Diagnostic accuracy of angiographic quantitative flow ratio across cardiac phases in patients with myocardial bridge: a retrospective cohort study

Background and purpose: The functional assessment of myocardial bridge (MB) is challenging and still unsettled. The objective of this study is to evaluate the feasibility and diagnostic performance of quantitative flow ratio (QFR) computed during different cardiac phases in patients with MB, using wire-based fractional flow reserve (FFR) as the reference standard. Methods: Patients with MB who underwent both coronary angiographies and FFR measurement were retrospectively enrolled. Murray law-based QFR (μQFR) of the MB artery was automatically calculated for the bridged artery using a single angiographic projection, empowered by artificial intelligence (AI) algorithms, in three different approaches: from systolic phase images (sys-μQFR), diastolic phase images (dias-μQFR), and from lumen deformation throughout the cardiac cycle at the milking segment (D-μQFR). Results: A total of 118 vessels from 118 patients were analyzed. When compared with FFR using a cutoff value of 0.80 to define ischemia, diagnostic accuracy was 79% (95% confidence interval [95% CI]: 71%–86%) for sys-μQFR, 86% (95% CI: 80%–93%) for dias-μQFR, and 92% (95% CI: 86%–97%) for D-μQFR. Sensitivities were 96% (95% CI: 81%–100%), 41% (95% CI: 22%–61%), and 67% (95% CI: 46%–84%), while specificities were 74% (95% CI: 63%–82%), 100% (95% CI: 96%–100%), and 99% (95% CI: 94%–100%) for sys-μQFR, dias-μQFR, and D-μQFR, respectively. The area under the receiver operating characteristic (ROC) curve (AUC) was 0.86 (95% CI: 0.79–0.92) for sys-μQFR, 0.92 (95% CI: 0.86–0.96) for dias-μQFR, and 0.94 (95% CI: 0.88–0.98) for D-μQFR. Conclusions: AI-aided μQFR enables rapid, wire-free functional assessment of MB. By quantifying lumen deformation dynamics across the cardiac cycle, D-μQFR achieves optimal diagnostic accuracy. This technology may serve as a clinically viable wire-free alternative to invasive FFR for MB evaluation.