Evaluation of modified window-based scatter compensation in quantitative 177Lu-SPECT for a ring-configured CZT SPECT-CT

Abstract Background : The aim was to evaluate bias and precision for 177 Lu-activity-concentration estimation for window-based scatter compensation in a ring-configured CZT gamma camera. This paper extends a previous study by applying modified scatter compensation (MSC) methods. Whilst the original compensation addressed tailing in the 208 keV peak only, separating primary and scatter photons within the scatter window, the current adaption extents the concept to the 113 keV peak. In the modified version, the estimated primary signal in the scatter window can optionally also be added to measured projections to boost the signal-to-noise ratio (MSC + P). Methods : No additional measurements were acquired, and only previously collected listmode-data for a uniform cylindrical phantom, an image quality NEMA phantom and an anthropomorphic phantom were reframed to accommodate the updated window settings recommended for MSC and MSC + P. Reconstructions were performed using OS-EM with two to 30 iterations (10 subsets) with compensation for attenuation, scatter (DEW/TEW, MSC and MSC + P), distance-dependent spatial resolution, and penetration at 208 keV. Volume-of-interest following the physical sphere size were defined and activity-concentration estimates for each sphere, a liver region and for total activity in the phantoms were assessed in terms of bias and precision for short (10 min) and long timeframes (60 min). Results : Imaging at 208 keV generally result in similar bias and precision and total activity estimates for all scatter-compensation methods. For 113 keV, a slightly netter precision is obtained for MSC and MSC + P but at the cost of larger bias compared with TEW. The major difference is seen for total activity where the MSC and MSC + P manages to decrease bias considerably. The MSC + P yields a decrease in bias of about 10 percentage points when comparing short versus long time frames for activity concentration estimates at 113 keV. Conclusions : Despite the lack of broader improvement in bias and precision when using MSC and MSC + P the findings suggest potential value in more targeted applications. For instance, MSC and MSC + P improves preservation of total activity. Although the findings indicate that using MSC + P is not advisable for estimating high activity concentrations, it may positively influence the estimation of low activity concentrations.