Research On Construction And Application Of Knowledge Graph Of Neurodegenerative Diseases Based On Ontology

Neurodegenerative diseases are common in the elderly population.The early symptoms of these diseases are often not obvious,and the disease duration is long.Early detection,diagnosis and treatment of the disease are important to slow down the disease progression.Building a disease knowledge graph based on ontology can provide a structured representation of disease knowledge.Based on which a question answering system,assisted diagnosis system and other application systems can be built to help the early diagnosis and treatment of diseases.Objectives:Firstly,to construct disease ontology by extracting disease diagnosis and treatment knowledge of neurodegenerative diseases,to provide reference for the construction of disease ontology in Western medicine,and to provide a schema layer for the knowledge graph of neurodegenerative diseases.Secondly,on the basis of the ontology,we extract disease diagnosis and treatment knowledge from clinical guidelines,medical textbooks and other knowledge sources to build a knowledge graph of neurodegenerative diseases for clinical scenarios.Finally,we design an automatic question answering system based on the knowledge graph to provide references for the research of question answering system based on the knowledge graph.Methods:Firstly,the current research status of disease ontology and knowledge graph construction was investigated by literature research,and then the data sources and construction process of ontology and knowledge graph construction were clarified.By referring to the seven-step method proposed by Stanford University and reusing resources such as UMLS,we used clinical guidelines and medical textbooks as data sources to construct disease ontology.Using the ontology as the schema layer of the knowledge graph,clinical guidelines,medical textbooks and web data as data sources,and Neo4j as the database for knowledge graph storage,the knowledge graph of neurodegenerative diseases is constructed through the process of knowledge extraction,knowledge fusion,quality assessment and knowledge storage.Results:In this study,a neurodegenerative disease ontology and knowledge graph for clinical diagnosis and treatment were constructed,and an automatic question answering system was designed based on the knowledge graph.The disease ontology contains 10 types of concept,17 types of object attribute and 7 types of data attribute;the knowledge graph of neurodegenerative diseases contains 2012 entities,including 668 clinical manifestation entities,199 disease examination item entities,205 disease examination result entities,249 disease treatment drug entities and 217 non-drug treatment entities.The automatic question answering system for neurodegenerative diseases designed based on the knowledge graph parses the user’s question by deep learning method,then converts the user’s question into cypher query language by template,and returns the specific answer to the user.Conclusion:This study proposes an ontology-based knowledge graph construction framework,and constructs a neurodegenerative disease ontology and knowledge graph.The ontology construction provides a reference for related disease ontology construction research in the field of disease,and provides resources that can be reused.Through the knowledge graph construction,the knowledge related to clinical diagnosis and treatment of neurodegenerative diseases is structurally represented.The knowledge of clinical manifestations,drug treatment,and non-drug treatment of diseases is classified and graded in the knowledge graph,and the types of entities such as examination items,examination results and treatment and the semantic information in them are comprehensively extracted.The construction of the knowledge graph promotes the management and utilization of disease knowledge,and provides reference for related researches.The design of the Q&A system clarifies the steps and methods of the implementation of the Q&A system,and provides a reference for the research of the disease Q&A system.