A method for constructing a universal health information data lake based on the MOF model

Abstract

The present invention discloses a method for constructing a national health information data lake based on the MOF model. First, the runtime model is refined, a meta-meta model is defined, a meta-meta model is constructed using the meta-meta model, an M1 model is constructed using the meta-meta model, a service model in the model is associated with a data model to be called, and a business model in the model is associated with a service model to be called; finally, the business model in the model is called to complete data entry and data exit from the lake; the present invention constructs a corresponding application model centered on personal health information by comparing the four-layer model in the above MOF with the medical and health information field, solving the problem of efficiently and flexibly storing national health information data of different granularities. The national health information data lake constructed according to the application model also greatly reduces the risk of becoming a data "swamp", and can also promote and accelerate the interconnection and sharing of medical and health information.

Description

Technical Field

[0001] The present invention belongs to the technical field of big data, and in particular relates to a method for constructing a national health information data lake based on a MOF model. Background Art

[0002] With the development of medical and health informationization in my country, medical and health information systems have accumulated a huge amount of data. In addition, since the data model structures used by various system platforms are different, it is difficult to use the gathered medical and health data efficiently without applying relevant theories to create a unified model. Therefore, in order to interconnect medical and health data, break data silos, and realize data sharing between system platforms under the background of big data and cloud computing, it is a very critical link to establish a unified data application model based on advanced model theory. At this stage, there are many data "swamps" in the data lake in the field of medical and health information because there is a lack of advanced model theory to support it.

[0003] At present, no similar patents have been found in the market. Due to the diversity and complexity of the business in the field of medical and health information, it is difficult to establish a unified model to support data storage. According to the MOF model theory, the MOF model theory is applied to the field of medical and health information. According to the MOF model theory, each layer of the model is mapped to the field of medical and health information and the corresponding MOF application model in the field of medical and health information is created. It can well solve the problem of high-quality unified storage of data in the field of medical and health information, because the editable and maintainable characteristics of the MOF application model instance can realize that the stored data is in accordance with the intentions of experts in the field of medical and health information, avoiding the embarrassing situation in which most of the stored data has become a data "swamp" and is difficult to exchange.

[0004] National Health Information Platform: Exchanges data with connected medical and health institutions, lower-level platforms, and other business organizations through the information access layer to achieve interconnection and information sharing, improve the level of information application in various regions and medical and health institutions, and is of great significance to accelerating the development of "Internet + Medical Health" and implementing the interconnection and sharing of national health information.

[0005] Data Lake: A data lake is a system or repository that stores data in its original format. A data lake is typically a single store for all enterprise data. It is used for tasks such as reporting, visualization, advanced analytics, and machine learning. A data lake can include structured data (rows and columns) from relational databases, semi-structured data (CSV, logs, XML, JSON), unstructured data (emails, documents, PDFs), and binary data (images, audio, video). Summary of the invention

[0006] In view of the deficiencies in the prior art, the present invention provides a method for constructing a national health information data lake based on the MOF model.

[0007] The present invention mainly solves the problem of flexible definition of medical and health information field data in the national health information data lake, flexible definition of medical and health information objects, and storage of flexible definition of medical and health information elements and flexible data access for different businesses. The application scenario of the present invention is the demand for building a middle platform for regional platform informatization, which solves the efficient reading and writing of data in the data lake by the middle platform, so that the data in the data lake will not easily become a "swamp" of data. The data stored in the data lake are basically strongly related to a certain clear field. The data can be defined and stored according to the level of field, object, and information element, and can provide diversified data services to the outside according to the model.

[0008] The present invention mainly applies the MOF four-layer model to implement customized control of model creation at different levels for data in the field of medical and health information. The creation of the model strictly complies with the MOF four-layer model, so that the health information data of the whole people can be collected and stored in the data lake according to the customized model. The data stored in the data lake conforms to the constraints of the MOF four-layer model, and then the corresponding data call service can be provided to the outside by selecting or customizing the service model and the business model. A system for custom creation and maintenance of a model is provided in the present invention. In addition to the M0 layer model when the system is running, the model also includes the meta-meta model of the M3 layer, the meta-model of the M2 layer, and the model of the M1 layer. The model of the M1 layer is divided into data model, service model and business model according to the category required for the informatization of the whole people's health.

