Abstract
P827
Introduction: Bone marrow (BM) is the dominant dose-limiting organ in 177Lu-PSMA therapy for patients with metastasized castration resistant prostate cancer, where BM dosimetry is challenging due to segmentation and activity quantification particularly from partial volume effect. This study aims to develop an automatic image-based segmentation method and apply a partial volume correction (PVC) method based on post-therapeutic sequential 177Lu-PSMA-617 images for personalized BM dosimetry.
Methods: After 6.7-8.4 GBq 177Lu-PSMA-617 injection based on fixed activities adjusted by patients situation, sequential quantitative SPECT/CT imaging at 2, 20, 40 and 60 (n=4) /200 (n=6) h were analyzed retrospectively for 10 patients in 10 treatment cycles (first: second: third cycle=7:2:1). PVC was performed on SPECT images using the reblurred van-cittert deconvolution method in the PETPVC toolbox [1], based on the reported system resolution [2]. CT images at latter time points were rigidly registered to those acquired at the first time point using Elastix [3], focusing on the spine region by applying the spine region mask. The resultant deformed vectors were applied to align the corresponding whole SPECT images. Voxel-S-value kernel with density correction was applied on the registered SPECT images to generate dose rate maps. On the other hand, X means clustering was applied on the segmented lumber spines (L1-L5) CT images to classify the lowest Hounsfield Unit (HU) region as BM. The resultant BM masks were then mapped to the dose rate maps. Mean BM absorbed dose was obtained using the bi-exponential curving fitting followed by integration.
Results: The median (range) %bone segmented as BM by X means clustering is 58.53% (48.57%-72.98%). The median of mean BM absorbed dose on SPECT with PVC is 0.28 Gy, 33.33% lower than that of measured on SPECT without PVC (0.42 Gy). The maximum of measured mean BM absorbed doses are 1.69 Gy and 1.77 Gy for with and without PVC on SPECT, all less than the suggested upper limit of 2 Gy [4].
Conclusions: X means clustering is feasible to segment the BM based on the CT images of post-therapy SPECT/CT. Generally, PVC leads to a lower BM absorbed dose estimation by potentially reducing the spill-out effect from the high uptake bone lesions adjacent to BM. BM segmentation results from experienced physicians are warranted to verify our proposed X means clustering segmentation method.
[1] Thomas, Benjamin A., et al. "PETPVC: a toolbox for performing partial volume correction techniques in positron emission tomography." Physics in Medicine & Biology 61.22 (2016): 7975.
[2] Huizing, Daphne MV, et al. "177Lutetium SPECT/CT: Evaluation of collimator, photopeak and scatter correction." Journal of Applied Clinical Medical Physics 21.9 (2020): 272-277.
[3] Klein, Stefan, et al. "Elastix: a toolbox for intensity-based medical image registration." IEEE transactions on medical imaging 29.1 (2009): 196-205.
[4] Sjögreen Gleisner, Katarina, et al. "EANM dosimetry committee recommendations for dosimetry of 177Lu-labelled somatostatin-receptor-and PSMA-targeting ligands." European Journal of Nuclear Medicine and Molecular Imaging 49.6 (2022): 1778-1809.
Research support: This work is supported by a research grant from Macao Science and Technology Development Fund (FDCT 0099/2021/A).
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.