基于R语言数据挖掘分析中医药治疗原发性失眠“症-药”结合规律*

作者：夏增嫒，王 丹，张芮萌，王 戈，唐启盛

单位：北京中医药大学第三附属医院，北京 100029

引用：

DOI：10.13862/j.cn43-1446/r.20251229.002

PDF： 下载PDF

摘要：目的：采用数据挖掘技术，基于失眠患者不同症状分类，探讨原发性失眠患者的用药规律。方法：检索中国知网（CNKI）、万方数据知识服务平台（Wanfang Data）、维普网（VIP Data）、中国生物医学文献数据库（CBM）中2009年1月至2024年12月期间治疗原发性失眠的医案及临床研究，用Excel软件构建数据库，并对药物功效、性味归经进行频数分析。运用R对使用频数排名前20的中药进行关联规则分析与聚类分析。结果：最终纳入医案142则，涵盖195味中药。频数分析中功效以补虚药为主，清热药与安神药次之；药性以寒多见，温、平次之；药味以甘多见，苦、辛次之；归经以肝经多见，心、肺、脾、胃、肾经次之。关联分析结果显示，“易入寐，茯苓寅柴胡”“眠浅易醒，龙骨寅牡蛎”“彻夜不寐，茯苓寅酸枣仁”关联度最高。聚类分析结果共4类。结论：逍遥散合酸枣仁汤加减可用于不易入寐型失眠，补中益气合柴胡加龙骨牡蛎汤加减可用于眠浅易醒型失眠，天王补心丹加减可用于彻夜不寐型失眠，黄连温胆汤加减可贯穿各种类型的失眠。

关键词：失眠；“症-药”结合；用药规律；R语言；数据挖掘

Abstract：Objective: To explore the medication rules for primary insomnia patients based on the classification of different symptoms using data mining technology. Methods: Medical records and clinical studies on the treatment of primary insomnia from January 2009 to December 2024 were retrieved from databases including China National Knowledge Infrastructure (CNKI), Wanfang Data Knowledge Service Platform (Wanfang Data), VIP Database (VIP Data), and China Biomedical Literature Database (CBM). A database was established using Excel software, and frequency analysis was performed on the efficacy, properties, flavor, and meridian tropism of the herbs. R software was employed to conduct association rule analysis and cluster analysis on the top 20 most frequently used Chinese herbal medicines. Results: A total of 142 medical cases were included, covering 195 types of Chinese medicines. Frequency analysis showed that tonifying herbs were the most commonly used in terms of efficacy, followed by heat-clearing herbs and mind-calming herbs. In terms of medicinal properties, cold-natured herbs were the most frequent, followed by warm and neutral-natured herbs. In terms of medicinal flavor, sweet flavor was the most common, followed by bitter and pungent flavors. In terms of meridian tropism, the liver meridian was the most frequently targeted, followed by the heart, lung, spleen, stomach, and kidney meridians. Association rule analysis identified the strongest correlations: "difficulty falling asleep, Fuling (Poria) → Chaihu (Bupleuri Radix) ", "light sleep with easy awakening, Longgu (Mastodi Ossis Fossilia) → Muli (Ostreae Concha) ", and "sleeplessness throughout the night, Fuling (Poria) → Suanzaoren (Ziziphi Spinosae Semen) ". Cluster analysis resulted in four categories. Conclusion: It is suggested that Modified Xiaoyao San combined with Suanzaoren Tang (逍遥散合酸枣仁汤加减) can be used for insomnia with difficulty falling asleep, Modified Buzhong Yiqi Tang combined with Chaihu Jia Longgu Muli Tang (补中益气合柴胡加龙骨牡蛎汤加减) is suitable for light sleep with easy awakening, Modified Tianwang Buxin Dan (天王补心丹加减) is applicable for sleeplessness throughout the night, and Modified Huanglian Wendan Tang (黄连温胆汤加减) can be used for various types of insomnia.

Key words：insomnia; "syndrome-medication" combination; medication rules; R language; data mining

发布时间：2025-12-30

点击量：15