文献类型：期刊文献

中文题名：基于隐结构结合关联规则的宫颈癌症-证-方-药分析

英文题名：An analysis of symptoms-syndrome-prescription-drug of cervical cancer based on hidden structure plus association rules

作者：胡安鑫[1];梁永林[1];马程[1];孙云娇[1];连冰冰[1];陈俞君[1];杨双琪[1];杨亮亮[1]

第一作者：胡安鑫

机构：[1]甘肃中医药大学,甘肃兰州730000

第一机构：甘肃中医药大学

年份：2023

卷号：15

期号：30

起止页码：54

中文期刊名：中医临床研究

外文期刊名：Clinical Journal Of Chinese Medicine

基金：国家级大学生创新训练项目(202110735001);校级大学生创新训练计划项目。

语种：中文

中文关键词：宫颈癌;关联规则;隐结构模型;证候;中药

外文关键词：Cervical cancer;Association rule;Hidden structure model;Syndrome;TCM medicinal materials

摘要：目的:文章运用隐结构结合关联规则,探讨名老中医医案中宫颈癌的常见证候和用药,为放化疗前后中医药治疗宫颈癌提供参考。方法:筛选归纳名老中医辨治宫颈癌医案的数据样本,使用隐结构法的Lantern网络版软件,基于BI算法,对整理后的中医证候进行模型学习及诠释;以中医专业知识为根基、综合聚类为方法探索隐变量,依据宫颈癌的症状规律归纳证型并推测相关方剂;利用SPSS 26.0软件对频次>13的中药进行系统聚类分析并由中药药组推断对应证型;后运用SPSS Molder18.0软件,对频次>13的中药使用关联规则得到治疗本病不同证型的药对。结果:由93份合格医案、141个证候构建隐结构模型,得到26个隐变量,聚类后为5个常见证型,分别是热毒证、瘀血证、脾虚湿盛证、肝肾阴虚证、湿热瘀阻证;频次>13的中药系统聚类为10个药组并由此得到5种常见证型,与上文一致;最后运用SPSS Molder 18.0软件得到上文5个证型的常用药对,佐证了方剂推断的正确性。结论:中医辨治宫颈癌的常见证型为热毒证、瘀血证、脾虚湿盛证、肝肾阴虚证、湿热瘀阻证,其对应方剂为逍遥散、六味地黄丸、参苓白术散、清化煎、益气养阴解毒汤。文章研究结果与相关中医理论规律基本吻合,为中医临床医者辨治宫颈癌提供了思路借鉴。
Objective:In this paper,the hidden structure combined with association rules were used to explore the common symptoms and medications of cervical cancer in the medical cases of famous veteran TCM practitioners,so as to provide a reference for TCM treatment of cervical cancer before and after radiotherapy.Methods:The data samples of TCM medicinal cases from famous veteran TCM practitioners for cervical cancer differentiation and treatment were screened and summarized,and the model learning and interpretation of the sorted syndroms were carried out by Lantern online software of hidden structure method based on BI algorithm.Based on the expertise of traditional Chinese medicine,the method of comprehensive clustering was used to explore hidden variables,and the syndrome types were summarized and related prescriptions were speculated according to the symptoms of cervical cancer.SPSS 26.0 software was used for systematic cluster analysis of Chinese medicinal materials with frequency>13,and corresponding syndrome types were inferred from Chinese medicinal materials groups.The association rule analysis by SPSS Molder 18.0 was performed to analyze the traditional Chinese medicine with frequency>13 to obtain the pairs of medicines for treating different syndrome types of the disease.Results:The hidden structure model was constructed from 93 qualified medical records and 141 syndromes,and 26 hidden variables were obtained,which were clustered into five common types,namely,heat toxicity syndrome,blood stasis syndrome,spleen deficiency and dampness excess syndrome,liver-kidney Yin(阴)deficiency syndrome,and dampness-heat stasis syndrome.The systematic clustering of chinese medicinal materials with frequency>13 was divided into 10 drug groups and 5 kinds of common types were obtained,which was consistent with the above.Finally,SPSS Molder 18.0 software was used to obtain the common drug pairs of the above 5 syndrome types,which proved the correctness of the prescription inference.Conclusion:The common types of TCM differentiation and treatment of cervical cancer are heat toxicity syndrome,blood stasis syndrome,spleen deficiency and dampness excess syndrome,liver-kidney Yin deficiency syndrome and dampness-heat stasis syndrome.The corresponding formulas are Xiaoyao San(逍遥散),Liuwei Dihuang Wan(六味地黄丸),Shenling Baizhu San(参苓白术散),Qinghua decoction(清化煎),Yiqi Yangyin Jiedu decoction(益气养阴解毒汤).The results of this study are basically consistent with the relevant theoretical laws of traditional Chinese medicine,which provides a reference for TCM clinicians to identify and treat cervical cancer.