Design And Implementation Of Ontology Research And Auxiliary Diagnosis Reasoning System In Gestational Diabetes Mellitus

With the continuous in-depth study of artificial intelligence,ontology is more and more used in the representation and sharing of domain knowledge.How to apply ontology to the medical field is one of the current research hotspots,with its focus more on knowledge sharing and query than on ontology-based reasoning.In China,gestational diabetes mellitus(GDM)is a high risk symptom in pregnant women,the application of ontology to diabetes is at its preliminary stage.Obstetricians generally make judgments based on relevant guidelines and combined with clinical experience in the diagnosis process.This diagnosis method is more dependent on the clinical experience of doctors,and there is a risk of misdiagnosis for inexperienced young doctors.To tackle this problem,this paper adds ontology evaluation to the established“seven steps”ontology construction method,and constructed GDM ontology in a top down manner.OWL(Web Ontology Language)has been used to describe the ontologies,the classes,attributes,individuals and their relationships are constructed via protégé(a ontology construction tool).Based on the constructed GDM ontologies,rules of inference for diagnosis of GDM have been constructed via the reasoning engine,which helps to unveil the hidden knowledge.Finally,A diagnosis tool prototype based on J2 EE and Jena semantic framework has been developed.The preliminary experimental results show that the auxiliary diagnosis tool constructed by "gestational diabetes ontology and rules of inference" can judge the patient’s GDM.This paper explores the construction of the domain ontology of GDM and builds a reasoning tool to assist clinicians’ diagnosis through Automatic reasoning,which can be used as an aid for obstetricians to diagnose the symptoms of GDM in pregnant women and as a reference for the follow-up ontological study in obstetrics.