Study And Practice On Hierarchical Medical Information Model

The improvement of medical service quality,the reduction of medical service costs,and even the advancement and development of medicine are increasingly dependent on the effective use of medical data.The semantic interoperability of medical information is the basis for the effective use of decentralized and heterogeneous medical data,and has become an important issue in medical data science.Traditional semantic interoperability methods based on message or document standards only standardize fragment or local information,and have poor completeness,which cannot flexibly adapt to changes in domain requirements.The hierarchical information model has high scalability and is suitable for information modeling in complex fields,which has gradually become the main method to solve this problem in recent years.The key to the hierarchical modeling of medical information is the hierarchical expression framework and collaborative modeling mechanism of the model.The former hierarchically expresses the semantics of medical information in a complex environment.Compared with the single-layer information model,it not only ensures the flexibility of information expression,but also constrains the consistency of information to some extent.The latter is committed to establishing a modelling mechanism for open collaboration among multi-regional and interdisciplinary people,so that it can fully cover the information needs of the field and ensure that the information expression has sufficient crowd recognition.At present,the existing hierarchical models such as HL7 V3 and openEHR have been applied in a certain range,but they have not achieved a good balance between flexibility and consistency.Therefore,they cannot flexibly adapt to the rapid growth of domain requirements,while at the same time ensuring the consistency of the developing model.In view of the above problems,this thesis conducts study and practice on the hierarchical modeling of medical information.Based on the analysis and comparison of the existing representative hierarchical medical information models,with openEHR as the research object,a domain-controlled hierarchical expression framework and an open hierarchical collaborative modeling mechanism are established;a medical information collaborative modeling method for semantic interoperability is proposed;and an electronic medical record standard information model and a standard information model for acute coronary syndrome are established.1.The semantic interoperability requirements of medical information and its influencing factors are analyzed and summarized,and the medical information semantic interoperability requirement model is established.Based on this criterion,the typical medical information hierarchical models such as HL7 V3,FHIR and openEHR are analyzed,and their advantages and disadvantages in hierarchical expression architecture and collaborative modeling mechanism are compared and analyzed.On this basis,the design solution of medical information hierarchical model based on openEHR model is proposed.2.Aiming at the problem that the openEHR hierarchical expression framework has a large degree of freedom of modeling,the original hierarchical expression framework is extended,and a set of archetype patterns for constraining information expression forms is defined and established.the medical domain ontology based on the openEHR reference model is established based on the integration of the existing ontology and model resources,and;a set of archetype patterns covering the concepts of clinical diagnosis and treatment is studied and established;and the case study proves the feasibility and effectiveness of modeling based on the archetype pattern.3.Aiming at the problem of insufficient model knowledge exchange and sharing in the openEHR collaborative modeling mechanism,a hierarchical open collaborative modeling mechanism is proposed and established based on the knowledge ecosystem theory.According to the Distribution,Interaction,Competition and Evolution(DICE)model of knowledge ecology,the problems existing in the openEHR collaborative modeling mechanism are analyzed;an open hierarchical modeling mechanism with multi-dimensional coordination among groups,between groups and between levels is designed to implement the medical information collaborative modeling platform.The results of modeling practice show that the mechanism can better achieve the collaboration of multiple professionals.4.A medical information collaborative modeling method for semantic interoperability is proposed,which defines the seven stages of iteration required for modeling:requirements analysis,concept design,archetype development,archetype review,archetype publish,archetype collaboration and archetype application.An evaluation scheme for verifying the model from model consistency and completeness is designed.5.Aiming at the modeling requirements of current electronic medical records,a set of electronic medical record standard information model including 962 data elements,45 domain concepts and 64 archetypes are developed with collaboration among multiple organizations using the modeling method proposed in the paper.The standard information model has been applied to multiple data repository projects shared by national,military,regional and hospital electronic medical records,and the feasibility and effectiveness of this thesis' research method is verified by data interoperability cases between different institutions.6.Taking the acute coronary syndrome as an example,through the collaboration of multiple professionals in the project and the collaboration with the New Zealand Acute Coronary Syndrome Database project team,26 archetypes of acute coronary syndrome are defined and a set of standard information models for acute coronary syndrome are constructed.The standard information models have been applied to the coronary clinical image big data platform participated by 16 medical institutions,and it can automatically obtain relevant disease information from the hospital electronic medical record and implement semantic interoperability with the New Zealand acute coronary syndrome database.The results verify the feasibility and effectiveness of the research method of this thesis.