Related Experiment Videos
A computer method for validating traditional Chinese medicine herbal prescriptions.
1Department of Computational Science, National University of Singapore Blk SOCI, Level 7, 3 Science Drivf 2, Singapore.
The American Journal of Chinese Medicine
|June 25, 2005
Summary
A new computer program uses a support vector machine (SVM) to predict Traditional Chinese Medicine (TCM) prescriptions. This AI tool accurately identifies valid TCM herbal formulas, aiding in their analysis and application.
Area of Science:
- Computational biology
- Pharmacology
- Traditional Chinese Medicine (TCM)
Background:
- Traditional Chinese Medicine (TCM) utilizes multi-herb prescriptions with complex therapeutic actions.
- Formulation of TCM prescriptions traditionally relies on herb properties: pharmacodynamic, pharmacokinetic, toxicological, and physicochemical.
Purpose of the Study:
- To develop a computer program for predicting the validity of multi-herb TCM prescriptions.
- To leverage statistical learning methods for analyzing TCM herbal formulas.
Main Methods:
- A support vector machine (SVM) model was developed and trained using 575 known TCM prescriptions and 1961 non-TCM recipes.
- The SVM model was tested on 72 TCM prescriptions and 5039 non-TCM recipes, and subsequently on 48 recent TCM prescriptions.
Main Results:
- The SVM system correctly classified 73.6% of TCM prescriptions and 99.9% of non-TCM recipes in initial testing.
- A further test showed 68.7% accuracy for classifying recently published TCM prescriptions.
- Classification accuracies are comparable to those achieved in other biological systems using SVM.
Conclusions:
- The study demonstrates the potential of support vector machine (SVM) for the computational analysis of Traditional Chinese Medicine (TCM) prescriptions.
- This approach can aid in verifying the authenticity and efficacy of TCM herbal formulas.