Practical PET respiratory motion correction in clinical simultaneous PET/MR

Respiratory motion during PET acquisition leads to blurring in images and quantification underestimation. We propose a practical, anatomy independent MR-based correction strategy for PET data affected by motion, and show it can improve image quality for both PET acquired simultaneously to the motion-capturing MR, and for PET acquired up to 1 hour earlier during a clinical scan. For our method, a short additional PET/MR sequence is acquired to form a patient-specific MR-based motion model, with a PET-derived respiratory signal and a gradient echo 2D multi-slice sequence. To evaluate our method, uncorrected and motion-corrected images are compared using line profiles, and quantitatively with SUV_peak in avid lesions. We show that clinical PET image quality can be improved using only a short additional PET/MR acquisition with no external respiratory hardware, standard MR sequences and an open registration method.