摘要
[主要目的]数据挖掘分析李济仁治疗胃系疾病方药规律。[资料来源]2014年12月至2015年3月皖南医学院附属弋矶山医院国医大师工作室门诊。[选择文献量及依据]中医和中西医结合临床确诊或经胃镜确诊胃系疾病的病案处方。复诊次数≥2次,只选择一次复诊病案。排除信息不全、方药组成不完整及临床疗效不显著病案。[数据提炼规则及应用方法]筛选处方录入"中医传承辅助系统(V2.5)",2位助手审核。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,分析方药使用频次及药味关联规则。药味频次:方剂药味按出现频次排序;药症关联:症状按出现频次排序;组方规律(支持度20%,置信度0.9):药味组合按出现频次排序,关联分析所得药对;聚类分析新药味组合(改进互信息法、复杂系统熵聚类):聚类分析前,选择合适相关度和惩罚度,"提取组合",发现新药味组合(基本算法示无监督熵层次聚类),网状结构图展示。[数据综合得出结果与结论]纳入样本80例,处方120首,用药117种。使用频次>25的前10味:炒白术>茯苓>蒲公英>乌药>木香>佛手>瓦楞子>香附>金钱草>全蝎。药症关联前7位症状:嗳气>腹胀>胃脘胀痛>大便不成形>胃脘隐痛>泛酸>嘈杂。16味联系紧密核心用药:黄芪、炒白术、茯苓、木香、瓦楞子、姜半夏、竹茹、八月札、佛手、香附、蒲公英、金钱草、沉香、乌药、延胡索、全蝎。聚类新药味组合共6对。[未来展望]多中心、大样本长时间调查研究,更好挖掘、传承名老中医经验。
[Objective] Professor LI Jiren treatments of gastric diseases drug law.[Data sources] From December 2014 to 2015 March in Wangnan medical college affiliated Yijishan hospital of Chinese medicine master Studio elinic.[Select the volume of literature and on the basis of] Chinese medicine and western medicine in clinical diagnosis of gastric disease prescription, or by electronic gastroscope for definite diagnosis of gastritis prescription.(1)The case of incomplete information and incomplete prescription.(2)To ensure accurate statistical results, were more than 2 times, only a referral medical record.(3)The repeated medical records.(4)The clinical curative effect of the medical record is not significant. [Data extraction rules and application methods] The screening prescription "TCM inheritance system (V2.5)" software, the entry is complete, in charge of the audit data by 2 assistants, to ensure the accuracy of the data. The application of "TCM inheritance system (V2.5)" software "data analysis" module "Prescriptions" ,function, drug law mining. Analysis of drug frequency, drug disease association, prescription law,new clustering.[The data obtained comprehensive results and conclusion] In 80 samples, 120 of the first 117 kinds of prescription medication. Using frequency〉25,came in before 10 flavour: stir fried white atractylodes rhizome〉Poria〉dandelion〉Lindera〉woody〉Bergamot〉conc ha Arcae〉nutgrass galingale rhizome〉Christina Loosestrife herb〉scorpion.7 symptoms of drugs in association: belching abdominal distension epigastric pain shapeless stool, epigastric pain, pantothenic acid, noisy.The drug combination data of 2805,including 16 kinds of traditional Chinese medicine. Mining to get a total of 6 new prescription combination clustering. This study based on TCM inheritance auxiliary system platform, using drugsregularity of professor LI Jiren data mining research,from objective data in professor LI in the treatment of gastric disease of department of medical records, prescription analysis, LI Jiren professor and gastrointestinal disease treatment experience in-depth exploration and inheritance to provide reference.[Future Prospect] Can be carried out muhicenter, large sample for a long period of time investigation and study, In order to better mining and provide new methods and ideas of inheritance of experience of famous doctor.
出处
《实用中医内科杂志》
2016年第5期1-4,共4页
Journal of Practical Traditional Chinese Internal Medicine
基金
国家中医药管理局重点学科中医痹病学建设项目(国中医药人教发[2012]32号)~~
关键词
胃病
慢性胃炎
李济仁
老中医经验
中医传承辅助系统(V2.5)
关联规则
Apriori算法
复杂系统熵聚类
无监督数据挖掘
药味频次
药症关联
组方规律
支持度
置信度
网状结构
相关度
惩罚度
循证医学
系统综述
stomach trouble
chronic gastritis
LI Jiren(李济仁)
the old doctor of traditional Chinese medicineexperience
inheritance of traditional Chinese medicine auxiliary system (V2.5)
association rules
the apriorialgorithm
complex system entropy clustering
unsupervised data mining
ingredients frequency
medicinedisease association
formula
support
confidence
the reticular structure
relevance
punishment
evidence-basedmedicine
systematic review