300: diagnosis server stage
400: communication network
120: processor
121: basic data acquisition unit
122: conversion data generation unit
123: diagnostic data generating unit

Claims (16)

The mediation server selects original data including the user's personal health record from an original collection group including a plurality of sources, which are hospitals providing the user's medical treatment and diagnosis records, and a target artificial intelligence model into which the original data is to be input obtaining basic data including a request signal;
The mediation server identifies a target artificial intelligence model in response to the request signal, analyzes items of the basic data to extract item data, and minimum parameter information constituting a target format of input data that can be processed by the target artificial intelligence model. The target artificial intelligence that analyzes, matches the item data and minimum parameter information corresponding to the item data, and enables compatibility between the item data and minimum parameter information based on a matching result satisfying the minimum parameter information. generating conversion data for the model; and
The mediation server transmits the converted data to a diagnosis server, the diagnosis server receives output data generated using a model learned as input data of the target artificial intelligence model, and diagnostic data based on the output data Generating a; Including,
The basic data item includes at least one of medical images, EMR data, biometric information, health data, and time series data;
The item data varies depending on the item, and when the item is a lung CT image, images other than lung CT images are removed and lung CT images are taken, and include different measurement parameters for each environment, but in the target artificial intelligence model It includes a minimum processable parameter, and the target format is information used in an algorithm for generating output data by the target artificial intelligence model, and includes at least a part of item data,
The output data is generated using the target artificial intelligence model selected in response to the request signal from among a plurality of artificial intelligence models,
The diagnostic data includes an operation progress of the user, an integrated mediation method for medical treatment.
delete delete delete delete According to claim 1,
The request signal includes an individual request signal of each of the plurality of sources,
Wherein the mediation server independently receives the individual request signals from each of the plurality of sources.
According to claim 1,
Generating the conversion data,
Standardizing a data structure format for the basic data for each of the plurality of sources; and
a noise processing step of removing noise including information unrelated to the minimum parameter of the target format with respect to the basic data for each of the plurality of sources;
Including further, the integrated mediation method for the treatment act.
According to claim 1,
Generating the diagnostic data,
generating an identification UI capable of identifying a source of the output data;
Matching phenotypes for each of the output data; and
Compiling the output data matching the phenotype and the identification UI corresponding thereto, and providing a comprehensive medical certificate;
Including, an integrated mediation method for the practice of treatment.
A request signal for selecting original data including the user's personal health record from the original collection group including a plurality of sources, which are hospitals providing the user's treatment and diagnosis records, and a target artificial intelligence model to which the original data is to be input a basic data acquisition unit that acquires basic data including;
Identifying a target artificial intelligence model in response to the request signal, analyzing items of the basic data to extract item data, and analyzing minimum parameter information constituting a target format of input data that can be processed by the target artificial intelligence model, Conversion of the target artificial intelligence model for matching the item data and minimum parameter information corresponding to the item data, and enabling compatibility between the item data and minimum parameter information based on a matching result satisfying the minimum parameter information a conversion data generating unit generating data; and
Diagnosis data for transmitting the converted data to a diagnosis server, receiving output data generated using a model trained as input data of the target artificial intelligence model in the diagnosis server, and generating diagnostic data based on the output data Including; generating unit;
The basic data item includes at least one of medical images, EMR data, biometric information, health data, and time series data;
The item data varies depending on the item, and when the item is a lung CT image, images other than lung CT images are removed and lung CT images are taken, and include different measurement parameters for each environment, but in the target artificial intelligence model It includes a minimum processable parameter, and the target format is information used in an algorithm for generating output data by the target artificial intelligence model, and includes at least a part of item data,
The output data is generated using the target artificial intelligence model selected in response to the request signal from among a plurality of artificial intelligence models,
The diagnostic data includes the user's surgical history, integrated mediation system for medical treatment. delete delete delete delete According to claim 9,
The request signal includes an individual request signal of each of the plurality of sources,
The basic data acquisition unit independently receives the individual request signal from each of the plurality of sources, an integrated mediation system for medical treatment.
According to claim 9,
The conversion data generating unit,
a standardization unit that standardizes a data format of the basic data for each of the plurality of sources; and
a noise processing unit to remove noise including information unrelated to the minimum parameter of the target format with respect to the basic data for each of the plurality of sources;
Further comprising, an integrated mediation system for medical treatment.
According to claim 9,
The diagnostic data generating unit,
generating an identification UI capable of identifying the source of the output data;
Matching the phenotype for each of the output data,
An integrated mediation system for medical treatment that provides a comprehensive diagnosis by synthesizing the output data matching the phenotype and the identification UI corresponding thereto.

Images