A Study On Diagnostic Model Of Hypertension Identified As Overabundant Liver-fire Syndrome Based On Objectification Of Tongue Diagnosis

ObjectiveIn this study,the residents who established health records in the community were used as research objects.The tongue-and-mouth information collection objective identification system was used to collect the tongue diagnosis information data,and the relationship between the objective parameters of tongue diagnosis and the overabundant liver-fire syndrome of hypertension was analyzed.Then combined with other relevant information to construct a diagnostic model of overabundant liver-fire syndrome in hypertension patients,further explore the characteristic variables and diagnostic accuracy of overabundant liver-fire syndrome in the whole sample,in order to provide methods and basis for more effective and accurate syndrome differentiation and treatment of clinical hypertension.MethodThis study used 486 cases.They were divided into overabundant liver-fire syndrome of hypertension group,non-overabundant liver-fire syndrome of hypertension group,overabundant liver-fire syndrome of non-hypertension group,and normal group.First,according to the research objectives,a case report form is prepared,collect basic information by means of questionnaires.Secondly,using the tongue and vein information collection body identification system to collect the objective parameter information of the tongue diagnosis.The collected data were entered into the database by double entry method,and the statistical analysis of the case data was carried out by reasonable statistical methods to study the specificity of the tongue diagnosis objective parameter in overabundant liver-fire syndrome of hypertension and the correlation between hypertension and related risk factors.The Random Forest data mining algorithm was used to construct the diagnostic model of overabundant liver-fire syndrome in hypertension patients,and the model was validated by the test set data.The accuracy of the model prediction was given,and the important variables that contributed to the model construction were selected.Further explore the construction of a full-sample liver-fire syndrome model based on the objective parameter of tongue diagnosis.ResultsIn this study,the diagnostic model of overabundant liver-fire syndrome in hypertensive patients and in the whole sample population were constructed.1.Construct a diagnostic model of overabundant liver-fire syndrome in hypertensive patientsA total of 259 patients were enrolled in this model,which include overabundant liver-fire syndrome of hypertension group 131 cases and non-overabundant liver-fire syndrome of hypertension group 128 cases.1.1 statistical descriptionObjectification parameters of tongue diagnosis,This study found that the objective parameters of RGB-G,RGB-B,HSV-H and LAB-L on the right tongue side in overabundant liver-fire syndrome of hypertension group were significantly lower than non-overabundant liver-fire syndrome of hypertension group.There was statistical difference(P<0.05);General information,This study found that there was no statistical difference in age,gender,marital status,occupation,and education level between overabundant liver-fire syndrome group and non-overabundant liver-fire syndrome group in hypertensive patients.However,in terms of occupation,we can see that overabundant liver-fire syndrome of hypertension group of mental labor accounted for more than non-overabundant liver-fire syndrome of hypertension group;General physical examination,This study found that there was a statistically significant difference in the abdominal circumference index between overabundant liver-fire syndrome of hypertension group and non-overabundant liver-fire syndrome of hypertension group(P<0.05),in height,weight,waist circumference,chest circumference,hip circumference,there was no statistical difference;Symptoms of TCM,This study found that the overabundant liver-fire syndrome of hypertension group was significantly more severe than the non-overabundant liver-fire syndrome of hypertension group in the dizziness syndrome and irritable irritability,which was statistically different;Relevant risk factors,overabundant liver-fire syndrome of hypertension group and non-overabundant liver-fire syndrome of hypertension group have significant statistical differences in whether the treatment factors are taking Chinese medicine,whether blood pressure is fluctuating,whether lifestyle is reducing salt,and whether to reduce oil consumption index(P<0.05).1.2 model constructionIn this study,Random Forest algorithm was used to construct a diagnostic model for overabundant liver-fire syndrome of hypertensive disease,which was based on overabundant liver-fire syndrome.A total of 1180 variables were used,and a total of 259 cases were randomly divided into 4:1 ratios.The training set and test set were 194 cases and 65 cases respectively.The training set is used to obtain the model,and then the test set is used to verify the model,and the important variables that contribute more to the model are selected.The OOB error of this research model is 29.35%,and the generalization ability is strong.For the training data,the total predictive rate(Accuracy)of the data outside the bag was 71%,71%for the overabundant liver-fire syndrome,and 71%for the recall;For the full sample,the total predictive rate was 72%,the accuracy of overabundant liver-fire syndrome was 71%,and the recall rate was 73%.In this study,MeanDecreaseAccuracy and MeanDecreaseGini were used to observe the importance of independent variables for the modle,and