Objective To use coronary CT angiography(CCTA)imaging technology based on deep neural network to analyze coronary stenosis,plaque and CT fractional flow reserve(CT-FFR),and to explore its value in evaluating myocardial ischemia.Methods A total of 190 patients with suspected coronary heart disease who underwent cardiac X-ray angiography(XA)and CCTA examinations in the hospital from Octo-ber 2021 to March 2023 were selected,and XA was used as the gold standard to divide them into myocardi-al ischemia group and non-myocardial ischemia group,there were 95 cases in each group.The artificial in-telligence(AI)was used to compare the difference between the 2 groups,and to evaluate its diagnostic ef-ficiency and correlation.Results There were statistically significant differences in coronary artery stenosis index,plaque length(PL),plaque volume(PV),minimum lumen area(MLA),minimum degree of ste-nosis(MLD％),lumen,plaque,lipid and fibrous lipid area,positive remodeling,low-attenuation plaque,napkin ring sign and VP content between the 2 groups.Coronary stenosis,CT-FFR and vulnerable plaque can improve the diagnostic efficiency of myocardial ischemia.In terms of the qualitative degree of CCTA lumen stenosis,the diagnosis of myocardial ischemia by physicians and AI software was consistent with that by XA(Kappa=0.853,P＜0.001).The degree of lumen stenosis of XA was significantly negatively correlated with CT-FFR(rs=-0.52),and positively correlated with MLD％max,LS and PL(rs=0.46,rs=0.42,rs=0.21),and the differences were statistically significant.Conclusion CCTA based on deep learn-ing framework has good value in the diagnosis of coronary myocardial ischemia,and the degree of lumen stenosis of CCTA is consistent with that of XA in the diagnosis of coronary myocardial ischemia.There is a significant cor-relation between myocardial ischemia and CCTA lumen stenosis,plaque and CT-FFR.
维普
万方数据