Keywords

1 Introduction

On February 4 and March 4, 2022, the 24th Winter Olympic Games and the 13th Winter Paralympic Games will be held in Beijing and Zhangjiakou respectively. The Winter Olympic Games proposes to establish a scientific and technological winter Olympic emergency medical security system, form an intelligent, standardized, and integrated emergency medical support capability, provide efficient and reliable technical support, and lay the foundation for the construction of a new model of emergency medical security for emergencies.

Ontology construction is to extract knowledge from a certain field to form semantic concepts, examples and relationships between data describing the field [1]. As a representation model of the medical concept system, the medical ontology is the core content of the medical information system. Constructing a medical ontology is conducive to promoting the sharing of medical knowledge, while supporting the application of medical intelligence. In the field of open medicine, UMLS [2] (Unified Medical Language System) integrates clinical medicine, basic medicine, pharmacy and other medical terms. OpenGALEN [3] (Generalized Architecture for Languages, Encyclopedias and Nomenclatures in Medicine) covers pathophysiology, surgery, medicine and other medical knowledge. CMeKG [4] (Chinese Medical Knowledge Graph) realizes the knowledge connection between diseases, symptoms, drugs, and diagnosis and treatment techniques.

The current medical ontology is mainly for the establishment of a general medical ontology for certain diseases or for the open medical field, and seldom constructs an ontology model for a specific application scenario. Based on the combination of the Winter Games emergency medical security scene and medical knowledge, this paper constructs the Winter Olympics emergency medical security knowledge ontology. It is convenient for users to quickly and accurately screen information during the subsequent Winter Olympics Games emergency medical security intelligent application process, and provide timely treatment plans, providing data support for the intelligent decision-making of the Winter Olympics Games emergency medical security.

2 Ontology Construction Methods and Models

2.1 Ontology Construction Methods

The current typical methods of ontology construction include five-step method, seven-step method, skeleton method, TOVE method, five-step cycle method, and cycle acquisition method. Among them, the seven-step method [5] was developed by Stanford University proposes a more mature ontology construction method in the medical field. The seven steps of the seven-step method are to determine the professional neighborhood and category of the ontology, and examine the reuse of existing ontology. Possibility, list the important terms in the ontology, define the class and class hierarchy, define the attributes of the class, define the facet of the attributes, and create instances. However, the seven-step method lacks the necessary evaluation and feedback links. The skeleton method [6] is the process of constructing enterprise ontology by the Institute of Artificial Intelligence, University of Edinburgh. The method of summary formation is mainly used to establish a collection of related enterprise terms and definitions. The main steps include determining the application purpose and scope of the ontology, establishing the ontology, evaluating the ontology, and archiving the ontology.

2.2 Four-Dimensional General Disease Ontology Model

Four-dimensional general disease ontology model [7] is an ontology-based auxiliary system established by Hadzic et al. of Curtin University in Australia, which is also the top-level ontology model of the disease neighborhood.The four-dimensional general disease ontology model mainly describes the disease from the four dimensions of disease type, phenotype, etiology and treatment methods.

In the Four-dimensional general disease ontology model, the disease type is used to describe the classification system of the disease, which is a hierarchical classification system structure; the phenotype mainly describes the various clinical manifestations of the disease; the etiology is the pathogenic factor describing the disease, mainly including genetic factors and There are two major types of environmental factors; treatment methods are possible effective treatment methods for specific diseases. The four dimensions are independent of each other, but they can be combined to describe the overall knowledge of a certain disease field.

3 Construction of Emergency Medical Ontology for Winter Olympics

Based on the skeleton method and the seven-step method, this paper proposes a new ontology construction process based on the application scenarios of the Winter Olympics Games emergency medical security, and reuses the above-mentioned four-dimensional general disease ontology to construct the ontology framework. The construction process is shown in Fig. 1. The following is a specific analysis of each stage:

Fig. 1.
figure 1

Ontology construction process.

