Fast Virtual Fractional Flow Reserve Based Upon Steady-State Computational Fluid Dynamics Analysis: Results From the VIRTU-Fast Study

Objectives (i) Develop and validate a method which computes fractional flow reserve (FFR) in <5 minutes on a personal computer; (ii) perform a sensitivity analysis of the physiological parameters which determine FFR. Background Percutaneous coronary intervention (PCI) guided by FFR is superior to standard assessment but remains underused. Computed, ‘virtual’ FFR (vFFR), may increase uptake of physiological guidance but is slow to compute with computational fluid dynamics (CFD) modelling. Methods Seventy-three 3-D arterial geometries were segmented from 20 patients undergoing PCI and FFR. The trans-lesional pressure drop was represented as a quadratic function of flow based upon paired steady-state analyses. Coronary microvascular impedance was determined from invasive pressure measurements. vFFR was computed using a novel pseudo-transient analysis protocol. Results were compared with fully transient CFD analysis and the clinically-measured FFR. Model sensitivity analysis was performed with the Sobol decomposition method. Results The newly developed computational protocol executed in a mean time of 189 seconds (SD 34s), whereas transient analysis took >26h. Mean error was <1%. Intraclass correlation coefficient between virtual and measured values was 0.99 (P<0.0001). Physiological lesion significance (FFR≤0.80) was discerned with 100% accuracy. Sensitivity analysis demonstrated that coronary microvascular resistance (CMVR) was the dominant influence upon vFFR. Conclusions Given an accurate value for CMV resistance, vFFR can be accurately computed from coronary angiography in <4 min. Transient physiological results can be approximated without performing time-consuming transient CFD analysis. vFFR is influenced less by geometric accuracy than by the accuracy of the representation of the distal CMV resistance.