An approach to calculation and visualization of efficacies of Traditional Chinese Medical formulae based on a semantic network

摘要

Traditional Chinese Medicine (TCM) is widely used and tested in ethnic communities for a long history. A large number of TCM formulae, as the centralized expression of “theory, cure, recipe, and medicine”, were preserved and recorded in way of relatively completed ingredients and lack of well-noted efficacies. The main objective of this paper is to translate dosage of each herb in a TCM formula into intensity of its efficacies in a quantitative and visual way. In this study, firstly, we estimated the intensity between a herb and its efficacies quantitatively based on clinical case data, which can reflect the difference between main and secondary efficacies of a herb. Then, to calculate and visualize the interaction between different efficacies from different herbs, we have proposed a method based on the Efficacy Semantic Network (ESN). It includes three steps: (1) Simple cumulation; (2) Adjustments; (3) Main efficacies selection. We tested on a dataset of 70 TCM formulae, results show that maximum performance measure reaches 0.818 which prove that the proposed approach could help standardize existing and predict potential efficacies of TCM in a quantitative way. It shows a great potential of semantic technologies to represent, calculate and visualize the TCM knowledge. However, more efforts are still required to investigate how to represent other fuzzy and rule-based knowledge and bring more data of TCM formulae for extensive tests.