Decision Tree-Based Body Constitution Diagnosis System for Traditional Chinese Medicine

Abstract

This study aimed to establish a method for fast and accurate determination of body constitution types from the body constitution questionnaire (BCQ) by employing a decision tree model. The model was trained for 4 classes, namely, Yin-Xu, Yang-Xu, Phlegm and Blood Stasis, and Normal, and it achieved 67% accuracy for the testing dataset. The model also reduced the required number of BCQ questions from 44 to 3–6, depending on the responses. Lastly, we developed the Traditional Chinese Medicine (TCM) body constitution online diagnosis system using our model to collect data digitally and use it more practically and efficiently. This system can assist doctors to improve the diagnosis and treatment in TCM practice.