Effects of CT-based Attenuation Correction on PET Images Using Data-driven Respiratory Gating

Abstract

We applied deviceless, positron emission tomography/computed tomography(PET/CT) data-driven respiratory gating (DDG) to validate the effects of misalignment between PET and CT at various respiratory phases. A lung lesion was simulated using an NEMA IEC body phantom in which the background comprised hot spheres containing polystyrene foam beads. We acquired PET images as the phantom moved downwards and then stopped. Attenuation on computed tomography images acquired at the inspiratory, stationary, and expiratory phases was corrected after the phantom stopped moving. Normalized mean square error (NMSE), recovery coefficients (RCmax and RCmean) and volume were analyzed on DDG-PET images using CT-based attenuation correction. The NMSE was closest to 0 in PET images corrected using the expiratory CT image. The RCmax was<1.0, and the RCmean was closest to 1.0 only in PET images corrected using the expiratory CT image. Volume was either underestimated or overestimated more according to the size of the spheres when the alignment of CT and PET images was greater. We recommend using the expiratory but not the inspiratory phase when using DDG for PET/CT correction.