基于人工智能优化与计算流体力学的无创血流储备分数对冠状动脉狭窄缺血的诊断价值

Diagnostic value of computed tomography angiography-derived fractional flow reserve based on artificial intelligence and computational fluid dynamics versus invasive fractional flow reserve for myocardial ischemia

目的:对比基于冠状动脉CTA获得的无创血流储备分数(FFRCT)和有创血流储备分数(FFR)心肌缺血的诊断效能。方法:本研究共纳入61例稳定性心绞痛患者,采用深圳睿心智能医疗开发的基于人工智能计算的专利系统(Ruixin-FFR)计算狭窄处的FFRCT值,并与冠状动脉造影术中测量的FFR值进行比较,以FFR≤0.8为具有血流动力学意义。通过准确性、敏感性、特异性、阳性预测值(PPV)和阴性预测值(NPV)并计算受试者工作特征(ROC)曲线下的面积(AUC)来评价FFRCT的诊断效能。结果:对每个患者的FFRCT与有创FFR相比的准确率为91.8%,敏感性为97.6%,特异性为73.7%,PPV为89.1%,NPV为93.3%。计算获得AUC值为0.890(P<0.0001)。FFRCT和有创FFR间相关系数r:0.652(P<0.0001),一致性检验较好(Kappa=0.757,组内相关系数r=0.770,P值均<0.001),Bland-Altman图显示两者具有良好的一致性。结论:FFRCT对比有创FFR具有良好的一致性及诊断效能,作为评估心肌缺血的一种高效、无创、准确的功能性参数。

Objective:We aimed at investigating the diagnostic value of non-invasive coronary blood fractional flow reserve based on coronary CTA imaging(FFRCT)compared with invasive coronary blood fractional flow reserve(FFR).Methods:61 patients were included in this study,The FFRCT value at each stenosis was calculated by Ruixin FFR(version 2.5.0 RUIXIN Technology Co.,Ltd.)and evaluated with accuracy,sensitivity,specificity,positive predictive value(PPV),negative predictive value(NPV)and the area under the receiver operating characteristic(ROC)curve(AUC),with FFR≤0.8 as the reference for the diagnosis of hemodynamically significant coronary stenosis.Results:The accuracy,sensitivity,specificity,PPV and NPV of Ruixin FFR versus invasive FFR were 91.8%,97.6%,73.7%,89.1%,and 93.3%.There was a significant correlation between FFR-CT and invasive FFR(r=0.652;kappa=0.757,ICC=0.770,all P<0.001),with AUC values of 0.890.Bland-Altman plot suggested the good consistency between FFRCT and FFR values.Conclusions:FFRCT based on coronary CTA imaging can be used as a non-invasive and accurate tool to evaluate the myocardial ischemia due to coronary artery stenosis.