[0009] A method for building a universal health information data lake based on the MOF model, the steps are as follows:

[0010] Step 1: Refine the runtime model;

[0011] Step 2: Define the meta-meta model;

[0012] According to the industry characteristics of the medical and health information field, which is strongly related to personal health information, the meta-meta model is defined as a series of components, including meta-domains, meta-nodes, meta-classes, meta-attributes, meta-operations, and meta-relations;

[0013] Step 3: Use the meta-meta-model to build the meta-model;

[0014] Step 4: Use the metamodel to build the M1 model;

[0015] Step 5: Associate the service model in the model with the data model to be called;

[0016] Step 6: Associate the business model in the model with the service model to be called;

[0017] Step 7: Call the business model in the model to complete data entry and data exit from the lake.

[0018] Entering the lake: The business model calls the service model, and the service model stores the incoming medical and health data into the data lake according to the data model.

[0019] Out of the lake: The data caller calls the service model through the business model, and the service model obtains medical and health-related data from the data lake according to the data model and returns it to the data caller.

[0020] The data entering the lake is processed by the model and stored in the data lake. The data leaving the lake is assembled by the model and returned to the caller.

[0021] Step 1:

[0022] Abandoning the modeling approach centered around the operation of medical and health institutions, modeling is centered around personal health information, focusing on and covering the information objects at system runtime related to the medical, health, public health, and elderly care fields and the entire population, thereby refining a runtime model for the medical and health field.

[0023] Step 3:

[0024] Define the metamodel, which consists of business domain, business line, business object, data entity, and data entity attribute. These five components are defined by the meta-nodes in the meta-metamodel, and the relationship between the previous node and the next node is defined by the meta-relationship in the meta-metamodel.

[0025] Step 4:

[0026] The metamodel is used to assemble the M1 model according to the specific data structure. The M1 model defines the data model, service model and business model according to the classification. The data attributes of the data model include data name, data domain, data description and data structure. The service attributes of the service model include service name, service version and service description. The business attributes of the business model include business name, business domain, business description and security policy.

[0027] Furthermore, the meta-meta model predefines inputs based on the characteristics of the medical and health information technology field; the meta-model automatically collects inputs from source databases in the medical, health, public health, and elderly care industries; the data model, service model, and business model in the M1 model are collected by experts in the medical and health information field according to the M1 layer model in the MOF model layer, and customized model instances are tailored to the medical and health information field.

[0028] The beneficial effects of the present invention are as follows:

[0029] The present invention constructs a corresponding application model centered on personal health information by comparing the four-layer model in the above MOF with the field of medical and health information, solving the problem of efficiently and flexibly storing national health information data of different granularities. The national health information data lake constructed according to the application model also greatly reduces the risk of becoming a data "swamp", and can also promote and accelerate the interconnection and sharing of medical and health information.

[0030] The present invention discloses a technical method for solving the storage problem of universal health information data in the field of medical and health information by building a data lake based on an MOF application model. The data aggregation of heterogeneous systems is well completed through a standard data model. Without adjusting the source data structure and business process, the demand can be met through flexible business model configuration, which greatly increases the flexibility of business expansion. The present invention provides the definition of the field of medical and health information, the definition of medical and health information objects, and the definition of medical and health information elements, and at the same time, it is very convenient to share and store data. Through this method, the application efficiency and value of data in hospitals and regions can be effectively improved, the pace of interconnection and interoperability of medical and health information can be promoted, the time and cost of data docking implementation can be reduced, and data services of different granularities in a certain field of medical and health information can be provided to the outside quickly and accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a structural diagram of the model of the present invention;

[0032] Figure 2 It is a call diagram for data in and out of the lake of the present invention;

[0033] Figure 3 It is a schematic diagram of the M0 layer model structure;

[0034] Figure 4 It is a schematic diagram of the M1 layer model structure;

[0035] Figure 5 It is a schematic diagram of the M2 layer model structure;

[0036] Figure 6 It is a schematic diagram of the M3 layer model structure;

[0037] Figure 7 This is a block diagram of the design solution. DETAILED DESCRIPTION

[0038] The method of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0039] like Figure 1As shown in the figure, the model structure includes runtime objects (M0), models (M1), metamodels (M2), and meta-metamodels (M3). Models are divided into data models, service models, and business models by type. Runtime objects (M0) are objects used when the system is running and are invisible to users. The model (M1) composed of data models, service models, and business models is a model created based on the business and is also the model that users pay the most attention to and participate in establishing. The metamodel (M2) is collected from source databases in the medical, health, public health, and elderly care industries. The meta-metamodel (M3) is created and maintained by developers.

