Partial volume correction for Lu-177-PSMA SPECT

Lu-177 PSMA therapy of metastatic prostate cancer requires sequential SPECT/CT imaging for dosimetry. However, the limited spatial resolution in SPECT images leads to the severe partial volume effect. In this study, we aimed to quantitatively analyze the influence of PVE and effectiveness of a partial volume correction (PVC) method in Lu-177 PSMA SPECT images. A cylindrical phantom (diameter: 30 cm) with seven spheres (diameter from 10 to 50 mm, center-to-center distance of 48-60 mm, activity concentration of 2 MBq/mL), an XCAT phantom with 10 spherical lesions (diameter from 10 to 30 mm) located in various locations of the spine were modelled based on the activity distributions of clinical data. Sixty projections were simulated for a clinical dual-head SPECT/CT with medium energy general propose collimators using SIMIND Monte Carlo code and reconstructed with OS-EM algorithm with dual energy window scatter correction, attenuation correction and geometric collimator-detector-response correction. Sequential quantitative SPECT/CT imaging at 2, 20, 40 and 60 h of five patients treated with 6.0-7.4 GBq Lu-177-PSMA-617 were analyzed retrospectively. The reblurred Van-Cittert (RVC) iterative deconvolution PVC method was applied on the post-reconstructed simulated and clinical SPECT images. In the geometrical phantom, the quantification error decreased as the sphere size increased. PVC reduced the quantification error from 87.7% to 85.7% for the smallest sphere and from 14.8% to 4.9% for the largest one. For the XCAT phantom, PVC reduced the activity quantification error of bone lesions from 50.1%±2.5 to 42.5%±3.7 and kidneys from 41.6%±4.7 to 37.8%±16.7 respectively. For clinical data, the mean activity difference w/ and w/o PVC was 24.4%±11.9 for tumor and 10.7%±5.6 for kidneys at 60 h. RVC-based PVC improves quantification accuracy in Lu-177 PSMA SPECT.