Quantification of Myocardial Perfusion Using Dynamic 64-Detector Computed Tomography

Abstract

The purpose of this study was to determine the ability of dynamic 64 slice multidetector computed tomography (d-MDCT) to provide an accurate measurement of myocardial blood flow (MBF) during first-pass d-MDCT using semiquantitative and quantitative analysis methods. Six dogs with a moderate to severe left-anterior descending artery stenosis underwent adenosine (0.14 mL · kg−1 · min−1) stress d-MDCT imaging according to the following imaging protocol: iopamidol 10 mL/s for 3 seconds, 8 mm × 4 collimation, 400 milliseconds gantry rotation time, 120 kV, and 60 mAs. Images were reconstructed at 1-second intervals. Regions of interest were drawn in the LAD and remote territories, and time-attenuation curves were constructed. Myocardial perfusion was analyzed using a model-based deconvolution method and 2 upslope methods and compared with the microsphere MBF measurements. The myocardial upslope-to-LV-upslope and myocardial upslope-to-LV-max ratio strongly correlated with MBF (R2 = 0.92, P < 0.0001 and R2 = 0.87, P < 0.0001, respectively). Absolute MBF derived by model-based deconvolution analysis modestly overestimated MBF compared with microsphere MBF (3.0 ± 2.5 mL · g−1 · min−1 vs. 2.6 ± 2.7 mL · g−1 · min−1, respectively). Overall, MDCT-derived MBF strongly correlated with microspheres (R2 = 0.91, P < 0.0001, mean difference: 0.45 mL · g−1 · min−1, P = NS). d-MDCT MBF measurements using upslope and model-based deconvolution methods correlate well with microsphere MBF. These methods may become clinically applicable in conjunction with coronary angiography and next generation MDCT scanners with larger detector arrays and full cardiac coverage.