Quantitative imaging with low-dose CT in the PET/CT system

Abstract

The use of PET/CT systems in clinical applications has rapidly increased of late. Not only has their diagnostic value increased due to their superior statistical noise characteristics; they have also shortened the acquisition time for attenuation correction compared to a single-modality PET system. Most commercially available PET/CT systems use the CT scan for attenuation correction instead of the transmission scan, which uses a radioactive source, such as 137Cs or 68Ge. The effective dose of CT attenuation correction, however, is up to 118 times larger than that of radioactive isotope attenuation correction. As such, an increased dose is expected during the CT scan. The purpose of this study is to evaluate the quality of PET images that have been reconstructed using a CT attenuation map in various CT tube currents. Image data were obtained, using a 3D Hoffman brain phantom and a cylindrical phantom, for both the GE DSTe PET/CT and Philips GEMINI PET/CT systems. The acquisition protocols of the DSTe and GEMINI systems for PET studies were based on the parameters used by Turkington et al. and Gregory et al., respectively, in the studies they conducted. The CT data that were used to create an attenuation map were acquired from various tube currents, from a diagnostic dose to an ultra-low dose. The CT image quality was evaluated by measuring the standard deviation (SD) of the cylindrical phantom filled with water. The PET image quality was evaluated by computing the activity ratio of the gray and white matter in the 3D Hoffman brain phantom images. In addition, the coefficient of variance (COV) was calculated to evaluate the CT and PET images. The SDs of the CT data and the COVs of the CT images decreased when the tube currents increased. The activity ratios of the gray and white matter with a ventricle area in the PET images remained almost constant for the CT attenuation maps that used various tube currents. The DSTe and GEMINI systems' activity ratios were approximately 3.76plusmn0.03 and 3.86plusmn0.03. The COVs of the PET images were almost the same in spite of the fact that the PET images' attenuations were corrected with the use of different CT data. The results indicated that the quality of the PET images whose attenuations were corrected using a low-dose CT was comparable to that of the PET images whose attenuations were corrected using a diagnostic-dose CT. In conclusion, if the CT exam does not require a high-quality diagnostic CT, the use of the low-dose CT rather than the diagnostic-dose CT is recommended.