MINING OF CLASSIC CHINESE MEDICINE PRESCRIPTIONS FOR CARDIOVASCULAR AND CEREBRO-VASCULAR DISEASES FOR PRECISION MEDICINE

Abstract

Traditional Chinese medicine (TCM) prescription has the characteristic of “one person, one prescription”. Mining the rules of its prescriptions can be helpful for clinical diagnosis and treatment and reflects “Precision Medical”. Taking the classic Chinese medicine prescriptions for cardiovascular and cerebrovascular diseases as samples, the author discussed the mining process of prescription rules and the methods and techniques of data preprocessing. The machine learning method is used to construct a mining model of prescription rules based on set relations, and the similarity between prescriptions is calculated, and the “Mean-Link” clustering algorithm is used for clustering. The results showed that 14 kinds of TCM prescriptions for cardiovascular and cerebrovascular diseases, such as hypertension and coronary heart disease, were grouped into four categories in terms of prescription compositions among which the cluster composed of HFP (Heart Failure Prescription), SHYF (Siheyifang), WSJXD (Wenshen Jiuxin Decoction) poly as one category, prescription HP1 (Hypotensive Prescription), SSP (Sinus Syndrome Prescription), YQJND (Yiqi Jiannao Decoction) cluster into one category, the similarity of the two types is 0.21, is the highest value. In terms of prescription indications, they are clustered into four groups, among which the HFP and WSJXD are clustered into one category, and the similarity is 0.60, the highest value. The results of the comprehensive composition and indication clustering can be seen: (1) BWJYD (Bawei Jiangya Decoction), TMGTD (Tianma Gouteng Decoction); (2) HP1, AP (Anemia Prescription), MVD (Myocardial Vital Drink); (3) HFP, WSJXD; (4) SSP, YQJND. The treatment prescriptions between the two diseases in these four groups have the significance of mutual reference and reference.