A Novel Angiography-Derived Computational Coronary Flow Reserve to Evaluate Non-Obstructive Coronary Artery Disease.

Abstract

Coronary flow reserve (CFR) is a key parameter for risk stratification in coronary artery disease but is limited by high cost, prolonged procedure time, and suboptimal reproducibility. We proposed a novel angiography-based method (Angio-CFR) to overcome these challenges and assessed its diagnostic performance. 107 consecutive patients underwent invasive coronary angiography with thermodilution-derived CFR (CFR_thermo) were prospectively enrolled. Flow velocity at hyperemia and rest was estimated from angiographic images, and Angio-CFR was calculated as their ratio. Angio-CFR correlated well with CFR_thermo (r = 0.72, p < 0.001), and showed good discrimination of CFR_thermo < 2.5 (AUC = 0.879, p < 0.001) with an optimal cut-off of 2.5. The accuracy, sensitivity, and specificity of Angio-CFR were 81.71%, 83.33%, and 81.25%, respectively. Angio-CFR provides a pressure-wire-free method for coronary function assessment, demonstrating promising diagnostic accuracy and offering a more accessible approach for CFR measurement in clinical practice.