The Theoretical And Empirical Research On Disease Differentiation Mode Of “1+5” In Traditional Chinese Medicine

ObjectiveThis study established a new disease differentiation mode in traditional Chinese medicine(TCM-DDM),which aims to build an artificial intelligence TCM diagnostic model,combining with big data and AI.This new mode is both practical and medically feasible to improve the diagnosis accuracy,to optimize the intelligent diagnostic system,and to edify somewhat to the reader.Methods1.The concept and characteristics of “disease” in TCM are recognized by systematic review literature retrieval method.The same method is used to explore the multi-axis disease classification pattern.2.The comparative analysis approach is used to investigated the practice situation of TCM diagnosis.The study shows that,a new TCM-DDM could be a solution.3.The theoretical connotation,logical structure and basic features of the new TCM-DDM are explained by theoretical deduction.4.Under the guidance of the new mode,support vector machine,multi-layer perceptron and other relevant algorithms were used to construct artificial intelligence TCM diagnostic model driven by data and knowledge for 7 common TCM diseases,and the accuracy of diagnosis was verified to provide data support for confirming the rationality and applicability of the mode.Results1.The concept of TCM “disease” is the imbalance of human body,the harmony between nature and humanity,the unity of space and time.The essences of TCM “disease” are out of balance,phenomenological,whole,complex,and multilayer.In this study,the diseases in internal medicine of TCM were divided into three types:diseases of viscera,viscera and organs.Combined with the different classification axes of etiology,main symptom,pathogenesis,disease trend and course,the classification system of lung diseases in TCM is reconstructed.2.At present,there are four modes of disease differentiation of TCM in different practical situations.The states quo of TCM disease diagnostic shows the short of TCM thoughs,and the lack of disease differentiation,diagnosis path,diagnostic criteria and differential diagnosis,which restrict the further development of TCM disease differentiation to a certain extent.3.This study bases on the disease identification parameters,the TCM disease name as the conclusion,the diagnostic logic as the intermediary,and the knowledge rules as the guarantee,constructing a new TCM-DDM of “1+5”.“1” represents the deep integration into a core concept,namely the overall concept;“5” represents the important steps of the 5 rings of collaborative linkage,which are the synthesis of multiple parameters,the extraction of disease characteristics,the skillful use of diagnostic logic,the dynamic identification and exclusion,and the rechecking of disease name diagnosis.The whole concept runs through the whole process,and the 5-ring steps are unfolded step by step to form a complete and closed-loop reasoning process of TCM diagnosis.Wholeness,dynamics and order are the basic characteristics of the model.4.For 7 common diseases in internal medicine of TCM,the name diagnosis of common diseases in internal medicine of TCM including 763 disease differentiation parameters was established.Under the guidance of the new model,an intelligent diagnosis model of TCM disease names driven by data and knowledge was constructed by using support vector machine,multi-layer perceptron and knowledge rules of TCM disease differentiation.The diagnostic accuracy of this model is up to 86%.Conclusion1.The proposed “1+5” TCM-DDM,fitting the thinking process of TCM diagnosis,can effectively meet the needs of modern clinical diagnosis and treatment,which has practical guiding significance for improving the level of doctors’ clinical disease differentiation.2.The intelligent diagnosis model of TCM driven by the data and knowledge in the guidance of the proposed “1+5” TCM-DDM,has good university and extensibility and could increase the accuracy and stability of TCM diagnosis.