[0040] Figure 2 The process of data entering and leaving the lake is described. Data enters and leaves the lake through business models. Business models call service models, and service models call data models to process the data accordingly.

[0041] A method for building a universal health information data lake based on the MOF model, the steps are as follows:

[0042] Step 1: Refine the runtime model;

[0043] Abandoning the previous modeling approach centered around the operation of medical and health institutions, innovative modeling is centered around personal health information, focusing on and covering information objects at system runtime related to the medical, health, public health, and elderly care fields and the entire population, thereby refining a runtime model for the medical and health field.

[0044] Step 2: Define the meta-meta model;

[0045] According to the industry characteristics of the medical and health information field, which is strongly related to personal health information, the meta-meta model is defined as a series of components, including meta-domains, meta-nodes, meta-classes, meta-attributes, meta-operations, and meta-relations;

[0046] Step 3: Use the meta-meta-model to build the meta-model;

[0047] Define the metamodel, which consists of business domain, business line, business object, data entity, and data entity attribute. These five components are defined by the meta-nodes in the meta-metamodel, and the relationship between the previous node and the next node is defined by the meta-relationship in the meta-metamodel.

[0048] Step 4: Use the metamodel to build the M1 model;

[0049] The metamodel is used to assemble the M1 model according to the specific data structure. The M1 model defines the data model, service model and business model according to the classification. The data attributes of the data model include data name, data domain, data description and data structure. The service attributes of the service model include service name, service version and service description. The business attributes of the business model include business name, business domain, business description and security policy.

[0050] Step 5: Associate the service model in the model with the data model to be called;

[0051] Step 6: Associate the business model in the model with the service model to be called;

[0052] Step 7: Call the business model in the model to complete data entry and data exit from the lake.

[0053] Entering the lake: The business model calls the service model, and the service model stores the incoming medical and health data into the data lake according to the data model.

[0054] Out of the lake: The data caller calls the service model through the business model, and the service model obtains medical and health-related data from the data lake according to the data model and returns it to the data caller.

[0055] The data entering the lake is processed by the model and stored in the data lake. The data leaving the lake is assembled by the model and returned to the caller.

[0056] Furthermore, the meta-meta model is predefined by developers based on the characteristics of the medical and health information technology field; the meta-model automatically collects input from source databases in the medical, health, public health, and elderly care industries; the data model, service model, and business model in the M1 model are collected by medical and health information field experts according to the M1 layer model in the MOF model layer, and customized model instances are tailored to the medical and health information field.

[0057] The following is a comparison between the four layers of MOF and the field of medical and health information to illustrate how to build these four layers of models in the universal health information data lake:

[0058] I.MOF four-layer model

[0059] The four layers of MOF are: M0 (runtime object), M1 (model layer), M2 (metamodel layer), and M3 (meta-metamodel layer). The latter layer models the former layer. The higher the level, the greater the granularity, which is very suitable as a theory for abstract construction of models in the field of medical and health information.

[0060] II.M0 Layer Model

[0061] The M0 layer model defines the objects during system runtime with personal health information as the center. It is the instantiation of the M1 layer model and also the most specific layer of model. The national health information data lake contains data in multiple fields such as personal files, medical care, health, public health, and elderly care. The present invention identifies more than 800 types of objects defined in the MO layer model, such as personal files, medical institutions, business departments, medical staff, medical orders, electronic medical records, and imaging pictures, covering structured, semi-structured, and unstructured data.

[0062] III.M1 Layer Model

[0063] The M1 layer model describes the data structure of the M0 layer object. Taking personal file query as an example, it defines the personal file query business model (business name: personal file query, business domain: personal file, business description: query personal file information, security policy: return the personal file information content corresponding to the permission according to the assigned permission), defines the personal file query service model (service name: personal file query, service version: V4.0, service description: query personal file information), defines the personal file data model (data name: personal file information, data domain: personal file, data description: information related to personal file, data structure: personal basic information, personal health file Case information, personal basic card information, personal address information, personal contact information, etc.), in the above example, personal basic information attributes include name, gender, date of birth, contact number, etc. In addition to personal file information, medical institution attributes include institution name, business license code, legal person name, institution address, etc., business department attributes include department code, department name, department description, etc., medical staff attributes include name, gender, title, department, etc., medical order attributes include order number, order date, order department, etc., electronic medical record attributes include medical record type, medical record name, medical record chapter, etc., image pictures include image code, image path, image file, etc. These classes and attributes are abstracted into metadata at the M1 layer. Different from the traditional relational table structure, the M1 layer model breaks up all metadata and manages them in a unified manner. The present invention abstracts a total of more than 16,000 similar metadata.

