Abstract
242413
Introduction: Reported SPECT scan times for Lu-177 dosimetry vary widely, with 60 – 120 projections recommended, frame durations commonly 30 – 40s (10 – 80s reported), and 1 – 3 bed positions [2]. Currently, our Lu-177 dosimetry protocol uses 4 timepoints, 3 bed positions, 128 total projections, and a frame duration of 15 s, resulting in scan times of about one hour. These long scan times can be challenging to accommodate, both for patients and a busy clinic. One option for reducing scan times is to shorten the frame duration. The objective of this study is to validate a retrospective frame-shortening reconstruction, then use it to investigate the impact of shortened frame durations on a clinical dosimetry case.
Methods: For the phantom validation, a NEMA IEC PET Body phantom (Fig. 1) was filled with a 7.5:1 hot:background ratio of Tc-99m. SPECT/CT scans were acquired with frame durations of 15s, 7s, and 3s. The 15s scan was processed with a vendor-provided retrospective frame duration shortening reconstruction to produce simulated scans with frame durations of 7s and 3s. MIM Software was used to create regions of interest (ROIs) within the six hot spheres and the background. Mean counts per minute per activity (CPM/μCi) was calculated and compared between different images and ROIs.
The retrospective frame shortening was similarly applied to one patient’s dosimetry scans from 177Lu-PSMA-617 treatment (Fig. 2). MIM Software was used for quantitative image reconstruction and dosimetry calculation. Mean 24 hr SUV and mean absorbed dose in the lungs, liver, kidneys, and two lesions from the acquired images and frame-shortened reconstructions were compared.
A one-way ANOVA with Tukey’s Honest Significant Difference post-hoc test was used to test for difference in means.
Results: CPM/μCi of the hot ROIs were not statistically significantly different between all images, while the background showed significant differences for: 15s scan vs 7s scan, 15s scan vs 3s scan, 7s scan vs 3s recon, 3s scan vs 3s recon, and 3s recon vs 7s scan. Notably, the only significant difference between acquired and retrospectively shortened scans was the background ROI for the 3s scan and reconstruction, which showed a difference of –2.5% CPM/μCi.
Mean 24 hr SUV in each ROI was significantly different for the 15s acquired image and the 7s and 3s reconstructions. Absorbed dose was significantly different for all ROIs except lesion doses from the 15s vs 7s images and lesion 1 dose from the 7s vs 3s images. Both mean SUV and absorbed dose showed an increasing trend with decreasing frame duration, with absorbed dose increasing 3-12% for the 7s reconstruction, and 19-31% for the 3s reconstruction.
Conclusions: The phantom study indicated that the retrospective frame-shortened reconstruction provides realistic results, especially in higher-activity areas. Although statistically significant differences were found in the background ROI, these differences were small and not clinically meaningful.
Applying frame-shortening to patient dosimetry scans showed an increasing trend in both mean SUV and mean absorbed dose with decreasing frame duration, likely due to the increase in noise [1]. In this patient, over-estimation of uptake and dose could have impacted clinical decision making given that the resulting cumulative extrapolated kidney dose estimate for 6 cycles would be 27.8 Gy, 30.4 Gy and 34.7 Gy for the 15s, 7s, and 3s scans, respectively. Analysis of additional dosimetry scans is required to further assess the apparent correlation between frame duration and SUV and dose estimates, and to determine whether frame duration shortening could be implemented in a way that does not affect clinical decision making.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.