Exploring Professor Zhang Minghe’s Experience In Treating Systemic Lupus Erythematosus Based On Data Mining

Objective: Based on Data Mining Technology,study on professor Zhang Minghe’s theory and treatment of systemic lupus erythematosus,as well as the laws of the Medicines,summary of academic thinking,to inherit.Method: Included between 2014 and 2016,In Shandong Province Chinese medicine hospital outpatient service,100 patients were diagnosed with systemic lupus erythematosus(SLE),Sort out the relevant treatment information,Such as current medical history,out-patient diagnosis,prescription medication,"Chinese medicine transmission support system V2.5" as a tool,build database,Through the use of the system integration frequency analysis,Apriori rule algorithm,unsupervised entropy hierarchical clustering method,complex system entropy clustering method and other data mining methods,Analysis of the patient’s symptoms,signs and traditional Chinese medicine prescription,Combined with Professor Zhang’s clinical teaching content to conduct a comprehensive discussion.Results: 1.Analysis results show that,The main clinical symptoms of 121 patients with frequency ≥20 were erythema rash,fatigue,joint pain,hair loss,Raynaud’s phenomenon,oral ulcer;Frequency ≥ 9 other symptoms of chest tightness,sweating,dry mouth,headache,light allergy and so on.2.Smell frequency results show,A total of 363 prescriptions were collected,Contain 173 traditional Chinese medicine,more than 100 of the frequency of traditional Chinese medicine taste of the 13.3.Drug sex,taste,return statistics results show,Drugs of four gas to cold(1494 times),temperature(1353 times)the frequency of drugs higher;Drug flavors to sweet medicine(2088),bitter herbs(1784 times)appear frequency is higher;The medicine to the liver by(2130),kidney(1715),spleen(1171)have a higher frequency.4.The method of association rule analysis is used,Set the Number of Supports to 91(ie,Support = 0.25),A total of 24 combined entries were drawn,12 traditional Chinese medicine.5.Analysis of drug association degree based on improved mutual information method,According to the degree of correlation and penalty,Analysis of complex system entropy clustering method,It was concluded that there were 22 3-6 flavor combinations.On this basis,Using unsupervised hierarchical clustering method,dig out 11 new prescriptions.Conclusion: Statistical results show that heat poisoning is the most important pathogenic factor of SLE,Professor Zhang Minghe that the core pathogenesis of this disease is congenital deficiency,Feel the heat of evil,Or within the Yun-heat heat poisoning.;Therefore,the basic principles of clearing heat and detoxicating,invigorating qi and tonifying kidney were established,At the same time pay attention to "disease and syndrome combination",With the card addition and subtraction,Identify the location,magical reference medicine.The use of data mining technology,Multi-angle,multi-faceted,transparent and accurate on the name of the old medical case in-depth analysis,So as to arrive at more objective and accurate,scientific,clinically meaningful conclusions.