Motion corrected abdominal PET/MRI using MR-based motion tracking

Abstract

Respiratory and bulk motion in abdominal PET studies lead to blurring of PET reconstructions. As a result, lesion volumes tend to be overestimated, while tracer uptake is underestimated and small lesions become undetectable. Motion correction may be used to correct PET images that would otherwise be blurred by motion. The purpose of this study is to investigate the feasibility of using MR-based motion tracking for motion correction on abdominal PET images acquired with PET/MRI. To study this, motion vector fields were generated using the novel MR sequence rCASPR. These vector fields were subsequently applied to simulated PET data acquired with the digital XCAT phantom. Motion correction of the reconstructions was done using the CASToR reconstruction framework complemented with a motion correction method based on deformation vector fields. The influence of only using either respiratory or bulk motion correction was studied on a simulation with both respiratory and bulk motion. The feasibility of using MRbased motion tracking with the rCASPR technique for motion correction of PET simulations was shown. Higher CRC values, CNR values and maximum uptake values and lower lesion volumes were found for the reconstruction with full motion correction than for the reconstruction without motion correction. Moreover, the relative impact of bulk motion correction versus respiratory motion correction could be assessed in this PET simulation containing both respiratory and bulk motion.