Study On Relationship Of Digital Thermal Change With Vascular Function And Its Modeling

Cardiovascular disease are threatening people's health seriously,its pathological foundation is the change of vascular function.Early detection of vascular dysfunction plays an important role in suppressing and slowing the progression of cardiovascular disease.Digital thermal monitoring could assess vascular function by detecting the changes of digital temperature during and after occlusion,and its reproducibility and availability have been demonstrated preliminarily.The current study focused on the indicators of finger temperature rebound after occlusion,but these indicators are unstable.Besides,the relationship between digital thermal curve and vascular function needs further study during and after occlusion of brachial artery.112 subjects were studied for the statistical model of digital thermal monitoring,including 75 healthy subjects and 37 patients.Firstly,Characteristic parameters of digital thermal curve were extracted and analyzed by multivariate linear Regression method and the receiver operating characteristic curve.Then,Logistic statistical models were established by combining characteristic parameters and the parameters such as age,systolic blood pressure and body temperature.Logistic regression models were building and two best models were screening by the ten-fold cross validation method.To verify the accuracy rates of these models,26 subjects as a validation group were selected for trial.To study the relationship of digital thermal monitoring,pulse wave velocity and blood flow-mediated dilatation,validation group was also tested by the latter two technologies.The parameters and results of statistical models were compared with those of pulse wave velocity and blood flow-mediated dilatation.Based on the data of flow-mediated dilatation,the brachial artery dilatation range of the subjects was obtained and substituted into the one-dimensional model of the upper arm to simulate the digital thermal change after occlusion.Digital thermal curve can be divided into three stages: the stage during the occlusion,the stage of after the occlusion to the minimum temperature,the stage after the minimum temperature.All these three stages included characteristic parameters that were significant between healthy control and cardiovascular patients.Two statistical models of the temperature characteristic parameters could predict the cardiovascular disease and the areas under the receiver operating characteristic curve were 0.932 and 0.913,respectively.When the parameters of validation groupwere substituted into statistical models,the accuracyrates of two models were 92.3% and 88.5%,respectively,and the areas under the receiver operating characteristic curve of two models in validation group were 0.970 and 0.977,respectively,which indicated that the two models had good predictive abilities of cardiovascular disease.There were some correlations between digital thermal monitoring,pulse wave velocity and blood flow-mediated dilatation parameters,and the three methods were consistent with the predictive abilities of cardiovascular disease.The theoretical model of arteries in the upper arm was consistent with the measured curve,and it can simulate the stage of after the occlusion to the minimum temperature.Based on the characteristic parameters of digital thermal monitoring curve,this study established statistical models that have good accuracy rates to predict cardiovascular disease,and studied the relationship between digital thermal monitoring,pulse wave velocity and blood flow mediated dilatation.Besides,the digital thermal monitoring curves simulated by theoretical model had good coherence with experiments data in the stage of after the occlusion to the minimum temperature.In summary,the digital thermal monitoring curve was divided further,and the statistical model and theoretical model were used to verify the importance of the stage of after the occlusion to the minimum temperature,which laid solid foundation for the application of digital thermal monitoring.