BACKGROUND: Absolute quantification of perfusion with cardiovascular magnetic resonance has not previously been applied in patients with coronary artery bypass grafting (CABG). Owing to increased contrast bolus dispersion due to the greater distance of travel through a bypass graft, this approach may result in systematic underestimation of myocardial blood flow (MBF). As resting MBF remains normal in segments supplied by noncritical coronary stenosis (<85%), measurement of perfusion in such territories may be utilized to reveal systematic error in the quantification of MBF. The objective of this study was to test whether absolute quantification of perfusion with cardiovascular magnetic resonance systematically underestimates MBF in segments subtended by bypass grafts.METHODS AND RESULTS: The study population comprised 28 patients undergoing elective CABG for treatment of multivessel coronary artery disease. Eligible patients had angiographic evidence of at least 1 myocardial segment subtended by a noncritically stenosed coronary artery (<85%). Subjects were studied at 1.5 T, with evaluation of resting MBF using model-independent deconvolution. Analyses were confined to myocardial segments subtended by native coronary arteries with <85% stenosis at baseline, and MBF was compared in grafted and ungrafted segments before and after revascularization. A total of 249 segments were subtended by coronary arteries with <85% stenosis at baseline. After revascularization, there was no significant difference in MBF in ungrafted (0.82±0.19 mL/min/g) versus grafted segments (0.82±0.15 mL/min/g, P=0.57). In the latter, MBF after revascularization did not change significantly from baseline (0.86±0.20 mL/min/g, P=0.82).CONCLUSIONS: Model-independent deconvolution analysis does not systematically underestimate blood flow in graft-subtended territories, justifying the use of this methodology to evaluate myocardial perfusion in patients with CABG.