Full size image

3.1 Analysis of Application Scenarios of Winter Olympics Games Emergency Medical Security Ontology Acronyms

Analysis of application scenarios of Winter Olympic Games emergency medical security ontology Although the Winter Olympic Games is a winter snow sports event, it has the common characteristics of general large-scale event emergencies [8]. The first is suddenness and uncertainty. It is difficult to accurately predict and grasp the time, scale, specific situation and degree of impact of the outbreak. The second is publicity, which is public for the impact of the Winter Olympic Games and the subjects involved. Even if the scope of the event is not directly involved in the public neighborhood, the development and results of the event will attract public attention. The third is the diversity of events. The types of emergencies for large-scale events generally include natural emergencies, accidents, disasters, public health incidents, social security incidents and terrorist attacks. In particular, most sports venues in the Winter Olympic Games have high altitude and temperature. Low characteristics, there is the possibility of frostbite and blindness. The fourth is the hazard. The hazards of emergencies in the Winter Olympic Games are reflected in the impact of the normal order of the games, the threat to the lives of the public, and negative public opinion. Therefore, higher requirements are put forward for emergency rescue during the Winter Olympics Games.

Emergency rescue in the face of emergencies in large-scale events should follow the principles of locality, proximity, safety and efficiency. First of all, strive to accurately assess the injury at the scene in the shortest possible time, and conduct timely and effective treatment. Secondly, in large-scale event emergencies, most of the wounded still need to be evacuated after on-site classification and treatment. During the evacuating process, accidents should be prevented and the wounded should be sent to the hospital safely and quickly. Finally, the wounded need the best handover work in designated hospitals so that the wounded can receive timely and effective treatment.

3.2 Winter Olympics Emergency Medical Security Ontology Corpus

Knowledge datas sources in the medical neighborhood are highly specialized, complex, and widely distributed in multi-source heterogeneous datas, such as medical books, electronic medical records, etc. At the same time, medical ontology requires high accuracy. Therefore, the professionalism, authority, effectiveness and comprehensiveness of hospital knowledge should be considered when selecting the emergency medical security library for the Winter Olympics Games. This paper selects medical books, literature, and ICD-10 as the ontology corpus of the Winter Olympic Games emergency medical security.

  • In terms of medical books, this article mainly selects “Huang Jiasi Surgery” [9] and “Surgery” [10]. The two surgical books as the required corpus contain a complete medical knowledge system of various diseases, concepts, physiology and pathology, and treatment of surgery, and are published by the People’s Medical Publishing House, with high authority and comprehensiveness. At the same time, in the application of medical emergency protection scenarios for the Winter Olympics, more emphasis is placed on first aid measures for the wounded. Therefore, the two books “Emergency Handbook” [11] and “Emergency and Disaster Medicine” [12] published by the People’s Medical Publishing House are selected for the treatment of physiology. While analyzing the pathology, it focuses on the knowledge of diagnosis basis and first aid measures, which is practical for the emergency medical insurance scene of the Winter Olympics. Through the analysis of the semantic relations in the above four medical books, the ontology framework of the Winter Olympics emergency medical support knowledge is formed.

  • Literature is one of the most important carriers of knowledge in the recording field. In the process of constructing the Winter Olympics Games emergency medical security ontology, we also used the medical literature as the source of the corpus. As the latest research results, medical literature contains many cutting-edge diagnostic methods and treatment plans, which can be used as an important supplement to other knowledge.

  • ICD-10 [13] (international classification of diseases) The tenth version of the International Classification of Diseases is a system that categorizes diseases according to rules based on their symptoms, signs and other characteristics, and expresses them by coding. This article uses ICD-10 to guide ontology construction.

3.3 Ontology Analysis and Definition of Emergency Medical Security for the Winter Olympics Games

The four-dimensional general disease ontology model covers the most basic characteristics of a certain disease, but it is not perfect for the application scenarios of the Winter Olympics Games. This paper improves the four-dimensional general disease ontology model from the following aspects, and defines an ontology model that conforms to the emergency medical insurance of the Winter Olympics Games.

  • The property dimension of disease is mainly the type of disease and the core concept of disease. Among them, ICD-10 standardizes and structure the names of diseases. It is conducive to subsequent fusion with other knowledge bases and knowledge fusion in the process of knowledge graph construction.

