Research On Electronic Patient Record Knowledge Base Based On Ontology

In recent years, study of Electronic medical records has become a hot issue of the construction of hospital informationization at home and abroad. Electronic medical records at present can only combine external characteristics information (such as name, gender, age, number of medical record, number of paper medical record, time) to inquire about patient's medical records in order to facilitate the use of doctors'browser to read. But it cannot realize semantic retrieval of the contents of characteristics of information (such as chief complaint, history of present illness, past history, personal history, family history, Physical examination, assistant Examination, initial diagnosis, final diagnosis, examination results, treatment process, change of condition) and potential pathological analysis. Electronic medical records system at present can be said an efficient management system, but it hasn't realized intelligence analysis function and it doesn't bring doctors the help of better medical record analyze.With the deepening of the application of electronic medical records, how to inquire about medical history, examination, diagnosis and content information in report card from electronic medical records in order to find valuable information and knowledge has become a hot issue of the development of electronic medical record system.This paper does a data conversion of the existing data and establishes electronic medical record repository based on ontology. This paper first built a electronic medical record ontology model with"seven steps"method and Protégé, in order to normalize and unify various concepts and terminology; then extracts some data from electronic medical records to do data mining through methods of rough set and decision tree, then establish a rule base of electronic medical records; finally this paper researches patient's information and semantic relations among these information and experts'empirical knowledge in this field , to reason and find potential lesions and effective early treatment.