A quality assurance pipeline for quantitative cardiovascular magnetic resonance myocardial perfusion imaging in myofit46

Abstract Background Quantitative cardiovascular magnetic resonance (CMR) myocardial perfusion imaging is a widely employed technique in both clinical and research settings, but a detailed quality assurance (QA) framework for its analysis is lacking. Purpose To develop and implement a comprehensive QA process for quantitative CMR stress perfusion imaging in MyoFit46–the cardiovascular sub-study of the National Survey of Health and Development. Methods 505 participants were prospectively recruited to undergo quantitative stress and rest perfusion CMR in MyoFit46, using adenosine as the stress agent and the dual cannulation method by default. Stress adequacy was assessed based on heart rate (HR), blood pressure (BP), oxygen saturation (SpO2), symptomatology and splenic switch off. QA also involved global hyperaemia grading of stress in-line maps and assessment of RR interval stability, arterial input function (AIF) curves, artefacts and respiratory excursion (full protocol in Figure 1). QA parameters were compared between sexes and between cannulation setups (dual vs single). Results 484 participants (76.28±0.81 years; male=271) completed CMR, of which 444 completed both rest and stress perfusion imaging, and 62.5% were fitted with two cannulas. All response variables differed between stress and rest, including increased HR and SpO2 (respectively 81 vs 64bpm; 98 vs 97%; both p0.001) and reduced central and peripheral systolic BP (130 vs 136mmHg; 139 vs 143mmHg; p0.001). Adequate global hyperaemia (regardless of regional ischaemia) was widely observed (94.6%), and the majority (52.5%) maintained a stable HR during stress. Stress AIF curves were single-peaked and well-defined, with the majority exhibiting sharp rising (98.7%) and falling limb (95.2%) inflexion points for later quantification. Artefacts were infrequent, with only 18.6% of stress and 12.9% of rest maps having at least one segment excluded. Females had higher stress and resting HRs compared to males (p0.001 and p=0.003, respectively) and lower central and peripheral diastolic BP at rest compared to males (both p0.001). A steady HR was more common in those with dual vs single cannulas (p=0.005, Figure 2). Conclusion QA of CMR quantitative perfusion in MyoFit46 confirmed adequate stress response in the majority of subjects and a high quality of acquired imaging data. A dual cannula approach appears to be associated with a steadier HR during image acquisition. Such comprehensive QA procedures should become a routine part of large-scale studies and clinical practice.Summary of quality assurance protocol HR steadiness in dual/single cannulation