Multicenter integrated positron emission tomography/computer tomography imaging quality assurance accreditation interobserver reliability study with the American College of Radiology phantom

Background: Integrated positron emission tomography/computer tomography (PET/CT) image quality assurance ensures accurate, reproducible, and quantitative assessment of comparable scanner performance. We performed a preliminary multicenter PET/CT imaging quality assurance test with a fillable tomographic phantom in six medical centers in Taiwan. Methods: The phantom was filled with fixed proportions of fluorine-18 radionuclide solution in the background and with different spheres to simulate cold and hot lesions, and body background radioactivity. Imaging acquisitions were performed by using recommended parameters in different sites according to different brand names of the instrument. All imaging was subjectively scored by eight experienced nuclear medicine physicians as the spatial resolution of four hot vials (score 0-4), six cold spheres (score 0-6), and six cold rod areas (score 0-6), and overall satisfaction (score 0-5). Interobserver correlation and receiver operating characteristic (ROC) curve were analyzed. Results: The detection ability of hot vials, cold spheres, and cold rods was 4.0 ± 0.1, 5.2 ± 0.8, and 3.8 ± 0.9, respectively. Overall satisfaction was 4.0 ± 0.8. The ROC analysis revealed that the area under the curve for hot vials, cold spheres, and cold rods was 0.984, 0.887, and 0.928 respectively. The interobserver correlation for detectability of cold spheres and cold rods was 0.88 and 0.96, respectively. Conclusion: The results of the study indicated that (1) PET/CT imaging quality assurance for comparable scanner performance could be established on the basis of a standard phantom and (2) good interobserver correlation can be observed for those with accurate and interpretable results.