AUTHOR=Wen Runze , Xie Qiang , Pan Bo , Ni Ming , Zhu Xingxing , Meng Xueer , Wei Zhenheng , Wu Xinai , Li Dan , Wang Xuemei TITLE=Simultaneous 13N-ammonia and gadolinium perfusion using integrated PET-MRI: diagnostic accuracy in coronary microvascular dysfunction JOURNAL=Frontiers in Medicine VOLUME=Volume 12 - 2025 YEAR=2025 URL=https://www.frontiersin.org/journals/medicine/articles/10.3389/fmed.2025.1649175 DOI=10.3389/fmed.2025.1649175 ISSN=2296-858X ABSTRACT=ObjectiveAbsolute quantification of myocardial perfusion and coronary flow reserve (CFR) with positron emission tomography (PET) has demonstrated diagnostic and prognostic value in patients with coronary microvascular dysfunction (CMD). However, no studies have compared magnetic resonance imaging (MRI) and PET perfusion imaging in patients with CMD using integrated PET-MRI imaging. The aim of this study was to assess the quantitative accuracy of cardiac perfusion measurements using MRI with simultaneous 13N-ammonia PET as reference with a fully integrated PET-MRI scanner.MethodsThirty patients with suspected CMD underwent simultaneous MRI and 13N-ammonia PET scans at rest and regadenoson-stress on an integrated PET-MRI scanner. Correlation and agreement between MRI-and PET-derived myocardial blood flow (MBF) and CFR values were evaluated using correlation and Bland–Altman analysis.ResultsMRI measurements of global rest and stress MBF showed moderate correlation to those obtained using 13N-ammonia PET (r = 0.50; p = 0.005 for rest MBF and r = 0.49; p < 0.006 for stress MBF). Bland–Altman analysis revealed a mean bias of −0.83 ± 0.47 mL/g/min for rest MBF and −1.84 ± 0.57 mL/g/min for stress MBF. The correlations between regional MBFMRI and MBFPET obtained during rest and stress were only poor to moderate (r = 0.26 and r = 0.43). The limits of agreement were wide for both global and regional MBF, with larger variability for high MBF-values. However, there was good agreement between MRI and PET with regard to global and regional CFR with moderate to strong correlation (r = 0.64, p < 0.001; r = 0.48, p < 0.001). MRI-derived CFR demonstrated an area under the curve (AUC) of 0.85 (95% CI: 0.67 to 0.95) and had an optimal cutoff value of 1.57 for detecting CMD, defined as 13N-ammonia PET-derived CFR ≤ 2.0.ConclusionCFR measurements were concordant between MRI and 13N-ammonia PET. For detecting significant CMD, CFRMRI and CFRPET demonstrated comparable and high accuracy. Nevertheless, MRI measurements of rest and stress MBF showed only poor to modest agreement to those obtained with 13N-ammonia PET. Therefore, although quantitative MRI has clinical utility, further refinements are still required.