Abstract 1081: Correcting the Underestimation of Absolute Myocardial Blood Flow by Magnetic Resonance Perfusion Imaging
Noninvasive assessment of absolute myocardial blood flow is important for evaluation of microvascular function and serial changes in perfusion. Absolute flow can be calculated from magnetic resonance (MR) perfusion studies, but a nonlinear relationship between signal intensity (SI) and gadolinium concentration ([Gd]) in the myocardium may cause underestimation. When the relationship between SI and [Gd] was mapped in chronically instrumented dogs, this was linear in the left ventricular (LV) bloodpool for a 0.005 mmol/kg Gd-DTPA minibolus but nonlinear in the myocardium for a 0.05 mmol/kg bolus. 14 dual-bolus MR first-pass perfusion studies were performed at rest and during vasodilation with variable degrees of left circumflex coronary artery stenosis. Absolute flow by Fermi function deconvolution of MR SI-time curves underestimated microsphere flow by nearly 50% (y = 0.52x + 0.07, r2 = 0.61). When the myocardial and LV SI-[Gd] relationships were used as calibration curves to convert SI values to [Gd], deconvolution of [Gd]-time curves yielded accurate myocardial flow values (y = 0.93x + 0.05, r2 = 0.68). As seen in the figure⇓, absolute flow measured in endo-, mid-, and epicardial layers progressively decreased from epi- to endocardium in the stenotic zone, (p = 0.0004) while flow did not vary across the myocardial wall in the normal zone (p = 0.72). Correcting the curvilinear myocardial signal response to gadolinium improves the accuracy of absolute myocardial flow quantification by MR perfusion imaging. The high spatial resolution enables measurement of transmural flow differences caused by impaired vasodilator reserve.
This research has received full or partial funding support from the American Heart Association, AHA National Center.