| BackgroundCoronary heart disease is a kind of ischemic heart disease.The incidence rate in my country is increasing year by year.It has the characteristics of high morbidity,high disability rate and high mortality rate.It has posed a serious threat to people’s life and health.CHD belongs to the category of "chest pain " and "true pain" in traditional Chinese medicine(TCM).TCM has always had unique advantages in the diagnosis and treatment of chest pain.TCM advocates treatment based on syndrome differentiation,which is to comprehensively analyze the information collected from different angles by the four methods of "looking,smelling,asking,and cutting" to determine the etiology,location,nature,and status of the disease and summarize it into a syndrome of a certain nature.This unique diagnostic method of TCM has the advantages of non-invasiveness and convenience,but its clinical application and promotion are affected by its strong subjectivity and lack of objective diagnostic criteria.Therefore,how to use modern technology and based on the concept of TCM to realize the objectification of syndrome differentiation is an urgent problem to be solved.Qi deficiency and blood stasis syndrome is the core syndrome of coronary heart disease and runs through the occurrence and development of coronary heart disease.In this study,the qi deficiency and blood stasis syndrome of coronary heart disease was taken as an example to explore the characteristics of tongue diagnosis and pulse diagnosis information of patients with qi deficiency and blood stasis syndrome of coronary heart disease and construct an objective syndrome differentiation model,so as to provide new ideas for the objective and standardization of TCM syndrome differentiation,so as to better guide clinical practice and improve clinical efficacy.Objective(1)Construction of coronary heart disease patients with common syndromes of multi-diagnosis information database.(2)Mining the objective data characteristics of pulse diagnosis and tongue diagnosis in patients with coronary heart disease with qi deficiency and blood stasis syndrome,and constructing the objective syndrome diagnosis model of coronary heart disease with qi deficiency and blood stasis syndrome,so as to provide objective basis for clinical syndrome differentiation of coronary heart disease.MethodClinical data collection:(1)The basic information of patients with coronary heart disease was collected by clinical questionnaire,including physical indicators,current medical history,past history,TCM symptoms,signs,subjective tongue and pulse conditions;(2)Three-probe intelligent pulse information acquisition device of TCM was used to collect the pulse information of the left and right hand,Guan and ulnar parts of the subjects;(3)The tongue image data of subjects were collected by TCM tongue and face diagnosis equipment.Data processing and model construction:(1)The gender,age,subjective tongue,pulse,TCM symptoms and signs of patients collected by clinical questionnaire were statistically analyzed.According to the quantitative diagnostic criteria of syndromes,the subjective tongue,pulse,clinical symptoms and signs were assigned and the syndrome types of patients with coronary heart disease were judged according to the diagnostic threshold of syndromes.(2)The pulse diagnosis data were preprocessed and divided into a single cycle,and the multi-cycle least squares regression method was used to establish the feature parameter extraction model of pulse diagnosis information.Extract 193 parameters including time domain parameters and frequency domain parameters.In this study,the principal component analysis,the least square method and the nested regression were used to mine the characteristics of pulse data,and the objective syndrome differentiation model of coronary heart disease with qi deficiency and blood stasis based on pulse information was established.(3)Segment the tongue body of tongue diagnosis data,eliminate the interference of tongue background.The convolution neural network BiSeNetv2 was used to learn the tongue image of patients with coronary heart disease with qi deficiency and blood stasis syndrome and to construct the syndrome recognition model.ResultsObjective study on pulse diagnosis information of coronary heart disease with qi deficiency and blood stasis syndrome:the accuracy of the objective syndrome differentiation model constructed based on pulse diagnosis information was:69%-98%for qi deficiency syndrome;68%-91%for blood stasis syndrome;67%91%for qi deficiency and blood stasis syndrome.Further study found that the differences of pulse diagnosis information between qi deficiency syndrome and nonqi deficiency syndrome were mainly concentrated in zuocun h1,zuocun s1/s2,zuoguan h1,youguan h1,youguan s1,youguan s,youchi energy 1,youchi energy 4;the difference in pulse diagnosis information between patients with blood stasis syndrome and patients with non-blood stasis syndrome is mainly manifested in zuocun w,zuocun h1,zuocun s2,zuoguan s1,zuoguan energy 5,zuoguan phase 7,zuochi phase 4,zuochi phase 9,youchi energy 12;the difference in pulse diagnosis information between patients with qi deficiency and blood stasis syndrome and those with non-qi deficiency and blood stasis syndrome are mainly manifested in zuocun h1,zuocun energy 4,zuoguan w,zuochi energy 5,zuochi phase 7,youchi energy 1.Objective study on tongue diagnosis information of coronary heart disease with qi deficiency and blood stasis syndrome:the recognition accuracy of coronary heart disease syndrome recognition model constructed by convolutional neural network for Qi deficiency syndrome and non-Qi deficiency syndrome was 75%.The identification accuracy of blood stasis syndrome and non-blood stasis syndrome was 79%;the recognition accuracy of qi deficiency and blood stasis syndrome and non-qi deficiency and blood stasis syndrome was 68%.Conclusions1.The accuracy of the objective syndrome differentiation model based on pulse diagnosis information is high.Several different analysis methods can be verified each other,which can provide reference for the objective identification of TCM syndrome of coronary heart disease;the pulse characteristics excavated by the model show that patients with qi deficiency syndrome have insufficient pulse strength and deficiency of spleen and kidney;patients with blood stasis syndrome have stagnation of liver qi and string astringency of pulse;patients with qi deficiency and blood stasis syndrome have the characteristics of deficiency of pulse strength,deficiency of kidney essence and kidney qi,and stringy pulse.The characteristics of pulse diagnosis information of the three syndrome types of coronary heart disease show that the patients with coronary heart disease are located in the heart,mainly due to the lack of heart qi and insufficient driving force to the blood,resulting in heart disease caused by poor blood flow.The pulse characteristics obtained can be confirmed with the theoretical point of view of traditional Chinese medicine,and can provide an objective basis for the theory of traditional Chinese medicine.2.Convolutional neural network has better feature learning ability for tongue diagnosis picture data.The objective syndrome recognition model based on convolutional neural network has a high classification accuracy,which can provide support for the realization of intelligent auxiliary diagnosis and remote diagnosis and treatment of traditional Chinese medicine,and provide a demonstration for the crossintegration of emerging disciplines and traditional Chinese medicine. |
PDF Full Text Request