Table 3.
Diagnostic performances of μQFR1, μQFR2, and 3D-μQFR of ≤0.80 in predicting FFR of ≤0.80.
| μQFR1 ≤ 0.80 | μQFR2 ≤ 0.80 | 3D-μQFR ≤ 0.80 | |
|---|---|---|---|
| Accuracy, % (95% CI) | 92.1 (89.0-95.3) | 92.5 (89.4-95.6) | 93.2 (90.3-96.2) |
| Sensitivity, % (95% CI) | 88.1 (80.2-93.7) | 88.1 (80.2-93.7) | 90.1 (82.5-93.7) |
| Specificity, % (95% CI) | 94.4 (90.0-97.3) | 95.0 (90.7-97.7) | 95.0 (90.7-97.7) |
| PPV, % (95% CI) | 89.9 (82.2-95.0) | 90.8 (83.3-95.7) | 91.0 (83.6-95.8) |
| NPV, % (95% CI) | 93.4 (88.7-96.5) | 93.4 (88.8-96.5) | 94.4 (90.0-97.3) |
| Positive LR (95% CI) | 15.8 (8.6-28.9) | 17.5 (9.2-33.3) | 17.9 (9.4-34.0) |
| Negative LR (95% CI) | 0.13 (0.07-0.2) | 0.13 (0.07-0.2) | 0.10 (0.06-0.2) |
| AUC (95% CI) | 0.96 (0.93-0.98) | 0.95 (0.92-0.98) | 0.95 (0.92-0.97) |
| Optimal cutoff by Youden index | ≤0.81 | ≤0.80 | ≤0.80 |
AUC, area under the ROC curve; FFR, fractional flow reserve; LR, likelihood ratio; NPV, negative predictive value; PPV, positive predictive value; ROC, receiver operating characteristic; μQFR, Murray law–based quantitative flow ratio; 3D, 3-dimensional.