Frame Duration Optimization in SPECT/CT Imaging with Lu-177 for Radioligand Therapy using Joint Dual Photopeak Reconstructions

Ziel/Aim: Dosimetry in radionuclide therapy using ¹⁷⁷Lu plays an important role in maximizing treatment efficacy and requires, for instance, serial SPECT/CT imaging. Typically, the 208-keV single photopeak (SP) signal is used in image reconstruction. Incorporating the 113-keV photopeak signal into a joint dual-photopeak (DP) reconstruction could enhance image count statistics and potentially allow for a reduction of the total scan time. This reduces patient burden and optimizes utilization of the department resources. The aim of this study was to evaluate the quantification accuracy for reduced frame durations using the DP signal in reconstruction.

Methodik/Methods: A NEMA phantom with six spheres (diameters: 10.1 to 37.4 mm) was prepared with 2.3, 1.1 and 0.60 MBq/mL of ¹⁷⁷Lu solution in the spheres and a signal-to-background ratio of 100, mimicking typical tumor uptake values at 1, 3 and 7 days after injection of 7.4 GBq ¹⁷⁷Lu-PSMA. Acquisitions were conducted for five different frame durations (60, 20, 10, 5 and 2.5 s) on a Siemens Intevo SPECT/CT system using a clinical standard protocol for image quantification. Images were reconstructed using the 208-keV SP and the DP signals. Oversize approach was applied for partial volume effect corrections. Quantification accuracy of the SP- and DP-reconstructed images were assessed using respective images with high counts (60-s frame duration) as a reference.

Ergebnisse/Results: The absolute percentage mean (median) deviations for SP vs. DP from the reference images over all spheres and three activity concentrations were 4.3%, (1.4%) vs. 4.8% (2.2%) for 20s, 9.2% (4.8%) vs. 7.3% (2.3%) for 10 s, 13.9% (6.8%) vs. 9.5% (8.2%) for 5 s, 28.3% (14.3%) vs. 31.9% (12.9%) for 2.5 s. No notable differences in the quantitative performance between the SP- and DP-reconstructed images were found. Considering both the mean and median deviations, quantification deviations were below 5% for 20-s and below 10% for 10-s frame durations.

Schlussfolgerungen/Conclusions: DP reconstructions offer no notable advantage in the quantification performance for acquisitions with a reduced frame duration. A reduction of the frame duration to 10 s appears viable in our phantom setting. Further investigations will be conducted for lower activity concentrations.

Publication History

Article published online:
12 March 2025