Program Information
Brain PET/CT Imaging On a Scanner with a Large Axial Field-Of-View
M Park*, V Gerbaudo , L Hamberg , K Seaver , M Kijewski , Brigham and Women's Hospital and Harvard Medical School, Boston, MA
Presentations
SU-D-9A-4 Sunday 2:05PM - 3:00PM Room: 9APurpose: Large axial field-of-view (FOV) PET/CT scanners are valued for high sensitivity. Brain PET image quality may depend on the head position within the FOV. We investigated the precision of activity estimation for brain PET imaging when the brain was positioned at the end (END) and in the middle (CEN) of the FOV. The additional CT dose for the CEN position was recorded.
Methods: An image quality (Jaszczak) phantom and a striatal phantom were filled with F-18 and positioned in END and CEN locations. For each phantom and each location, we acquired a ~1-hr listmode PET, rebinned the data into 10 frames with equal number of coincidence events, and reconstructed each frame using an iterative algorithm. For the striatal phantom, END and CEN were compared by drawing on each image three regions of interest (ROI) in axially separated uniform areas. The standard deviation of the activity estimation within each ROI was averaged over the 10 images. The coefficient of variation (CV) for activity estimation was calculated at each position. Image quality was assessed by inspecting the resolution bar pattern in the Jaszczak phantom at two different head positions.
Results: The CV was the lowest for ROIs near the center of the FOV. For slices near the end, not only was the CV highest, but also the resolution pattern was degraded. CTDIvol summarized in the dose report indicated that the CT dose was ~ 10% higher for CEN as compared to END position.
Conclusion: Positioning the brain in the middle of the FOV in a large FOV PET/CT scanner allows more precise measurement of tracer uptake and better image quality at the cost of increased CT dose. For the end location longer scan times may minimize image quality degradation without any additional CT dose.
Contact Email: