Based On Data Mining, The Exploration And Analysis Of The Symptoms And Medication Rules Of Tibetan Medicine

Objective:by using appropriate methods of data mining,to dig the rules of symptoms and the law of clinical drug using from collected medical records of Longzhibu disease.Then summary the core symptoms of its hot syndrome and find out its drug combination.Methods:to collect the clinical data of Longzhibu disease by Tibetan medicine treatment oriented reduction respectively in order for the "symptom" and "medicine" collection of related,and combined with the method of descriptive statistics,to understand clinical distribution of the two.According to the symptoms by system cluster analysis which is based on different kinds of distance calculation of data,to obtain the clustering results which reveal overall symptoms and general symptoms according to the correlation between the application and the simple rules combination of strong correlation between the symptoms.According to the clinical medication,firstly uses the system clustering analysis based on many kinds of the distance calculation of data on the use of all drugs,to form the clustering results which show the similarity between medication;secondly,using simple rules combination relationship between drugs to understand the basic situation of drug's usage;in line with the characteristics of Tibetan medicine clinical medication timing,further using the sequence in the dawn?morning?noon ?night time of the disease medication combination rule.Results:Based on the Longzhibu disease clinical retrospective survey results,descriptive statistical analysis obtained the distribution of the symptoms of the disease,the results show that the disease symptoms are very complex and in large quantities,and uneven performance,mainly focus on hot symptoms,while the symptoms of cold syndrome are fewer.Then gather the symptoms which have similar attributes into the same cluster by cluster analysis,mining thermal performance is obtained such as urine yellow,urine red,quickly disappeared foam,string pulse,full pulse,red tongue,tongue dry,headache and dizziness were gathered for a class;cold symptoms and other complicated performances for another class.As a clear correlation between the symptoms and further reveal the quantitative correlation strength between the symptoms,then use the simple association rules to discovery rules of combination between the symptoms of dizziness,headache,red tongue,taut pulse,yellow urine,foam quickly disappeared.Such a group of symptoms has a strong correlation,concluded that this group of symptoms can be as the core symptoms of the heat syndrome.To reveal the similarity between general symptoms,cluster analysis of the reflected a combination of fundamental pathogenesis of Longzhibu disease symptoms-headache,dizziness,language dysfunction,numbness or weakness of limbs and limb movement disorder.Through a comprehensive analysis of the above results,the initial formation of the headache,dizziness,language dysfunction,numbness or weakness of limbs and limb movement disorder disease is the general rule of Longzhibu disease diagnosis,while the red tongue,taut pulse,yellow urine,foam quickly disappeared as its heat syndrome diagnostic behaviors.According to the first part of Longzhibu disease heat syndrome core symptoms group and the subject of the previous research achievements of "The diagnostic scale of Longzhibu type of Tibetan Sazhibu "-judge review survey results in long stagnant heat syndrome cases of Longzhibu disease.Descriptive statistics were used to analyze the distribution of drug usage in patients with heat syndrome,that is,there are many kinds of drugs.This central tendency is conducive to the development of mining data,so the system clustering analysis is used to get drugs such as Seventy Flavors Pearl Pill,Twenty Flavor Eaglewood Pill,Ruyizhenbao Pill and other high-frequency drugs as a class,which has certain contact tips.Then,through the simple association rule analysis,we find out a group of drug combinations with strong correlation,and get the combination rule of clinical treatment of the disease.According to Tibetan medicine clinical medication time(dawn,morning,noon,and evening)features,choose suitable means of mining sequence association to find regular medication combined for heat syndrome of Longzhibu disease in sequence:Seventy Flavors Pearl Pill(dawn)>Twenty-four Flavor Eaglewood Pill(morning)>Ruyizhenbao Pill(noon)>Twenty Flavor Eaglewood Pill(evening).Conclusion:Data mining is a method that can be used not only in TCM researches of symptom and medicine rules of stroke,but also in Tibetan Medicine researches of this disease.By using this scientific and objective method,the rules have been found in this study are accorded with the Tibetan Medicine basic theories and clinical experiences.All in all,combining traditional Tibetan Medicine with advanced data mining methods will promote the transformation of the Tibetan Medicine clinical treatment to knowledge and help to spread the consequences of it.