Research And Construction Of Tongue Image Diagnosis And Treatment System Based On Knowledge Graph

With the rapid development of society and the continuous improvement of the quality of life,people pay more and more attention to health.As a Chinese national quintessence,Chinese medicine is increasingly favored by people.Traditional Chinese medicine remote diagnosis products with the theme of TCM diagnosis continue to emerge,but the existing TCM tongue diagnosis system has limited ability to obtain patient symptom information,which affects the diagnosis effect.With the advent of the era of big data,the continuous maturity of artificial intelligence technology and the continuous improvement of computer technology make the integration of traditional Chinese medicine and computer technology become inevitable.The knowledge map of Chinese medicine can integrate the knowledge of traditional Chinese medicine and be understood and applied by computers.The tongue-and-eye diagnosis and treatment system based on the knowledge map can give the dialectical results of TCM and the recommended drugs according to the symptoms input by the user and the uploaded tongue-like photos to simulate the Chinese medicine diagnosis process.In this thesis,the construction of TCM knowledge map is deeply studied,and a TCM knowledge map is constructed and applied to the tongue diagnosis system.The main research contents are as follows:First,with the concept of ontology,the model layer of TCM knowledge map is constructed.When designing the construction mode layer,referring to the semantic type and TCM professional knowledge of the domestic "Chinese medicine language system",the basic types,attributes and relationships of knowledge maps were determined,and the conceptual model of TCM knowledge atlas was formed.Second,the natural language processing technology is used to construct the data layer of the TCM knowledge map.Through the processing and integration of different data sources(original Chinese medicine knowledge base,diagnostic criteria for TCM symptoms,Baidu Encyclopedia,electronic medical records),entity identification,attribute extraction,relationship extraction and knowledge fusion are completed.In the entity identification process,the CRF is used to identify the electronic medical records,use the CRF++ tool to extract the entities in the electronic medical record,use the crawling technology to realize the attribute extraction,and use the rule template and themachine learning model Att-BiLSTM to achieve the relationship extraction.Based on attribute vector-based entity alignment method,knowledge fusion of different data sources is carried out,and then a knowledge map of Chinese medicine is constructed by using Neo4 j tool.Finally,the design of the tongue diagnosis system based on knowledge map is implemented.Using Spring Boot to complete the system framework,and using Spring Data Neo4 j and Spring Data JPA to connect Neo4 j and MySQL data sources,in the tongue diagnosis section based on user input symptoms and tongue diagnosis results to achieve TCM syndrome,and finally give advice Complete the diagnosis process with medication and precautions.In order to evaluate the effect of the system,100 users were selected for testing.The test results found that the system can realize the tongue image diagnosis based on the knowledge map on the one hand,and the speed of the tongue-diagnosis in the tongue diagnosis process compared with the traditional relational database.Significant improvements,especially when querying across multiple relationships,is even more significant.