Han Yushui, Ahmed Ahmed Ibrahim, Schwemmer Chris, Cocker Myra, Alnabelsi Talal S, Saad Jean Michel, Ramirez Giraldo Juan C, Al-Mallah Mouaz H
Debakey Heart & Vascular Center, Houston Methodist Hospital, Houston, Texas, USA.
Computed Tomography-Research & Development, Siemens Healthcare GmbH, Erlangen, Bayern, Germany.
Open Heart. 2022 Mar;9(1). doi: 10.1136/openhrt-2021-001951.
Advances in CT and machine learning have enabled on-site non-invasive assessment of fractional flow reserve (FFR).
To assess the interoperator and intraoperator variability of coronary CT angiography-derived FFR using a machine learning-based postprocessing prototype.
We included 60 symptomatic patients who underwent coronary CT angiography. FFR was calculated by two independent operators after training using a machine learning-based on-site prototype. FFR was measured 1 cm distal to the coronary plaque or in the middle of the segments if no coronary lesions were present. Intraclass correlation coefficient (ICC) and Bland-Altman analysis were used to evaluate interoperator variability effect in FFR estimates. Sensitivity analysis was done by cardiac risk factors, degree of stenosis and image quality.
A total of 535 coronary segments in 60 patients were assessed. The overall ICC was 0.986 per patient (95% CI 0.977 to 0.992) and 0.972 per segment (95% CI 0.967 to 0.977). The absolute mean difference in FFR estimates was 0.012 per patient (95% CI for limits of agreement: -0.035 to 0.039) and 0.02 per segment (95% CI for limits of agreement: -0.077 to 0.080). Tight limits of agreement were seen on Bland-Altman analysis. Distal segments had greater variability compared with proximal/mid segments (absolute mean difference 0.011 vs 0.025, p<0.001). Results were similar on sensitivity analysis.
A high degree of interoperator and intraoperator reproducibility can be achieved by on-site machine learning-based FFR assessment. Future research is required to evaluate the physiological relevance and prognostic value of FFR.
CT和机器学习的进展使得能够在现场对血流储备分数（FFR）进行无创评估。
使用基于机器学习的后处理原型评估冠状动脉CT血管造影衍生的FFR在不同操作者之间和同一操作者内部的变异性。
我们纳入了60例有症状且接受了冠状动脉CT血管造影的患者。在使用基于机器学习的现场原型进行训练后，由两名独立操作者计算FFR。如果没有冠状动脉病变，则在冠状动脉斑块远端1 cm处或节段中间测量FFR。组内相关系数（ICC）和Bland-Altman分析用于评估FFR估计值中不同操作者之间的变异性影响。通过心脏危险因素、狭窄程度和图像质量进行敏感性分析。
共评估了60例患者的535个冠状动脉节段。每位患者的总体ICC为0.986（95%CI 0.9至0.992），每个节段为0.972（95%CI 0.967至0.977）。FFR估计值的绝对平均差异为每位患者0.012（95%一致性界限CI：-0.035至0.039），每个节段为0.02（95%一致性界限CI：-0.077至0.080）。Bland-Altman分析显示一致性界限较窄。与近端/中间节段相比，远端节段的变异性更大（绝对平均差异0.011对0.025，p<0.001）。敏感性分析结果相似。
通过基于机器学习的现场FFR评估可实现高度的不同操作者之间和同一操作者内部的可重复性。未来需要开展研究以评估FFR的生理相关性和预后价值。
