State-of-the-Art of Myocardial Perfusion by CMR: A Practical View

• Published in: Reviews in cardiovascular medicine Vol. 23; no. 10; p. 325
• Main Authors: Pons-Lladó, Guillem, Kellman, Peter
• Format: Journal Article
• Language: English
• Published: Singapore IMR Press 01.10.2022
• Subjects: computational intelligence; ischemic heart disease; perfusion magnetic resonance imaging; Review; computational intelligence; ischemic heart disease; perfusion magnetic resonance imaging
• ISSN: 1530-6550; 2153-8174; 2153-8174; 1530-6550
• DOI: 10.31083/j.rcm2310325

Ischemic heart disease (IHD) outstands among diseases threatening public health. Essential for its management are the continuous advances in medical and interventional therapies, although a prompt and accurate diagnosis and prognostic stratification are equally important. Besides information on the anatomy of coronary arteries, well covered nowadays by invasive and non-invasive angiographic techniques, there are also other components of the disease with clinical impact, as the presence of myocardial necrosis, the extent of pump function impairment, and the presence and extent of inducible myocardial ischemia, that must be considered in every patient. Cardiovascular Magnetic Resonance (CMR) is a multiparametric diagnostic imaging technique that provides reliable information on these issues. Regarding the detection and grading of inducible ischemia in particular, the technique has been widely adopted in the form of myocardial perfusion sequences under vasodilator stress, which is the subject of this review. While the analysis of images is conventionally performed by visual inspection of dynamic first-pass studies, with the inherent dependency on the operator capability, the recent introduction of a reliable application of quantitative perfusion (QP) represents a significant advance in the field. QP is based on a dual-sequence strategy for conversion of signal intensities into contrast agent concentration units and includes a full automatization of processes such as myocardial blood flow (MBF) calculation (in mL/min/g), generation of a pixel-wise flow mapping, myocardial segmentation, based on machine learning, and allocation of MBF values to myocardial segments. The acquisition of this protocol during induced vasodilation and at rest gives values of stress/rest MBF (in mL/min/g) and myocardial perfusion reserve (MPR), both global and per segment. Dual-sequence QP has been successfully validated against different reference methods, and its prognostic value has been shown in large longitudinal studies. The fact of the whole process being automated, without operator interaction, permits to conceive new interesting scenarios of integration of CMR into systems of entirely automated diagnostic workflow in patients with IHD.