[0064] VI.M2 layer model

[0065] The M2 layer model is a description of the M1 layer model. It models the M1 layer metadata and establishes the relationship between the M1 layer metadata. For example, the personal basic information metadata is abstracted into a table containing many attributes, the personal name metadata is abstracted into an attribute consisting of no more than a specified number of characters, and the personal basic information metadata and the personal name metadata are included. There is a dependency relationship between the metadata of the department to which the medical staff belongs and the metadata of the business department, etc. The data of this model layer is automatically collected from the source database of the medical, health, public health, and elderly care industries.

[0066] V.M3 layer model

[0067] The M3 layer model is a meta-meta-model used to define the M2 meta-model and provide a basic meta-meta-model to quickly assemble an M2 meta-model. Because the field of medical and health informationization is data-intensive and data structure-focused, the meta-meta-model components are relatively limited and stable, mainly including meta-domains, meta-classes, meta-attributes, meta-behaviors, meta-operations, and meta-relationships. In addition, the relationship between meta-classes is also defined, mainly including: inclusion, inheritance, type reference, and dependency. In the above example, personal information metadata is defined by meta-classes, and name metadata, gender metadata, date of birth metadata, and contact phone metadata are defined by meta-attributes. Meta-classes and meta-attributes are connected by inclusion relationships.

[0068] The MOF model layer (M1) is divided into data model, service model and business model according to the needs of the national health information data lake. The business model calls the service model, and the service model completes the entry and exit of data into the lake through the data model. The present invention can adjust the content of various models at any time according to the needs of applications in the field of medical and health information, and has openness, scalability and interoperability.

Claims

1. A method for constructing a national health information data lake based on the MOF model, characterized in that: Here are the steps: Step 1: Refine the runtime model; Step 2: Define the meta-meta model; According to the industry characteristics of the medical and health information field, which is strongly related to personal health information, the meta-meta model is defined as a series of components, including meta-domains, meta-nodes, meta-classes, meta-attributes, meta-operations, and meta-relations; Step 3: Use the meta-meta-model to build the meta-model; Step 4: Use the metamodel to build the M1 model; Use the metamodel to assemble the M1 model according to the specific data structure. The M1 model defines the data model, service model and business model according to the classification. The data attributes of the data model include data name, data domain, data description and data structure. The service attributes of the service model include service name, service version and service description. The business attributes of the business model include business name, business domain, business description and security policy. Step 5: Associate the service model in the model with the data model to be called; Step 6: Associate the business model in the model with the service model to be called; Step 7: Call the business model in the model to complete data entry and data exit from the lake; Entering the lake: The business model calls the service model, and the service model stores the incoming medical and health data into the data lake according to the data attributes of the data model; Out of the lake: The data caller calls the service model through the business model. The service model obtains medical and health-related data from the data lake according to the data attributes of the data model and returns it to the data caller. The data entering the lake is processed by the model and stored in the data lake; the data leaving the lake is assembled by the model and returned to the caller.

2. According to claim 1, a method for constructing a national health information data lake based on the MOF model is characterized in that: Step 1: Abandoning the modeling approach centered around the operation of medical and health institutions, modeling is centered around personal health information, focusing on and covering information objects at system runtime related to the medical, health, public health, and elderly care fields and the entire population, thereby refining a runtime model for the medical and health field.

3. A method for constructing a national health information data lake based on the MOF model according to claim 2, characterized in that: Step 3: Define the metamodel. The metamodel consists of business domains, business lines, business objects, data entities, and data entity attributes. These five components are defined by meta-nodes in the meta-metamodel, and the relationship between the previous node and the next node is defined by the meta-relationship in the meta-metamodel.

4. A method for constructing a national health information data lake based on the MOF model according to claim 3, characterized in that: The meta-meta model predefines inputs based on the characteristics of the medical and health informationization field; the meta-model automatically collects inputs from source databases in the medical, health, public health, and elderly care industries; the data model, service model, and business model in the M1 model are collected by experts in the medical and health information field according to the M1 layer model in the MOF model layer, and customized model instances are tailored to the medical and health information field.

Images