Abstract
1466
Objectives To investigate the relationship between injected dose (ID) & BMI on NECR & noise in PET images.
Methods Two groups of 90 patients were imaged on a GE DSTE & a DRX PET/CT scanner respectively. The patients in each group were divided into 9 subgroups according to 3 BMI (20-25, 25-30, >30) & 3 ID (8-12, 12-15, >15mCi) ranges, resulting in 10 patients/subgroup. All PET data were acquired in 3D and reconstructed using the VuePoint HD fully 3D OSEM (2it, 21SS). Noise & NECR for the bed positions covering the liver were calculated for each patient. Noise was determined as the STD of a fixed ROI drawn in the liver of each patient. NECR was calculated from the trues, randoms & scatters recorded in the DICOM header. A t-test assessed the noise & NECR from different scanners but with the same BMI & ID. An ANOVA test compared the results of patients with different BMI but same ID & scanner as well as different ID but same BMI & scanner.
Results T-test showed a significant difference in NECR between scanners for all BMI & ID subgroups (p<0.0005). No such findings were observed for noise. The ANOVA results showed a significant difference in NECR between different BMI for all ID & scanner subgroups (p<0.0005). However, there was a weak difference in noise and no statistically significant difference in NECR across different ID for a fixed BMI & scanner type.
Conclusions Initial results in the liver suggests that dose reduction from 15 to 8-12 mCi has minimal impact on image quality independent of the scanner used. A reduction in dose decreases patient exposure as well as radionuclide cost.
- © 2009 by Society of Nuclear Medicine