基于数据挖掘技术分析当代中医名家痤疮验方经验研究

发布时间：2018-04-11 11:26

本文选题：痤疮 + 医案　；参考：《南京中医药大学》2016年博士论文

【摘要】：[目的]病证结合乃当代中医临床的主要模式,“审证求机”在辨证论治过程中最为关键。探究痤疮的病证特点并其理、法、方、药,不单能有效指导临床,更能提高疗效。文献研究发现,古今医家对痤疮的病位认识大体一致,归为肺脾肾三脏,涉及心肝。但对其症状、治法及处方用药的认识各家观点庞杂而欠系统。有见及此,本研究选择当代中医名家治疗痤疮的有效医案,通过强化FP-Growth算法的关联分析,索出其组方配伍规律,并发掘出一些新的认识,为中医临床与中医科研提供参考及线索。[方法]检索发表于1988年2月～2014年11月于中国知网有关中医痤疮的医案,并依据纳入排除标准,构建中医痤疮医案数据库；在进行数据挖掘、统计分析前,事先对原始医案数据进行原始资料查证、医学术语规范、药名规范、错字纠正等预处理方法；再将符合标准入选共448诊次的中医痤疮医案,采用Medcase V3.2仓公诊籍国医脉案数据记录挖掘系统,建立医案讯息采集、储存数据库,并对数据库纳入的医案讯息进行录入规范及多次的数据清洗。在规范症状术语词汇共361条、累计规范2699频次的前提下,建立集采集、储存、梳理、分析、挖掘于一体的、非单一的模块和数据挖掘方法组成的医案数据挖掘平台；运用Medcase V3.2系统,透过加强关联规则数据挖掘运算模型、Medcase Record Mining系统、Association Rule Analysis平台、Xminer Operation Tool运算工具及强化FP-Growth算法,就着统计分析结果,从舌象、脉象、治法、用药等各方面作中医痤疮医案经验的总结。[结果]频数分析方面,当代医家的痤疮医案依次以口干、面部油腻、烦躁、大便干和瘙痒为高频临床症状；.皮损依次有丘疹、脓疱、结节、瘢痕和色素沉着；病变部位依次为面部、颊部、额部、颈项、胸部、背部；皮损颜色依次为红、淡红、暗红；舌象依次为舌质红、舌尖红、舌淡红、舌淡、舌暗红、舌暗、舌边红、舌边有齿痕；脉象方面依次为弦、滑、细、数、沉、缓、小、浮、濡、弱、涩、大、实、虚、促、洪。当代医家治疗痤疮的医案中出现的首十五种高频治法依次为清热、解毒、活血、散结、化瘀、凉血、清肺、健脾、疏肝、化痰、利湿、祛湿、滋阴、清胃、通腑。对高频治法整合规范后,发现痤疮高频治法与上海科学技术出版社陆德铭主编的《中医外科学》中所列痤疮治法相符,说明痤疮临床治法仍以清肺、散风、清热、化湿、化痰、健脾、渗湿为主；研究发现当代医家的医案治法灵活,反映痤疮病情多复杂多变。当代医家治疗痤疮的医案中首十五味高频中药依次为牡丹皮、连翘、甘草、赤芍、丹参、当归、黄芩、茯苓、生地黄、薏苡仁、蒲公英、白术、柴胡、白花蛇舌草、陈皮等。参照上海科学技术出版社雷载权主编的《中药学》将上述高频中药分类,高频药物种类为：凉血、活血、清热解毒、清热燥湿、健脾渗湿、疏肝、化痰、清热、排脓药等。药物内关联方面,除了发现连翘与丹参、金银花与连翘、牡丹皮与茯苓、白术与茯苓这些经典药对外,还发现了四药联用,如牡丹皮、赤芍、甘草、连翘及牡丹皮、黄芩、连翘、甘草；三药联用则有生地黄、牡丹皮、赤芍及金银花、连翘、丹参。综合治则治法与药物外关联而言,疏肝与柴胡、牡丹皮有较高置信度；清胃与桑白皮、枇杷叶有较高置信度；解毒与生地黄、金银花、茵陈有较高置信度。化瘀、活血与桂枝有较高置信度；温阳、潜阳与制附子有较高置信度；滋阴、养阴与黄芩有较高置信度；调和清木、解郁与柴胡有较高置信度。清肺与桑白皮、清热与黄芩、利湿与茵陈、健脾与党参、疏风与荆芥均有较高置信度。[结论]通过对选取的当代医家治疗痤疮的医案进行统计分析和关联规则分析后,得出的结论大部分属于临床常用的、现存的、符合传统理论的认识,由此反证应用数据挖掘进行医案分析乃有效确当。可透过高频药物组来总结分析痤疮常用中药的配伍应用关系。利用关联规则分析,使治则与中药、症状与中药之间的组合关系,透过定量的数理关系来表示,将“症法药”相互连系,令挖掘结果可作为临床的参考及应用。在比对挖掘结果与传统中医药理论后,发现了许多新视角、新观点,其实用性可供临床参考,说明数据挖掘及强化FP-Growth算法对于中医医案的规范化处理和再认识是既可取而又值得推广的一种方法。
[Abstract]:Based on the correlation analysis of FP - Growth algorithm , this study selects the effective medical record of Chinese traditional Chinese medicine for treating acne , and finds out some new knowledge to provide reference and clue for Chinese medicine clinical and traditional Chinese medicine research .
Prior to the data mining and statistical analysis , the original data of medical records were verified , and the methods of pre - treatment , such as original data verification , medical terminology specification , drug name specification and error correction , were carried out in advance ;
A medical record data mining platform consisting of collection , storage , carding , analysis , mining and integration , non - single module and data mining method is established on the premise of standardizing symptom terms and vocabulary of 361 and cumulative specification of 2699 frequency .
