Forecasting Effects Of Chinese Prescription And Medicine Based On Symptoms

Treatment based on syndrome differentiation is the basic principle of understanding and treating diseases in TCM.TCM syndrome refers to the differentiation of the etiology,nature,location and the relationship between pathogenic factors according to the disease information obtained from the four diagnoses.TCM refers to these diagnosis results as "syndromes",believing that "syndromes" reflect the essence of pathological changes at a certain stage in the development of the disease,and can reveal the essence of the disease more than "symptoms".Treatment is based on the dialectical results of the treatment method,the main method of traditional Chinese medicine for the treatment of diseases is prescription,that is,according to the"syndrome" to give corresponding function of drugs,combined with the compatibility theory to give the corresponding function of the prescription.In the process of TCM clinical diagnosis and treatment,TCM syndrome is a key step,premise and basis of treatment methods.However,the TCM syndrome results depend on the doctors’personal clinical experience and cognitive understanding of TCM theories.Different doctors may prescribe different or even opposite curative effects to the same patient,resulting in the lack of consistency of TCM diagnosis and treatment results and evaluation criteria of curative effects.So how to reduce the influence of doctors’subjective factors in the process of TCM diagnosis,improve the overall quality and level of TCM treatment is a hot issue in current TCM clinical research.In view of this problem,this paper puts forward the solution to reduce the TCM syndrome,recommend the prescription of traditional Chinese medicine directly from the symptoms.In order to verify the feasibility of this idea,it is necessary to study the relationship between TCM symptoms and prescription-medicine effects.Combined with the data of modern Chinese medicine prescription,Chinese herbal medicine and ancient Chinese medicine prescription,this paper used machine learning method to establish the prediction model which can predicted the relationship between TCM symptoms and prescription-medicine effects.Finally,analyzes the correlation between TCM symptoms and prescription-medicine effects according to the prediction results.The main research results include the following three aspects:(1)First,analyzes the present situation and deficiency of the research on the function of TCM based on symptom group,and proposes a method to predict the relationship between symptom and function based on machine learning algorithm.(2)Built a standard data set based on Modern prescription case and TCM pharmacopoeia.Standardized the efficacy and symptom in order to reduce the problems of non-standard and vague expression.Used Cosine similarity and Dice distance to realize the standardization of symptom features,solve the noise problems of polysemy,ambiguity,cross-meaning or coverage in data features,and realize the quantitative transformation of Chinese medicine text data based on TF-IDF algorithm.(3)Experiment established predictive models of the relationship between prescription-medicine effects and symptoms based on machine learning algorithm,used SVM algorithm,Naive Bayes algorithm,KNN algorithm,CART algorithm and other general machine learning algorithms to establish the TCM symptom-effects prediction model.Achieved the TCM,symptom-effects prediction,compared and evaluated the prediction results and model performance.The average accuracy of the prediction model constructed in this paper reaches 80%,and the research results show that there is a strong correlation between TCM symptoms and effects,which lays a necessary theoretical foundation for TCM prescription recommendation model and the development of TCM auxiliary clinical decision-making system based on TCM symptoms.