the top 15 variables of importance ranking were screened.The results showed that the objectification parameters of tongue diagnosis accounted for the largest proportion,and others were impatient.Irritability,hip circumference,reduction of cooking oil,dizziness,systolic blood pressure,diastolic blood pressure,abdominal circumference and other indicators.The test data prediction model has an accuracy of 0.73 and AUC value of 0.849,which has a high prediction accuracy.2.Construct a diagnostic model of overabundant liver-fire syndrome in full-sample populationA total of 486 cases were enrolled in this model.The combination of overabundant liver-fire syndrome of hypertension group and overabundant liver-fire syndrome of non-hypertension group were combined into overabundant liver-fire syndrome group,a total of 267 cases,and the combination of non-overabundant liver-fire syndrome of hypertension group and normal group were combined into non-overabundant liver-fire syndrome group,a total of 219 cases.2.1 statistical descriptionObjectification parameters of tongue diagnosis,This study found that the HSV-S index of overabundant liver-fire syndrome group was significantly higher than that of non-overabundant liver-fire syndrome group in the whole sample.(P<0.05);General information,this study found that there was no statistical difference in age,gender,marital status,occupation,and education level between the overabundant liver-fire syndrome group and the non-overabundant liver-fire syndrome group;General physical examination,This study found that there was a statistically significant difference in the chest circumference,left arm heart rate,right arm systolic pressure,and right arm heart rate between the overabundant liver-fire syndrome group and the non-overabundant liver-fire syndrome group(P<0.05);Symptoms of TCM,this study found that the overabundant liver-fire syndrome group was significantly more severe than the non-overabundant liver-fire syndrome group in terms of irritable irritability,dizziness,headache,and dry mouth syndrome,with statistical differences;Related risk factors,the overabundant liver-fire syndrome group was significantly shorter in the course of disease than,the non-overabundant liver-fire syndrome group and there were statistical differences.There was a statistically significant difference in Other factors whether to take drug treatment,exercise therapy,whether to take Chinese medicine,whether to reduce salt,whether to reduce oil,quality of sleep between two groups.(P<0.05).2.2 model constructionIn this study,Random Forest algorithm was used to construct a diagnosis model of overabundant liver-fire syndrome in the whole sample population.The results were based on whether the overabundant liver-fire syndrome was the outcome.A total of 486 cases were randomly divided into training set and test set according to the ratio of 4:1.389 cases,97 cases.The training set is used to obtain the model,and then the test set is used to verify the model,and the important variables that contribute more to the model are selected.The OOB error of this research model is 24.70%,and the generalization ability is strong.For the training data,the total predictive rate(Accuracy)of the data outside the bag is 77%,60%for the overabundant liver-fire syndrome,and the recall rate is 77%;For the whole sample,the total predictive rate(Accuracy)is 73%,the accuracy rate for overabundant liver-fire syndrome was 60%,and the recall rate was 72%.In this study,MeanDecreaseAccuracy and MeanDecreaseGini were used to observe the importance of independent variables for the model,and the top 15 variables of importance ranking were screened.The results also showed that the objective parameters of tongue diagnosis accounted for the largest proportion.The test data prediction model has an accuracy of 0.74 and an AUC value of 0.802,which has a high prediction accuracy.ConclusionIn this study,the random forest algorithm was used to construct the diagnosis model of hypertensive disease of overabundant liver-fire syndrome.The accuracy of model prediction was high.It was found that the objective parameters of tongue diagnosis had the highest contribution in the construction of the model,indicating the objectification parameters of tongue diagnosis that is of great significance for the diagnosis of hypertensive disease,such as overabundant liver-fire syndrome,which provides some guidance for the study of hypertension syndrome.2.This study is based on the objective parameters of tongue diagnosis,combined with general information,TCM symptoms,physical examination,lifestyle and other relevant information for comprehensive analysis,construction of syndrome model,feasibility,and the risk of hypertension related factors,The analysis provides a basis for the prevention and treatment of hypertension;3.This study further explored the full-sample overabundant liver-fire syndrome diagnosis model,and found that the objective diagnosis parameters of tongue diagnosis have the highest contribution rate to the diagnosis model of overabundant liver-fire syndrome of the two groups of people,and in the two models,the symptoms of Chinese medicine are "impatient".The irritability and the objective parameters of the tongue diagnosis are repeated on the right side of the tongue HSV-H,the right side of the tongue RGB-G,the right side of the tongue HSV-S,the left side of the tongue LAB-B,the tip of the LAB-A index,indicating These indicators have important significance for the diagnosis of overabundant liver-fire syndrome,and provide some ideas for the research of other TCM syndrome diagnosis.