Case based reasoning using fuzzy set theory and the importance of features in medicine

We have developed an expert system for diagnosis of lung diseases using case based reasoning (Nguyen et al., 2000). At present, the case base of this system consists of 1026 patient records of 18 types of lung diseases and it is applied successfully at the Vietnam National Institute of Tuberculosis and Lung diseases. An improvement of the case based reasoning for medical diagnosis using fuzzy set theory (Nguyen et al., 2000) is proposed by including the importance of features. A feature in the former work is that a symptom and all important symptoms for the disease take value 1, that means that the important degree of all symptoms are the same. In this paper an important degree for each symptom according to each disease is included. The fitness of this method is confirmed by experiments in real medical applications