Joint Research Of PET Myocardial Perfusion And Metabolic Imaging In The Diagnosis Of Obstructive Coronary Artery Disease

Coronary artery disease?CAD?has become the main cause of chronic disease among residents in China.Its morbidity and mortality are increasing year by year,and it is showing a trend of youth,which has brought a heavy medical and social burden to China.Thus,effective and accurate early diagnosis is significant for the effective control of CAD.Positron emission tomography?PET?,as a non-invasive examination method,uses radionuclides to image physiological,biochemical,and metabolic activities in the body,and is used in the diagnosis,risk stratificationand prognosis of patients with suspected or confirmed CAD.PET myocardial perfusion imaging?MPI?can non-invasively and accurately measure clinical diagnostic information such as myocardial blood flow and myocardial flow reserve,which can meet the quantitative needs of clinical risk stratification and treatment decisions.18F-FDG PET myocardial metabolic imaging?MMI?is currently the "gold standard" for judging the survival of myocardium,which can effectively and accurately assess the viability of surviving myocardium,and can be used to guide the revascularization of CAD and evaluate the prognosis.PET MPI includes rest and stress imaging:rest MPI means imaging after injection of tracer in a calm state;stress MPI means imaging after injection of tracer under the exercise or pharmacologic stress state.However,patients with moderate to severe CAD,especially the elderly,are prone to cardiac malignant events,such as fatal arrhythmia and cardiogenic shock,during exercise or pharmacologic stress tests.Hence,doctors and patients are under tremendous psychological stress and risk.Therefore,the clinical nuclear medicine departments of our country?taking the Department of Nuclear Medicine of Guangdong Provincial People's Hospital as an example?mostly use the scanning method combining rest MPI and MMI.Doctors identify myocardial survival based on clinical experience and indicators,but they did not use it as a diagnosis of obstructive CAD.Coronary angiography?CAG?is the "gold standard" for diagnosis of obstructive CAD,but it has disadvantages such as invasiveness,contrast agent allergy,and vascular injury.In view of this,this study proposes to joint the quantitative indicators of PET rest MPI and MMI to build a machine-learning model of coronary artery classification,which provides more reliable,safe and accurate guidance for the noninvasive diagnosis of obstructive CAD,and improves the diagnostic effectiveness of existing clinical indicators.The work of this paper mainly includes the following two aspects:?1?We analysis the value of quantitative "perfusion-metabolism mismatch"?MIS?indicator for the diagnosis of obstructive CAD by using logistic regression method.In order to investigate whether the quantitative MIS can further improve the accuracy of obstructive CAD diagnosis on the basis of rest MPI,this work constructed logistic regression classification models.The results show that the quantitative MIS is an effective predictor of obstructive CAD diagnosis.A MIS of 9%provided the best trade-off between sensitivity and specificity for indentifying obstructive CAD.The results of the multivariable model including MIS achieved the highest diagnostic accuracy?AUC:0.839?,and there are significant differences compared with the control model?p=0.0034?,indicating that the quantitative MIS can provide more effective information for the diagnosis of obstructive CAD and can improve the accuracy of diagnosis.?2?We analysis the value of joint myocardial perfusion and metabolic imaging in the diagnosis of obstructive CAD based on seven machine learning methods.The quantitative indicators of rest MPI and MMI are randomly combined to build 4 univariate and 11 multivariate models.The models are trained using seven machine learning algorithms and the model performance is compared in the validation set.The results show that Model₂34 based on the SVM method revealed the best classification performance,achieved the highest sumScore,AUC,accuracy,and sensitivity.In addition,in order to observe the performance of this model in specific populations,four subgroups analysis are performed,and the results show that this model consistently achieved significantly higher AUC for four specific subgroups.The final results show that the joint imaging of rest MPI and MMI can show high accuracy in the diagnosis of obstructive CAD,and it is safe and non-invasive.It can be used as the preferred method for the nonivasive diagnosis of obstructive CAD to reduce unnecessary invasive surgery.