Making Sense of Multicenter PET Trials Through Harmonization

Background: Differences in scanner capabilities and image processing result in variations in standardized uptake values (SUVs) determined from positron emission tomography (PET) 18 F -fluorodeoxyglucose (FDG) studies performed at different facilities. For meaningful multicenter trial results, there must be methods for comparing and normalizing SUVs (harmonization). Objective: To evaluate 3 PET quality-control phantoms and several SUV approaches to determine which provide the best harmonization for PET multicenter clinical trials. Methods: All PET studies were acquired with the same time-of-flight PET/CT imaging system, and the data were reconstructed with 5 different protocols to obtain images with a range of noise and spatial resolutions. The 3 phantoms investigated were the American College of Radiology (ACR) PET phantom, the National Electrical Manufacturer's Association (NEMA) NU-2 PET image-quality phantom, and the (Society of Nuclear Medicine (SNM) Clinical Trials Network anthr more...