Fully automated pixel-wise quantitative CMR-myocardial perfusion with CMR-coronary angiography to detect hemodynamically significant coronary artery disease

ObjectivesWe applied a fully automated pixel-wise post-processing framework to evaluate fully quantitative cardiovascular magnetic resonance myocardial perfusion imaging (CMR-MPI). In addition, we aimed to evaluate the additive value of coronary magnetic resonance angiography (CMRA) to the diagnostic performance of fully automated pixel-wise quantitative CMR-MPI for detecting hemodynamically significant coronary artery disease (CAD).MethodsA total of 109 patients with suspected CAD were prospectively enrolled and underwent stress and rest CMR-MPI, CMRA, invasive coronary angiography (ICA), and fractional flow reserve (FFR). CMRA was acquired between stress and rest CMR-MPI acquisition, without any additional contrast agent. Finally, CMR-MPI quantification was analyzed by a fully automated pixel-wise post-processing framework.ResultsOf the 109 patients, 42 patients had hemodynamically significant CAD (FFR <= 0.80 or luminal stenosis >= 90% on ICA) and 67 patients had hemodynamically non-significant CAD (FFR > 0.80 or luminal stenosis < 30% on ICA) were enrolled. On the per-territory analysis, patients with hemodynamically significant CAD had higher myocardial blood flow (MBF) at rest, lower MBF under stress, and lower myocardial perfusion reserve (MPR) than patients with hemodynamically non-significant CAD (p < 0.001). The area under the receiver operating characteristic curve of MPR (0.93) was significantly larger than those of stress and rest MBF, visual assessment of CMR-MPI, and CMRA (p < 0.05), but similar to that of the integration of CMR-MPI with CMRA (0.90).ConclusionsFully automated pixel-wise quantitative CMR-MPI can accurately detect hemodynamically significant CAD, but the integration of CMRA obtained between stress and rest CMR-MPI acquisition did not provide significantly additive value.