Head motion is a well known problem in brain PET imaging. In human PET, paediatric patients are especially prone to motion and must be often be sedated or anaesthetized to prevent motion artifacts. A similar situation applies in pre-clinical imaging, where laboratory animals are invariably anaesthetized prior to the PET imaging procedure to avoid motion.
There has been considerable effort recently in the pre-clinical setting to develop ways to image small animals with PET in their unanaesthetized state, motivated by the fact that anaesthesia can significantly alter the physiological parameters under study. As part of this effort we have adapted methods previously developed for motion correction in human PET to small animal imaging, and demonstrated the feasibility of motion correction in this environment. Motion correction is achieved by acquiring the PET data in list mode, and applying spatial transformations, derived from simultaneously acquired motion tracking data, to individual events. We show that with accurate synchronization of PET and motion data, and an adequate motion sampling rate, it is possible to compensate very effectively for head motion, even when the motion is rapid and continuous. These results suggest that a motion correction scheme may be equally effective in paediatric studies. This could enable paediatric scans to be performed successfully with less frequent recourse to sedation and anaesthesia.