作者: Huiyan Wang , Yiyu Cheng
DOI: 10.1109/IEMBS.2005.1615774
关键词: Pulse diagnosis 、 Pulse (physics) 、 Patient diagnosis 、 Biomedical engineering 、 Medicine 、 Data mining 、 Traditional Chinese medicine 、 Bayesian network
摘要: The pulse diagnosis is one of the most important examinations in traditional Chinese medicine (TCM). Due to subjectivity and fuzziness TCM, quantitative systems or methods are needed modernize diagnosis. But up now, effective models that can classify types according waves automatically have not been reported, which undoubtedly limits practical applications clinical medicines. In this article, a novel system for was constructed based on Bayesian networks (BNs) build mapping relationships between types. results show obtains relative reliable predictions types, its predictive accuracy rate reach 84%, testifies model used our feasible be expected facilitate popular TCM