Comparison of image quality with 62Cu and 64Cu-labeled radiotracers in positron emission tomography whole-body phantom imaging.

OBJECTIVE PET imaging is possible with copper (Cu) isotopes, (60)Cu, (61)Cu, (62)Cu and (64)Cu. Although (62)Cu- and (64)Cu-labeled radiotracers are often used for preclinical and clinical PET studies, we do not know which radiotracers have better image quality for tumor imaging. In this study, we compare image quality between (62)Cu and (64)Cu imaging with a different acquisition mode and reconstruction algorithm using a whole-body phantom for tumor imaging. METHODS In a National Electrical Manufacturers Association (NEMA) 2001 whole-body phantom, the concentration of (62)Cu-ATSM and (64)Cu-ATSM was, respectively, approximately 2.7 and 1.8MBq/mL in all the spheres and approximately 0.9 and 0.6MBq/mL in the background. After adjustment for true coincidence events between (62)Cu and (64)Cu, two-dimensional (2D) and three-dimensional (3D) PET scan data were acquired for 10min. The data were reconstructed using filtered back projection (FBP) and the ordered subset expectation maximization (OSEM) algorithm. Image quality of (62)Cu and (64)Cu was compared using recovery coefficient (RC), sphere-to-background ratio (SBR) and coefficient of variation (%COV). RESULTS There were little significant differences between (62)Cu and (64)Cu imaging, visually. Recovery coefficients of (64)Cu images were higher than those of (62)Cu images. The RC of (64)Cu images with 3D acquisition mode and OSEM was the highest in all experiments. No SBR values were significantly different from the true value of 3.0 in 37mm sphere diameters, but 3D acquisition and OSEM yielded slight overestimations compared with 2D acquisition and FBP, the gold standard for quantification in PET studies. Percentage COV values of (64)Cu with OSEM were significantly lower than those of (62)Cu. CONCLUSIONS Copper-64 radiotracers provide higher image quality than (62)Cu-radiotracers in whole-body tumor imaging only when the 3D acquisition mode and OSEM algorithm are applied. However, the quantitative values for smaller tumors may be slightly overestimated.