The effect of respiratory gating on quantitative PET imaging in lung cancer

256 Objectives: In PET/CT, motion in tissues such as the lungs results in overestimation of tracer-avid volumes of interest and underestimation of their intensities. Respiratory gating has been shown to reduce these effects in phantoms and sample patients. We demonstrate its effect in a clinical setting with a larger number of patients. Methods: Using a GE Discovery LS PET/CT scanner equipped with a Varian RPM respiratory gating system 18 patients with known pulmonary cancer received both ungated and gated 18FDG-PET/CT scans as part of their regular staging. Lesion volumes were determined in CT by direct measurement and in PET using adaptive thresholding (Erdi et al. 1997) with software running on GE AW and Xeleris workstations. Maximum, minimum and average SUVs were determined. Statistical significance was tested using a two-tailed paired Student’s t-test and confirmed with a Wilcoxon test. In order to rule out a confounding influence of image noise on the gated studies, signal-to-noise-ratios (SNRs) were compared. Results: With respiratory gating, the average lesion area per slice as well as extent in the Z direction were significantly lowered (by 15.5% and 10.3%, p = 0.014 and p = 0.007), resulting in a decrease in measured tumor volume of 44.5% (p = 0.025). The difference between lesion volumes in gated PET and CT was significantly lower than the difference in ungated PET and CT (-126.6%, p = 0.025). Gating increased SUVmax by 22.4% (p