The influence of radioactivity ratio on harmonization of PET image

Purpose: The standardized uptake value (SUV) is primarily used for analysis of fluorodeoxyglucose (FDG) PET. The SUV is varied with PET scanner characteristics, acquisition protocols, and reconstruction algorithms. Harmonization is used to adjust the recovery coefficient of each PET scanner into the reference range and minimize the variation in SUV measurements. However, few studies, which have evaluated the influence of different radioactivities on harmonization using a phantom. The purpose of this study is to evaluate the influence of radioactivity ratio on the harmonization of a PET image using a phantom. Materials and Methods: A National Electrical Manufacturers Association (NEMA) body phantom was filled with different concentrations of 18F solution. The radioactivity ratios of hot spheres in comparison to the background were 4:1 and 2:1. The value 2.65 kBq/mL was chosen to represent the background activity. PET data were acquired from a 30 min scan using a Biograph mCT flow scanner (S1) and a Celesteion scanner (S2). The S1 PET images were reconstructed by using the ordered-subsets expectation maximization (OSEM) algorithm, OSEM with time-of-flight (TOF), OSEM with point spread function (PSF), and OSEM with PSF+TOF algorithms. On the other hand, S2 images were reconstructed by using the OSEM, OSEM with TOF, and OSEM with PSF+ TOF algorithms. The full-width at half-maximum of the Gaussian filter (GF-FWHM) varied from 1 mm to 10 mm. The harmonization of the PET image was evaluated by SUVmax. The SUVmax was plotted as a function of the sphere diameter and compared to the reference range proposed by the European Association of Nuclear Medicine (EANM). Results: At the radioactivity ratio of 4:1, GF-FWHM was that is included in the EANM reference range was changed by the reconstruction algorithms and scanners. For S1, the FWHM ranges of the OSEM, OSEM+PSF, OSEM+TOF, and OSEM+PSF+TOF were 5-8 mm, 6-9 mm, 6-8 mm, and 7-9 mm, respectively. For S2, the value of OSEM, OSEM+TOF, and OSEM+PSF+TOF were 3-5 mm, 7-9 mm, and 8-9 mm, respectively. At the radioactivity ratio of 2:1, the FWHM range of the OSEM, OSEM+PSF, OSEM+TOF, and OSEM+PSF+TOF for S1 were 3-4 mm, 4-6 mm, 3-4 mm, and 4-5 mm, respectively. For S2, the value of OSEM, OSEM+TOF, and OSEM+PSF+TOF were 4-5 mm, 5-6 mm, and 5-6 mm, respectively. When the radioactivity was low, the GF-FWHM that is to adjust in the reference range were small. Conclusions: The results of this study suggested that the harmonizati