  • The dimensions of disease visits are mainly to provide decision-making plans for the evacuations of the wounded in the Winter Olympic Games emergency, including departments and hospitals.

  • The epidemiological dimension of the disease is mainly to apply epidemiological investigations during the Winter Olympics Games, to quickly and accurately trace the cause of the disease, and to control it to minimize the harm, including susceptible population, susceptible age, Disease-prone areas, transmission route and source of infection.

  • The physiopathological dimension of the disease is an improvement of the four-dimensional ontology general model, and the prone site is added on the basis of the cause, because after the ABBCS assessment is completed in the first aid field, the heart, breathing, abdomen, spinal cord, Nine parts of the head, kidneys, limbs, arteries, and nerves were used to preliminarily classify the wounded.

  • The clinical manifestation dimension of the disease adds injury staging, differential diagnosis and complications based on the representation dimension in the four-dimensional general disease ontology model. The severity of the disease varies in different injury stages, especially in batches. In the case of injury, it is more necessary to determine the priority of treatment according to the clinical manifestations of the disease at different stages. In addition, the occurrence of a disease may cause some complications at the same time. These complications are similar to the original disease in clinical manifestations, which may lead to clinical misdiagnosis. Besides, a disease may have similar clinical manifestations to other diseases, which can also lead to clinical misdiagnosis. At this time, it is necessary to differentiate between the two diseases.

  • The significance of the dimension of the diagnosis basis of the disease is that the systematic and comprehensive data collection and evaluation in the emergency department are crucial to the emergency problem. The clinical datas analysis includes medical history, diagnostic points, laboratory examinations, impact studies, and physical examinations. And other auxiliary inspections in six aspects.

  • The dimension of disease treatment methods is based on the four-dimensional general disease ontology model, which adds three ontology of on-site first aid, on-the-way treatment, and hospital treatment. The current emergency medical service system in my country is a three-in-one model of pre-hospital emergency care, hospital emergency care, and critical care. Pre-hospital first aid is the medical treatment of emergency or (and) trauma patients at the scene and during transportation. Therefore, the ontology model is divided into three ontology, field first aid, on-the-go treatment, and hospital emergency, to introduce treatment measures in different treatment scenarios.

  • The dimension of disease prevention mainly refers to preventive measures against pathogenic factors. By taking effective preventive measures against diseases that may occur in the Winter Olympics Games, the probability of disease occurrence can be reduced.

Based on the above analysis and on the basis of the Winter Olympic Games emergency medical corpus, we have improved the four-dimensional general model. The improved ontology model is shown in Fig. 2.

Fig. 2.
figure 2

The improved model body

Full size image

The improved Winter Olympic Games emergency medical security ontology model based on the four-dimensional general ontology model not only meets the professionalism and accuracy of the medical ontology, but also meets the needs of the Winter Olympic Games emergency medical security scene. For example, in the follow-up question and answer system, input the clinical manifestations, you can get all the disease names, diagnosis basis, and first-aid measures related to the clinical manifestations, assist the doctor in timely diagnosis, and then take timely and effective first-aid measures.

3.4 Analysis and Definition of the Ontological Relationship of Emergency Medical Security for the Winter Olympic Games

Based on the reuse of the existing ontology model, this paper analyzes a large number of Winter Olympic emergency medical corpus manually and evaluates by experts, and determines that the relationship between the various ontologies of diseases is divided into two categories, namely, hierarchical and non-hierarchical relationships.

The hierarchical relationship mainly refers to the tree-like hierarchical structure between concepts. Each concept in the tree forms a subordinate relationship with other concepts. The upper concept is called the genus concept, and the lower concept is called the species concept, which reflects the inclusion relationship between concepts [14]. It mainly includes species relationship and instance relationship. For example, the sub-concepts of “disease” include “burn”, “frostbite”, “pneumothorax”, etc., among which “frostbite” also includes “freezing cold injury” and “non-freezing cold injury” Wait.

