Performance evaluation of quantitative SPECT/CT using NEMA NU 2 PET methodology

摘要

Although PET is routinely evaluated using NEMA NU2 as standard in the clinic, standard methodology for evaluating the performance of quantitative SPECT systems has not been established. In this study, the quantitative performance of the Symbia Intevo SPECT/CT was evaluated for two common isotopes (Tc-99m, Lu-177) and benchmarked against the performance of a PET/CT. A further aim was to demonstrate the utility of adapting NEMA NU2 PET measurements to SPECT. In addition, dead-time and resolution recovery were evaluated to provide more complete system evaluations. Spatial resolution of the SPECT system at 1 cm from the center in the transverse direction was 13.1 mm and 22.4 mm for Tc-99m and Lu-177 respectively, compared with 4.3 mm (F-18) and 5.8 mm (Ga-68) for PET. Sensitivity at the center of the FoV was 119 cps MBq(-1) and 48 cps MBq(-1) ((99)mTc,Lu-177) for SPECT and 9632 cps MBq(-1) and 8216 cps MBq(-1) (F-18,Ga-68) for PET. Scatter fraction was 0.25 and 0.36 ((99)mTc,Lu-77) for SPECT and 0.32 and 0.29 (F-18, Ga-68) for PET. Contrast recovery coefficient in the largest spheres was 0.79 and 0.65 (Tc-99m,Lu-177) for SPECT, 1.00 and 0.97 (F-18,Ga-68) for PET and the background variability was 2.7%, 6.5% (Tc-99m,Lu-177), 1.5% and 1.6% (F-18,Ga- 68), respectively. Partial volume effect was evaluated using the NEMA IQ phantom with six sphere inserts (diameter: 37 mm, 28 mm, 22 mm, 17 mm, 13 mm and 10 mm). Full contrast recovery was reached with the 17 mm for F-18, while SPECT did not reach full recovery for any sphere. Count rate losses were 2% for Tc-99m at 1 GBq and 11% for Lu-177 at 8.5 GBq which are well below the typical activities for clinical applications. We concluded NEMA NU2 methodology can be easily adapted to SPECT/CT as a routine quality assurance procedure in the clinic.