Background Coronary MR angiography (CMRA) at 3 T offers higher signal to noise ratio and contrast to noise ratio compared with 1.5 T. CMRA at 5 T may provide better diagnostic performance. Purpose To assess the feasibility and clinical application of 5-T noncontrast whole-heart CMRA and compare 5-T acquisition with 3-T acquisition. Materials and Methods From September 2023 to April 2024, patients scheduled for coronary CT angiography (CCTA) and volunteers were prospectively recruited. CCTA served as the reference standard in patients. CMRA was performed using a 3-T spectral attenuated inversion-recovery (3T_SPAIR) sequence and 3-T Dixon (3T_Dixon) sequence with routine spatial resolution (3T_{SPAIR-routine} and 3T_{Dixon-routine}, respectively), and 5-T Dixon (5T_Dixon) with routine and high spatial resolution (5T_{Dixon-routine} and 5T_Dixon-high, respectively). The study evaluated image quality, coronary artery calcium (CAC), the severity of coronary artery disease (CAD) graded according to Coronary Artery Disease Reporting and Data System, and the presence of ≥50% coronary stenosis. The nonparametric paired Wilcoxon signed rank test, McNemar test, generalized estimating equation model, and kappa test were used. Results Eight volunteers and 79 patients were included (mean age, 52 years ± 11 [SD]; 48 male). Image quality was higher for 5T_{Dixon-routine} compared with 3T_{SPAIR-routine} and 3T_{Dixon-routine} (P < .001 for both) and similar for 5T_Dixon-high (P = .60). The per-segment sensitivity for CAC was higher at 5T_Dixon-high than 5T_{Dixon-routine} (78.3% vs 53.3%; P < .001), with no difference in specificity (98.6% vs 98.6%; P > .99). In grading the severity of CAD, 5T_{Dixon-routine} showed better consistency with CCTA than 3T_{SPAIR-routine} (κ = 0.46 vs 0.13) and 3T_{Dixon-routine} (κ = 0.55 vs 0.42). For detecting ≥50% stenosis, the per-patient sensitivity, specificity, and accuracy were as follows: 5T_{Dixon-routine} versus 3T_{SPAIR-routine}, 88.9% versus 55.6%, 86.5% versus 62.2%, and 87.0% versus 60.9% (P = .55, .01, and .18, respectively); 5T_{Dixon-routine} versus 3T_{Dixon-routine}, 90.0% versus 80.0%, 86.8% versus 71.1%, and 87.5% versus 72.9%, respectively (P > .05 for all). Conclusion Noncontrast CMRA at 5 T shows potential to evaluate CAC and coronary stenosis simultaneously and demonstrates superior diagnostic performance compared with at 3 T. © RSNA, 2024 Supplemental material is available for this article. See also the editorial by Rahsepar and Kim in this issue.

Dong Wang

Publications