Motion effects on SUV and lesion volume in 3D and 4D PET scanning

To assess the effect of lesion motion and respiration rate on Standardised Uptake Value (SUV) and the ability of 4D PET to restore any loss in SUV and distortion of lesion volume on two PET/CT systems. A Perspex phantom with four cylindrical reservoirs filled with 18F-FDG was used in this study. The cylinders measured 5, 10, 15, and 20 mm in diameter. A GE Discovery STE8 (GE Medical Systems Milwaukee, WI) and a Siemens Biograph 64/40 (Siemens Medical Solutions, Erlangen, Germany) scanner was used to acquire a stationary un-gated PET scan of the phantom. Multiple 10 min list mode 4D PET scans were acquired using the Varian RPM on the GE camera and the Anzai Gating system on the Siemens camera. The phantom was scanned at five different respiratory rates and motion amplitudes in a sinusoidal fashion, 15 RPM/1 cm, 15 RPM/2 cm, 15 RPM/4 cm, 30 RPM/2 cm and 7.5 RPM/2 cm (RPM-respirations per minute). Each scan was reconstructed into ten bins and as an un-gated static image. The SUVmax, SUVmean and volume were measured for all four reservoirs using Siemens TrueD analysis software. With increasing lesion movement the SUVmax and SUVmean decreased and the volume increased with the SUVmax in the smallest lesion underestimated by up to a factor of four. The SUVmax, SUVmean and volume were mostly recovered using 4D imaging regardless of amount of lesion displacement. The larger lesions showed better count recovery and volume correction than the smaller lesions. The respiratory rate had no effect of SUV or volume. Un-gated imaging of moving lesions decreases apparent SUV in small lesions significantly and overestimates volumes. 4D PET scanning recovers most of the apparent loss in SUV and distortion of volumes.