Purpose: Respiratory motion of organs during PET scans is known to degrade PET image quality, potentially resulting in blurred images, attenuation artefacts and erroneous tracer quantification. List mode-based gating has been shown to reduce these pitfalls in cardiac PET. This study evaluates these intrinsic gating methods for tumour PET scans.
Methods: A total of 34 patients with liver or lung tumours (14 liver tumours and 27 lung tumours in all) underwent a 15-min single-bed list mode PET scan of the tumour region. Of these, 15 patients (8 liver and 11 lung tumours in total) were monitored by a video camera registering a marker on the patient's abdomen, thus capturing the respiratory motion for PET gating (video method). Further gating information was deduced by dividing the list mode stream into 200-ms frames, determining the number of coincidences (sensitivity method) and computing the axial centre of mass of the measured count rates in the same frames (centre of mass method). Additionally, these list mode-based methods were evaluated using only coincidences originating from the tumour region by segmenting the tumour in sinogram space (segmented sensitivity/centre of mass method). Measured displacement of the tumours between end-expiration and end-inspiration and the increase in apparent uptake in the gated images served as a measure for the exactness of gating. To estimate the accuracy, a thorax phantom study with moved activity sources simulating small tumours was also performed.
Results: All methods resolved the respiratory motion with varying success. The best results were seen in the segmented centre of mass method, on average leading to larger displacements and uptake values than the other methods. The simple centre of mass method performed worse in terms of displacements due to activities moving into the field of view during the respiratory cycle. Both sensitivity- and video-based methods lead to similar results.
Conclusion: List mode-driven PET gating, especially the segmented centre of mass method, is feasible and accurate in PET scans of liver and lung tumours.