Quantitative pixel-wise measurement of myocardial blood flow: a comparison of methods to correct for surface coil-related field inhomogeneity

Abstract

Background Pixel-wise quantitative analysis of cardiovascular magnetic resonance (CMR) perfusion images allows myocardial blood flow (MBF) to be measured at the level of approximately 30 μL of myocardium. However B1-field inhomogeneity, induced by phased array surface receiver coils, potentially confounds MBF measurements. The aim of this study was to compare the effect of: 1. no surface coil intensity correction (No-SCIC); 2. SCIC performed using saturation recovery steady-state free precession images (SSFP-SCIC); and 3. SCIC performed using proton density-weighted images (PD-SCIC); on pixel-wise measurements of MBF in patients with confirmed significant coronary artery disease (CAD) and healthy volunteers. Methods 26 subjects, comprising 18 patients (11 male; age 58 ± 12 years) with significant CAD (> 70% stenosis on invasive coronary angiography) in 1 (56%) or 2 (44%) major epicardial arteries (78% LAD, 50% RCA, 17% Cx) and 8 healthy volunteers (7 male; age 24 ± 10 years), were included. Perfusion imaging was performed using a motion-corrected saturation recovery steady-state free precession, dual sequence method at 1.5T (Espree, Siemens Medical Solutions) during Regadenoson vasodilator stress and at rest, using 0.05 mmol/kg Gadolinium-DTPA. A proton density-weighted image was acquired at the beginning of the perfusion sequence. B1-field inhomogeneity was approximated using a third-order surface fit to myocardial and body signal intensities on the PD image or the first SSFP image. The estimated intensity bias field was