An approximate reasoning using probability-possibility transformation for medical expert systems

We present PROPO-an experimental expert system using probability-possibility transformation for pulmonary tuberculosis diagnosis. We describe a transformation between the probabilistic logic and the possibilistic logic as a suitable reasoning engine for medical expert systems. Further more, a structure of the system including knowledge base and inference of the system is proposed. An example of performance of the system testing the knowledge base on pulmonary tuberculosis diagnosis is illustrated.