Traditional Chinese medical diagnosis based on fuzzy and certainty reasoning

Diagnosis in traditional chinese medicine (TCM) is the foundation of every TCM clinical branch. It is related to basic TCM theories, and guides TCM treatments. In this paper methodologies developed for quick and reliable TCM diagnosis are described. The diagnosis process is divided into two stages. In the first stage the scope of possible diseases and syndromes of a patient will be quickly narrowed down by a modifiable multilevel fuzzy sieving method. This sieving method follows TCM differentiation principles. In the second stage diseases and syndromes will be reliably determined by a multi-agent diagnosis helping system (MADHS) that simulates cooperated experts and makes joint decisions. Fuzziness and uncertainty are incorporated into decision trees to form the reasoning mechanism of agents. A prototype of MADHS has been implemented and successfully tested. It is anticipated that the system and methodologies may be widely used in computer aided diagnosis and TCM education software.