基于数据挖掘对血管性痴呆方药组成规律的文献研究  

The study of medicine composition rule on treating vascular dementia based on data mining

作　　者：李强 耿秀超 于文涛 宫志强 鲁琴 靳林静 

机构地区：[1]河北中医学院针灸推拿学院,河北石家庄050200 [2]河北中医学院国医堂,河北石家庄050091 [3]河北中医学院基础医学院/河北省心脑血管病中医药防治重点实验室,河北石家庄050200 [4]河北中医学院信息技术中心,河北石家庄050200

出　　处：《中医临床研究》2018年第6期34-37,共4页Clinical Journal Of Chinese Medicine

基　　金：河北省中医药管理局科研计划资助项目(NO:2016014)

摘　　要：目的:从文献研究角度探讨临床治疗血管性痴呆(VD)方药组成规律。方法:利用CNKI数据库检索到中医药治疗VD文献101首,运用"中医传承辅助系统"对该文献方药组成进行数据挖掘分析,挖掘方法包括药物出现频次统计、药物组合关联规则分析、熵聚类分析。结果:101首方剂中共涉及中药167味,其中使用频次超过10次以上的药物共计33味,按照频次由多至少前10味药分别为石菖蒲、黄芪、丹参、川芎、远志、茯苓、何首乌、熟地、红花、当归、地龙;当支持度为20条件下,共得到药对14个;在置信度≥0.9的条件下,共得到药对1个;选择相关度为8,惩罚度为2,进行聚类分析,得到关联系数大于0.02的药对32个;以改进的互信息法分析结果为基础,按相关的和惩罚度约束,基于复杂系统聚类,得到药物核心组合34个;运用无监督熵层次聚类算法,得到17个新处方。结论:活血、补肾、化痰为临床治疗VD最常用的药物,经数据挖掘得到的药物组合和配伍,对VD的临床治疗和方药选择可提供有益的线索。Objective:To discuss medicine composition rule on treatment of vascular dementia from the perspective of literature research.Methods:101 Chinese medicine treatment literatures of vascular dementia were retrieved by CNKI database,of which the medicine composition was data mined by Traditional Chinese Medicine Inheritance Support System,the method included prescription drugs frequency statistics,association rules analysis,and entropy clustering analysis.Results:33 kinds drugs were used over 10 times in 167 kinds of Chinese medicine drugs of 101 prescriptions,and according to the frequency,the first 10 TCM medicines were stone calamus,astragalus membranaceus and salvia miltiorrhiza,rhizoma ligustici wallichii,polygala,poria cocos,fleece-flower root,prepared rhizome of rehmannia,safflower,Chinese angelica,earthworm respectively;14 couple-medicines were included under the condition of 20 support degree;1 new prescription was got by using the improved mutual information method,and the regular entropy clustering of complicated system,and the algorithm of unsupervised entropy clustering.Conclusion:The most common used drugs for clinical treatment of vascular dementia were invigorating the circulation of blood and the kidney,and eliminating phlegm,the clinical treatment of VD and choosing of the formulas could provide useful clues by the drug combination and compatibility of data mining.

关 键 词：血管性痴呆 数据挖掘 中医传承辅助系统 

分 类 号：R277.7[医药卫生—中医学]