Causal Model Based Analysis of TCM Syndrome Studies
基于因果模型的中医证候研究探讨

Dialectical treatment is a basic feature of TCM. Unfortunately, traditional dialectical methods are always characterized with subjectivity and ambiguity, which leaves many treatment experiences unrepeatable. To overcome the unneeded part of traditional methods, both objective and quantitative research methods are introduced to differentiate syndromes. Recently, causal inference has paid more attention to statistics, epidemiology, and data mining. Its two models: structural equation model and Bayesian network, constitute an objective and quantitative approach of TCM study. The two models have found increasing applications in the study of syndrome differentiation. In this article, authors analyze the principles and methods of building a structural equation model and the Bayesian network, and present a review of applications of the two models in syndrome differentiation.