146 Chapter 7 integrating into daily clinical workflow. Nevertheless, simplified protocols with a nondiluted pre-bolus of small volume have been successfully implemented.8-10 Until recently, QP CMR required cumbersome and time-consuming operator-dependent analysis, which made it prone to substantial interobserver variability and prevented its widespread use in clinical routine.11-25 To address these limitations, the fully automated pixel-wise QP CMR image processing framework has been established, which is reproducible, time-efficient and does not require manual delineation of regions of interests (ROIs) or myocardial image segmentation.12, 17, 19, 20, 26-28 In this study, we clinically implemented a ready-to-use QP CMR workflow with a simplified dual-bolus GBCA administration scheme and fully automated QP post-processing. The study aimed to evaluate (1) the image quality, and (2) results and diagnostic accuracy of the proposed QP CMR workflow, to identify obstructive coronary artery disease (CAD) in patients with suspected CAD who underwent adenosine stress perfusion CMR. METHODS The study retrospectively analyzed 25 patients with suspected myocardial ischemia who underwent both adenosine stress perfusion CMR imaging with the simplified dualbolus scanning protocol and invasive coronary angiography (ICA) or coronary computed tomography angiography (CCTA) as a clinical routine at Amsterdam UMC, location VUmc. ICA was used to confirm or exclude presence of obstructive CAD, while CCTA was used only to exclude obstructive CAD. Obstructive CAD was defined as ≥ 70% stenosis in at least one major coronary vessel or > 50% stenosis in the left main coronary artery by visual assessment of the coronary angiogram, or fractional flow reserve (FFR) ≤ 0.8. Nonobstructive CAD was defined as: (1) < 70% stenosis in ICA, or (2) FFR > 0.8, or (3) < 30% stenosis and a calcium score 0 in all major coronary arteries in CCTA. Performance of FFR during ICA was on the discretion of the operator, and according to the currently available guidelines and recommendations. The study was approved by the Medical Ethics Review Committee of the Amsterdam UMC, location VUmc. All methods were performed in accordance with the relevant guidelines and regulations. Informed consents from all participants were obtained. CMR image acquisition Images were acquired using a 3 T whole-body MR scanner (Magnetom Vida, Siemens Healthcare, Erlangen Germany). Patients were instructed to avoid products containing caffeine or xanthine for at least 24 h before the examination.
RkJQdWJsaXNoZXIy MTk4NDMw