基于数据挖掘的《千金方》治疗腰痛用药规律研究  

Medication rules of Qianjin Fang for low back pain based on data mining

作　　者：于子涵 孙鼎[1] 苏励[1] YU Zihan;SUN Ding;SU Li(Longhua Hospitial,Shanghai University of Traditional Chinese Medicine,Shanghai200032,China)

机构地区：[1]上海中医药大学附属龙华医院,上海200032

出　　处：《上海中医药杂志》2020年第7期47-51,共5页Shanghai Journal of Traditional Chinese Medicine

基　　金：国家中医药管理局中医临床研究基地业务建设专项课题(JDZX2015193);上海中医药大学附属龙华医院第二批优秀青年人才临床能力提升计划项目(RC-2018-01-02);上海中医药大学附属龙华医院基于大数据与中医经典“知识元”的在线人工智能处方平台建设项目(KY1918)。

摘　　要：目的基于中医传承辅助系统软件,总结《千金方》治疗腰痛的遣方和用药规律。方法整理《千金方》中治疗腰痛的处方,采用关联规则Apriori算法、复杂系统熵聚类等无监督数据挖掘方法,分析处方中药物的使用频次及药物之间的关联规则。结果共筛选出38首处方;使用频次在前5位的药物为肉桂、甘草、杜仲、茯苓、当归,挖掘出核心组合14个,候选新处方7个。结论《千金方》所载的治疗腰痛方剂多补泻兼施、气血同治,重视温阳散寒,兼顾行气活血和补益肝肾,辅以补气利水和祛风解表。Objective To summarize the prescriptions and medication rules of Qianjin Fang(Essentially Treasured prescriptions) for treating low back pain based on the software of traditional Chinese medicine inheritance assistance system. Methods The prescriptions for the treatment of low back pain in Qianjin Fang were sorted out and the usage frequency of various Chinese herb medicines in the prescriptions and the association rules among Chinese herb medicines were analyzed by unsupervised data mining methods such as apriori algorithm of association rules and entropy clustering of complex system. Results A total of 38 prescriptions were selected and the Chinese herb medicines used frequently in the top 5 were Rougui(Cortex Cinnamomi), Gancao(Radix Glycyrrhizae), Duzhong(Cortex Eucommiae), Fuling(Poria), Danggui(Radix Angelicae Sinensis);14 core combinations of Chinese herb medicines and 7 new candidate drug groups were extracted. Conclusion The prescriptions for low back pain in Qianjin Fang mostly have the functions of tonifying and purging, simultaneously regulating qi and blood, and focus on warming yang to dissipate cold, promoting qi flow to activate blood, tonifying the liver and kidney, invigorating qi to promote urination and dispelling wind to relieve exterior syndrome.

关 键 词：《千金方》 孙思邈 腰痛 用药规律 数据挖掘 关联规则 聚类算法 

分 类 号：R289.5[医药卫生—方剂学]