Research On Professor Yang Dongdong’s Diagnosis And Treatment Of Insomnia Based On Machine Learning

Objective: In this study,insomnia syndrome,one of the dominant diseases in TCM diagnosis and treatment,is taken as the research goal,and the outpatient medical records of insomnia syndrome accumulated by Professor Yang Dongdong for several years are taken as the sample data.Data mining and analysis are carried out by using machine learning method,so as to reveal Professor Yang Dongdong’s TCM diagnosis and treatment rules and ideas in the diagnosis and treatment of insomnia syndrome.In order to explore the feasibility of machine learning method in the research of medical record data of TCM related dominant diseases and lay the foundation for inheriting Professor Yang Dongdong’s TCM diagnosis and treatment ideas.Methods: Firstly,more than 2000 medical records of Professor Yang Dongdong in the Department of Neurology,Affiliated Hospital of Chengdu University of traditional Chinese medicine from 2016 to 2021 were collected through the outpatient medical record system,and 804 medical records were screened out.Three professional TCM doctors,including Professor Yang Dongdong,analyzed,coded and classified the medical records of 805 research samples that met the standards.Data sets were made and input into the database,and then imported into the machine learning analysis software Python.Using the programming method to quantify the sample data,according to professor Yang Dongdong to treat insomnia syndrome characteristics of traditional Chinese medicine diagnosis and treatment to study sample data are divided into basic information,Four diagnostic methods of TCM,Treatment based on syndrome differentiation of TCM and treatment plan of four plates,respectively using frequency analysis,association rule,clustering analysis and the method of random forest part points analysis,finally establish models to the discussion of the whole forest.Results: Basic information section: the proportion of female patients was significantly higher than that of male patients,female patients accounted for 76%,male patients accounted for 24%,the mean age was 47 years old,the maximum was 81 years old,the minimum was 14 years old,the mean square deviation was 11.46;Four diagnostic methods sector: after using the association rules analysis,"inspection " and "auscultation and olfaction","interrogation","palpation" four no obvious external links between the various data,four diagnostic information are relatively independent.At the same time,the information of the four diagnostic methods and the information of treatment based on syndrome differentiation were correlated.A total of 14 related items with confidence greater than 0.7 were screened,and the analysis showed that there was an obvious correlation between the two.Treatment based on syndrome differentiation section: use association rules to screen out a total of 23 related items whose total item placement degree is greater than 0.9,showing obvious correlation among various dialectical factors;Treatment plan section: the framework of main prescriptions was taken as the core discussion content,and a total of 13 main prescriptions were obtained by cluster analysis.Diagnosis and treatment idea ideas: a total of 6 different random forest algorithm models were established in two processes.The accuracy of predicting the main party through the four diagnostic methods and the treatment based on syndrome differentiation was 0.88,and the accuracy of predicting treatment based on syndrome differentiation through the four diagnostic methods was 0.88,0.93,0.87,0.87,0.86,respectively.AUC value was used to verify the validity of each model,and Micro-F1 score was used to evaluate the accuracy of each model.AUC value and Micro-F1 score of the six models were all higher than0.85.The random forest algorithm model was used to extract characteristic values to explore the weight of various factors affecting the final main prescription of Professor Yang Dongdong in the diagnosis and treatment of insomnia syndrome,among which the five zang-organs dialectics was the most important factor affecting the final main prescription.Conclusions: In this study,the machine learning method is used to mine and analyze the insomnia medical records data of Professor Yang Dongdong from 2016 to 2021.From a large number of clinical outpatient medical records data to grasp the important information and key rules of the diagnosis and treatment process,and on this basis to sort out Professor Yang Dongdong’s treatment of insomnia syndrome,including diagnosis,dialectics,medication,etc.And transform clinical experience into systematic diagnosis and treatment methods.To provide data and method support for the inheritance of Professor Yang Dongdong’s clinical experience in treating insomnia,and provide a new idea and direction for the research of clinical diagnosis and treatment technology of traditional Chinese medicine.The results show that the method adopted in this study can effectively filter unimportant diagnosis and treatment information,and help TCM clinicians to quickly and efficiently obtain valuable content and key rules from a large number of clinical medical record data.In addition,similar methods can be used to explore the diagnosis and treatment rules of other TCM dominant diseases because the processes,including designed research procedures,medical records data collection and data analysis methods,are highly standardized.