Monte Carlo estimation of scatter effects on quantitative myocardial blood flow and perfusable tissue fraction using 3D-PET and 15O-water

Abstract

In clinical cardiac positron emission tomography using 15O-water, significant tracer accumulation is observed not only in the heart but also in the liver and lung, which are partially outside the field-of-view. In this work, we investigated the effects of scatter on quantitative myocardium blood flow (MBF) and perfusable tissue fraction (PTF) by a precise Monte Carlo simulation (Geant4) and a numerical human model. We assigned activities to the heart, liver, and lung of the human model with varying ratios of organ activities according to an experimental time activity curve and created dynamic sinograms. The sinogram data were reconstructed by filtered backprojection. By comparing a scatter-corrected image (SC) with a true image (TRUE), we evaluated the accuracy of the scatter correction. TRUE was reconstructed using a scatter-eliminated sinogram, which can be obtained only in simulations. A scatter-uncorrected image (W/O SC) and an attenuation-uncorrected image (W/O AC) were also constructed. Finally, we calculated MBF and PTF with a single tissue-compartment model for four types of images. As a result, scatter was corrected accurately, and MBFs derived from all types of images were consistent with the MBF obtained from TRUE. Meanwhile, the PTF of only the SC was in agreement with the PTF of TRUE. From the simulation results, we concluded that quantitative MBF is less affected by scatter and absorption in 3D-PET using 15O-water. However, scatter correction is essential for accurate PTF.