Abstract
1816
Purpose: Image uniformity of PET using pure water phantom is hard to visualize with PET/MR 3T unlike PET/CT. The water phantom for the PET/MR needs appropriate additives to control susceptibility and conductivity in the water. The purpose of this study was to investigate the optimal concentration of gadolinium diethylenetriamine pentaacetic acid (Gd-DTPA) and sodium chloride (NaCl) solution for acquiring good image uniformity with a water phantom on PET/MR. Materials and
Methods: A cylindrical phantom of 9.52L was filled with demineralized water, 21 kBq/ml F-18 solution, Gd-DTPA as a T1 relaxation modifier and NaCl as a conductivity modifier. Various concentrations of Gd-DTPA (0-0.2 mM) and NaCl (0-200 mM) mixtures were prepared for the phantoms. DIXON MR sequences were simultaneously performed with PET acquisitions. PET/MR image of each phantom was acquired for 10 minutes per bed position. PET data were reconstructed using OSEM with 2 iterations, 21 subsets and a 6 mm FWHM of Gaussian filter. The MR Image uniformity was estimated with a percent integral uniformity (PIU) and the nonuniformity of PET image was evaluated by a coefficient of variation (CV). Results: MR-based attenuation correction (MRAC) maps that acquired from the water phantom without Gd-DTPA or NaCl solution caused image artifacts along to a phase direction in MR. When Gd-DTPA was only added in the water phantom, artifacts of the MRAC showed a tendency to decrease but did not disappear completely. A mixture of 0.2 mM Gd-DTPA and 150 mM NaCl was the most suitable concentration for water filled uniformity phantom at the PET/MR image without artifacts. The PIU corresponding to the MRAC map and the DIXON image was 100%, 83.1% respectively. The CV of PET was 6.3%. Conclusion: We suggest that Gd-DTPA and NaCl are useful additives which are easily accessible and handle safely in clinical practice when performing a uniformity phantom test with water on PET/MR.