OBJECTIVESQuantitative myocardial perfusion imaging by computed tomography (CT) was recently introduced to calculate myocardial blood flow (MBF). Because absolute MBF thresholds may be affected by technique, methodology, and the microvasculature, we investigated whether a relative measure of MBF improves accuracy to identify hemodynamically significant coronary stenosis. MATERIALS AND METHODSIn this prospective study, 42 patients (mean SD age, 62.38.7 years; 8 women) with suspected or known coronary disease underwent dynamic CT myocardial perfusion imaging using adenosine vasodilation, before invasive angiography (coronary angiography) with fractional flow reserve (FFR). Within each myocardial territory MBF, the MBF relative to remote myocardium (MBFratio) was calculated and compared with coronary angiography and FFR. RESULTSOf the 91 vessels interrogated by FFR (median, 0.81; interquartile range, 0.73–0.94), 45 vessels (49%) had an FFR value lower than 0.8 and were considered hemodynamically significant. Hyperemic MBF was lower in ischemic territories75.6 ± 22.5 mL per 100 mL/min versus 98.3 ± 23.1 mL per 100 mL/min (P < 0.0001). The MBFratio correlated better with FFR (P = 0.76) than the absolute MBF did (P = 0.52). Receiver operating curve analysis showed better discrimination by MBFratioarea under the curve of 0.85 versus 0.75 (P = 0.02). The MBF of remote myocardium varied between 60.7 and 167.2 mL per 100 mL/min and was lower in patients without heart rate acceleration (P = 0.0035). CONCLUSIONSThe MBFratio seems to better identify hemodynamically significant coronary artery disease than does the absolute MBF determined by dynamic CT perfusion imaging. This may be caused by microvascular status or related to the methodology.