a nonsignificant trendp = 0.124
Deeks’ tests for publication bias demonstrated a nonsignificant trend toward small‐study effects for Angio‐FFR (bias estimate = ‐7.37 ± 4.74, p = 0.124), and no detectable bias was observed for ICI‐FFR (bias estimate = 3.04 ± 4.78, p = 0.533). 3.4 Sensitivity and Specificity of Computational FFR According to the ANOVA model, IVUS‐FFR had the highest ASEN of 0.92 (0.91, 0.94), and Angio‐FFR and OCT‐FFR performed similarly with the ASEN of 0.84 (0.84, 0.85) and 0.82 (0.79, 0.84), respectively.