摘要
目的探讨基于深度学习无创血流储备分数(CTFFR)和冠状动脉计算机断层血管成像(CCTA)的斑块成分定量分析对冠心病的诊断价值。方法收集急性冠状动脉综合征(ACS)与慢性冠状动脉综合征(CCS)各22例,分析冠状动脉斑块成分和病变段FFR_(CT)值。结果ACS组纤维斑块体积(FPV)和百分比(FPV%)、脂样斑块体积(LPV)和百分比(LPV%)、总斑块体积(TPV)和百分比(TPV%)以及血流储备差值(ΔFFR_(CT))高于CCS组,血管/斑块体积比值(L/P)和FFR_(CT)低于CCS组(P<0.001);ACS组FFR_(CT)、ΔFFR_(CT)、L/P、TPV%、TPV联合诊断AUC优于单一指标(P<0.001)。结论FFR_(CT)联合CCTA的斑块特征对ACS与CCS有一定的鉴别,为临床提供冠心病危险分层。
Objective To investigate the diagnostic value of non-invasive fractional flow reserve(CTFFR)based on deep learning combined with the quantitative analysis of plaque components in coronary computed tomography angiography(CCTA)for coronary heart disease.Method A retrospective collection of 22 cases each of acute coronary syndrome(ACS)and chronic coronary syndrome(CCS)was conducted.The coronary plaque components and lesion segment FFR_(CT) values were analyzed.Results The volume and percentage of fibrous plaque(FPV,FPV%),the volume and percentage of lipid-rich plaque(LPV,LPV%),the total plaque volume and percentage(TPV,TPV%),as well as the flow reserve difference(ΔFFR_(CT))in the ACS group were significantly higher than those in the CCS group,with the vessel/plaque volume ratio(L/P)and FFR_(CT) values being lower in the ACS group(P<0.001).The combined diagnosis of FFR_(CT),ΔFFR_(CT),L/P,TPV%,and TPV in the ACS group for AUC was superior to a single indicator(P<0.001).Conclusion The combination of FFR_(CT) and CCTA plaque characteristics offers certain discriminative power between ACS and CCS,providing clinical risk stratification for coronary heart disease.
出处
《浙江临床医学》
2024年第4期506-508,共3页
Zhejiang Clinical Medical Journal
关键词
人工智能
血流储备分数
冠心病
Artificial intelligence
Fractional flow reserve
Coronary heart disease