文献类型：期刊文献

中文题名：陈修园《医学实在易》内服复方用药规律复杂网络关联系统综述

英文题名：A Summary of CHEN Xiuyuan's(陈修园) "Medical Reality"(医学实在易) in the Complex Network Association System of Oral Administration

作者：彭希宸[1];李廷保[2];杨鹏斐[2]

第一作者：彭希宸

机构：[1]甘肃中医药大学中医临床学院,兰州730000;[2]甘肃中医药大学中医外科学教研室,兰州730000

第一机构：甘肃中医药大学中医临床学院

年份：2018

卷号：32

期号：9

起止页码：16

中文期刊名：实用中医内科杂志

外文期刊名：Journal of Practical Traditional Chinese Internal Medicine

基金：甘肃中医药大学教学改革立项课题(2013-ZB06)~~

语种：中文

中文关键词：用药规律;陈修园;医学实在易;内服复方;老中医经验;中医传承辅助系统(V2.5);复杂网络关联;无监督熵层次聚类算法;使用频次;循证医学;系统综述

外文关键词：medication rule;CHEN Xiuyuan（陈修园）;medical practice（医学实在易）;internal service compound;old Chinese medicine experience;TCM inheritance auxiliary system （V2.5）;complex network association;unsupervised entropy hierarchical clustering algorithm;frequency of use;evidence-based medicine;systematic review

摘要：[主要目的]数据挖掘分析陈修园《医学实在易》内服复方用药规律规律。[资料来源]陈修园《医学实在易》。[选择文献量及依据]纳入药物组成齐全的中药内服复方。排除药名不全的方剂、外用方剂、单味药方、未注明药物组成的方剂。[数据提炼规则及应用方法]筛选出方剂录入"中医传承辅助系统(V2.5)"。通过该软件"方剂分析"功能,研究组方用药规律;"频次统计"每味药出现频次排序。"规则分析"(支持度个数45,置信度0.8)药物间关联规则,聚类分析核心算法及复杂系统熵聚类分析(相关度8,惩罚度2),挖掘组合新方。[数据综合得出结果与结论]纳入内服复方253首,用药234味。使用频次>15的有23味药,前6位是甘草、茯苓、人参、白术、生姜、当归。药物组合出现频次前3位是"人参+甘草";"白术+甘草";"甘草+茯苓"。无监督熵层次聚类算法得到4个新处方。[未来展望]运用大数据、循证医学理念,强化互联网思维,发挥计算机人工智能作用,结合临床实际分析灵活应用关联规则和聚类研究得到组方规则、核心组合和新处方,能更好、更全面地继承古代名医学术思想及经验。
[Main purpose] Data mining analysis CHEN Shuyuan＇s ＂Medical Reality＂ is the law of oral administration of compound drugs. [Source] CHEN Xiuyuan ＂Medical Reality＂. [Select the amount and basis of the literature] Incorporate the Chinese medicine oral administration compound with complete drug composition. Except for prescriptions with incomplete drug names, topical prescriptions, single-flavor prescriptions, and prescriptions without drug composition. [Data refinement rules and application methods]Screen out the prescriptions to enter the ＂Chinese Medicine Heritage Support System（V2.5）＂. Through the software ＂prescription analysis＂ function, study the prescriptions of the prescriptions; ＂frequency statistics＂ frequency of each patch appeared. ＂rule analysis＂（support score 45, confidence 0.8） drug association rules,cluster analysis core algorithm and complex system entropy cluster analysis（correlation degree 8, penalty degree 2）, mining combination new party. [Data synthesis results and conclusions] 253 oral and compound prescriptions, 234 flavors. There are 23 flavors with frequency 〉15. The top 6 are licorice, wolfberry, ginseng,atractylodes, ginger and angelica. The top three frequency of drug combinations are ＂ginseng＋licorice＂; ＂baizhu＋licorice＂; ＂licorice＋Fuling＂. The unsupervised entropy hierarchical clustering algorithm obtained four new prescriptions. [Future Prospect] Using big data and evidence-based medicine concepts,strengthening Internet thinking, exerting the role of computer artificial intelligence, and combining clinical practice analysis with flexible application of association rules and clustering research to obtain group rules,core combinations and new prescriptions, can be better, More comprehensive inheritance of ancient medical ideas and experience.