18F, 11C and 68Ga in small animal PET imaging

Aim: The partial volume effect (PVE) significantly affects quantitative accuracy in PET. In this study we used a micro-hollow sphere phantom filled with F-18, (11)c or Ga-68 to evaluate different partial volume correction meth{ ods (PVC). Additionally, phantom data were I applied on rat brain scans to evaluate PVC methods on in vivo datasets. Methods: The I four spheres (7.81, 6.17, 5.02, 3.90 mm inner diameter) and the background region were filled to give sphere-to-background (sph/bg) activity ratios of 20:1, 10:1, 5:1 and 2:1. Two different acquisition and reconstruction protocols and three radionuclides were evaluated using a small animal PET scanner. I From the obtained images the recovery coefficients (RC) and contrast recovery coefficients (CRC) for the different sph/bg ratios were calculated. Three methods for PVC were evaluated: a RC based, a CRC based and a volume of interest (VOl) based method. The most suitable PVC methods were applied to in vivo rat brain data. Results: RCs were shown to be dependent on the radionuclide used, with the highest values for F-18, followed by C-11 and Ga-68. The calculated mean CRCs were generally lower than the corresponding mean RCs. Application of the different PVC methods to rat brain data led to a strong increase in time-activity curves for the smallest brain region (entorhinal cortex), whereas the lowest increase was obtained for the largest brain region (cerebellum). Conclusion: This study was able to show the importance and impact of PVE and the limitations of several PVC methods when performing quantitative measurements in small structures.