Non-hierarchical relationships reflect some semantic relationships between ontologies, and there are many types. The method of extracting non-hierarchical relationships on the ontology is to select a part of the text in the Winter Olympic Games emergency medical corpus to find the relationship between the two concepts. Analyze the relational words, and summarize the relations between related ontology. The non-hierarchical relationship between the disease and other ontology mainly includes:

  • The medical visit relationship, which is mainly to study the hospital and department to which the injury occurred.

  • the cause relationship, which mainly studies the etiology of the disease, that is, the direct factor or the cause of the disease. Indirect factors.

  • The examination relationship is mainly to study the examination items of the disease, including medical history, physical examination, laboratory examination and so on.

  • The clinical manifestations relationship, mainly studying the clinical manifestations and complications of diseases.

There are 7 kinds of relationships in total, and the relationships between each ontology and each other are shown in Table 1.

Table 1. The relationship between ontology
Full size table

3.5 Analysis and Definition of the Ontological Attributes of Winter Olympics Games Emergency Medical Security

Attributes mainly describe the internal structure of the ontology, which is a deep-level description of the ontology. After analysis, the main attributes of the core ontology are determined, which mainly include: the attributes of treatment ontology include treatment principles and basic treatment, and describe the guiding ideology and necessary treatment measures in the treatment process. The attributes of the diagnosis basis include: principle, operation method, and normal value, which is convenient for making accurate judgments of the disease. The attributes common to all ontology include the attributes common to all classes, aliases and definitions. Aliases refer to names other than the usual ones. Definition refers to a brief description of the basic attributes of the ontology.

3.6 Evaluation of Winter Olympic Games Emergency Medical Security Ontology

In this paper, the ontology evaluation stage is mainly based on the evaluation of the ontology proposed by GaGangemi et al. [15]. from the three dimensions of function, structure and usability. The structural dimension mainly considers the structural integrity and logic of the ontology, and mainly evaluates the grammar and formal semantics. The functional dimension considers the use of the built ontology, usually considering the application requirements and target tasks of the ontology, and is related to other functional indicators. The usability dimension is mainly the interaction between the application of the ontology and the user. Evaluate users’ understanding, use, maintenance and reuse.

3.7 Establishment of Emergency Medical Body for Winter Olympics

Experts will inspect all the ontology in the emergency medical security of the Winter Olympics Games in step 2 according to the evaluation criteria proposed in the ontology evaluation. For the ontology that meets the requirements, the neighborhood ontology is established, and for the ontology that does not meet the standard, it is transferred to the ontology analysis and definition stage. carried out. Until all the ontology meets the standard.

4 Instance Creation

Examples are the concretization of concepts and inherit the attributes of ontology. After a large number of examples are constructed, knowledge graphs or knowledge bases in related fields are gradually formed. In the later stage, it can be combined with the rule library and natural language processing to finally form the Winter Olympics Games emergency medical security knowledge graph. In this paper, ontology is divided into 8 categories, namely attributes, medical treatment, epidemiology, pathophysiology, clinical manifestations, diagnosis basis, treatment principles, and prevention. There are a total of 22 ontology subcategories, 7 types of relationships between ontology, and 7 ontology attributes.

In this paper, 15 kinds of diseases are extracted based on manual extraction. After manual proofreading, a total of 2646 concepts and 2064 concept relationship triples are extracted. The visual display of Winter Olympics Games emergency medical security knowledge is shown in Fig. 3.

Fig. 3.
figure 3

Winter Olympics emergency medical security knowledge graph

Full size image

5 Conclusion

In this paper, the existing ontology construction methods and ontology models are combined with specific Winter Olympics Games emergency medical security application scenarios, and the seven-step method and skeleton method are integrated to construct ontology, and on this basis, starting from the four-dimensional general medical ontology model, Combining the analysis of the application scenarios and corpus of the Winter Olympics Games emergency medical security, formulating rules and constructing a relatively complete Winter Olympics Games emergency medical security ontology model, laying the foundation for the construction of the follow-up Winter Olympics Games emergency medical security knowledge graph and the development of an intelligent question-and-answer system.