Page 34 - FiM_03_2016

reached at the first round in 63% of the cases, while in 31% and 6% were required two and more than five iterations respectively. For the PET/CT scanners fulfilling the CTQ the difference between measured and expected BAC in the uniformity phantom was 0.4±5.0% (CI 95%: -10.3% to +11.1%, ISV=21.4%). The difference between measured and expected BAC in the IQ phantom was - 3.4±9.3% (CI 95%: -23.5% to +16.7%, ISV=40.2%). The difference between measured versus expected SBR in the IQ phantom was -3.7±9.3% (CI 95%: -16.3% to +23.8%, ISV=40.1).
The 68Ge phantom was circulated across 20 scanners. Twenty out of 20 (100%) scanners fulfilled the CTQ requirements. For qualified scanners the CTQ was reached at the first round in 95% of the cases.
Differences between measured and expected BAC and SBR were –1.1±5.8% (CI 95%: -13.3% - +11.1%; ISV=24.4%) and 4.5±4.7% (CI 95%: -5.4% - +14.4%; ISV=19.8%), respectively.
CONCLUSIONS
The difference between expected and measured BAC was lower than ±10% and ISV was 21.2%. But when BAC was measured in the NEMA IQ phantom, which was not used for qualification, we observed a much higher variability with ISV 67.8%. RC curves were between the minimum and maximum RC limits. The high variability of SBR, ISV=50.6%, was accounted by the high probability of error in the filling procedures and the objective difficulty in phantom preparation. In particular the highest variability was found in small-sized spheres.
The results of the Spanish PET/CT and worldwide sites were similar to the data obtained from the Italian sites.
The low PET/CT site compliance of the 18F phantom-based CTQ program encouraged us to adopt the 68Ge phantom approach, so as to simplify the process and further reduce the inter-scanner variability.
Besides the high variability, several PET sites (20% of scanners) declared themselves unable to accomplish the CTQ, and up to 49% of the scanners required repeated iterations with the core lab while only 26% of them were qualified in a single round. The more critical aspects were the time needed for the CTQ procedure and the absence of dedicated and trained personnel to perform the tests. Some sites lacked the dedicated UQP or IQ phantoms, some could not cover the cost of the procedure for 18F phantom preparation, and some lacked time to dedicate to quality control of PET/CT scanners. These problems were resolved with the 68Ge phantom. It was shipped to the sites as a sealed source. Unpacking and positioning the phantoms required about 5 minutes as compared
to the nearly 2 hours needed for 18F phantom preparation. Moreover, the radiation exposure for the 26
Tesi di Specialità