Using semantic technologies for an intelligent medical trainer

Abstract

Diagnosis is the basis of medicine. Medical schools must evaluate their students competencies in clinical reasoning in order to assess how medical knowledge is applied by the student. It is also necessary that this knowledge is available and shared by health professionals. For sharing and representing the knowledge exist semantic technologies as terminologies and ontologies. Based on an inference system we extracted from a medical knowledge base a list of disease ontologies, and their related signs/symptoms and diagnostic test. We used the cosine similarity metric to find closeness between diseases. A diagnosis training module was developed, where a disease and its findings are shown to the medicine student. Then, the student must select the corresponding disease from a list of four possible similar diseases. Ontologies work great in the representation of medical knowledge and its applications, however there is little evidence of its uses on medical student training.