326. PET scanner qualification for clinical trial: Comparison between Italian and worldwide experience

Abstract

Purpose The aim of this work was to compare the results obtained by the procedure for qualifying PET/CT scanners for Clinical Trial Qualification (CTQ) adopted by the Italian Foundation on Lymphoma (FIL), the Grupo Espanol de Linfomas/Transplante Autologo de Medula Osea (GELTAMO) and the International Extranodal Lymphoma Study Group (IELSG). Methods Uniformity (UQP) and image quality (IQ) NEMA/IEC phantoms filled with 18F and acquired by local personnel were uploaded to a central server and analyzed within the CoreLab. Background activity concentration in UQP (BACUQP), Background activity concentration (BACIQ) and sphere to background ratio (SBR) in the IQ phantom were compared to expected values. Variability of these indexes among different PET/CT scanners was estimated in term of inter-scanner variability (ISV) at 95% confidence level. Results The number of PET/CT site qualified and the number of iteration necessary for the CTQ are summarized in Table 1. The BACUQP ISV at 95% confidence level among different scanners were 21.5%, 20.9% and 20.3% for Italian, Spanish and Worldwide PET/CT respectively. The BACIQ ISV were 63.0%, 36.4% and 40.4% and the SBR ISV were 57.9%, 57.5% and 35.5% for Italian, Spanish and Worldwide PET/CT, respectively. Conclusions The CTQ is a robust and reproducible procedure to verify inter-scanner calibration. The BACUQP is at least two times smaller than BACIQ demonstrating that the uniformity phantom, that is used for CTQ, is more accurately prepared by local sites. SBR demonstrated a great variability (ISV from about 35.5% to 57.9%) because it accounts both for variability in phantom preparation and in reconstruction algorithm tuning.