Using Medcase V3.2 system , through strengthening association rules data mining operation model , Medcase Record Mining system , Association Rule Analysis platform , Xishan Operation Tool operation tool and enhanced FP - Growth algorithm , the authors summarize the experience of Chinese traditional Chinese medicine acne medical records from aspects of tongue image , pulse image , treatment method and medication .
and the like . The lesions in turn have papules , pustules , nodules , scars and pigmentation ;
The lesion site is in turn the face , cheek , forehead , neck , chest , back ;
the color of the skin loss is in turn red , light red and dark red ;
The tongue image is in turn the tongue is red , the tongue is red , the tongue is reddish , the tongue is pale , the tongue is dark red , the tongue is dark , the tongue is red , and the tongue edge has a tooth mark ;
The first fifteen kinds of high - frequency treatment methods in the medical record of acne include clearing away heat , removing toxic substance , promoting blood circulation , dispersing pathogenic accumulation , dispersing blood stasis , cooling blood , clearing away lung , strengthening spleen , dispersing stagnated liver qi , eliminating phlegm , clearing away heat , eliminating dampness , nourishing yin , clearing stomach and dredging viscera .
In this paper , the relationship between traditional Chinese medicine and traditional Chinese medicine is analyzed by means of correlation rule analysis , and the results can be used as reference for clinical reference . The results of comparison show that the data mining and the enhanced FP - Growth algorithm are both desirable and worthy of popularization .
the combination of three medicines is prepared from Radix Rehmanniae , Cortex Moutan , Radix Paeoniae Rubra , Flos Lonicerae , Fructus Forsythiae , and Saviae Miltiorrhizae Radix .
Qingwei had higher confidence level with mulberry bark and loquat leaf .
The detoxification and Sheng Di Huang , Flos Lonicerae and Herba Artemisiae Scopariae have high confidence level , and have high confidence in removing blood stasis , promoting blood circulation and Ramulus Cinnamomi ;
There is a high degree of confidence in warming yang , latent yang and Radix Aconiti Lateralis Preparata ;
has high confidence in nourishing yin , nourishing yin and radix scutellariae ;
A Chinese medicinal composition for treating acne is prepared from Cortex Moutan , Fructus Forsythiae , Radix Glycyrrhizae , Radix Paeoniae Rubra , Radix Salviae Miltiorrhizae , Radix Angelicae Sinensis , Scutellariae Radix , Poria , Radix Rehmanniae , Radix Angelicae Sinensis , Scutellariae Radix , Poria , Radix Rehmanniae , Radix Angelicae Sinensis , Scutellariae Radix , Poria , Radix Rehmanniae , Radix Angelicae Sinensis , Herba Taraxaci , Rhizoma Atractylodis Macrocephalae , Radix Bupleuri , Rhizoma Atractylodis Macrocephalae , and Poria .

【学位授予单位】：南京中医药大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：R249;R275.9

【参考文献】