Medical physics · 2025
Validation of the collapsed-cone superposition for whole-body patient-specific dosimetry in [177Lu]Lu-PSMA-617 radionuclide therapy.
СтатьяВалидационное исследованиеНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Terro Aya, Perret Solène, Dumouchel Arthur, Tonnelet David, Edet-Sanson Agathe, Vera Pierre, Decazes Pierre, Dieudonné Arnaud
Аннотация
BACKGROUND: Patient-specific dosimetry in radiopharmaceutical therapy (RPT) offers a promising approach to optimize the balance between treatment efficacy and toxicity. The introduction of 360° CZT gamma cameras enables the development of personalized dosimetry studies using whole-body single photon emission computed tomography and computed tomography (SPECT/CT) data.
PURPOSE: This study proposes to validate the collapsed-cone superposition (CCS) approach against Monte Carlo (MC) simulations for whole-body dosimetry of [177Lu]Lu-PSMA-617 therapy in patients with metastatic castration resistant prostate cancer (mCRPC).
MATERIALS AND METHODS: Thirty patients with mCRPC were retrospectively included in this study. SPECT/CT images were acquired after the infusion of [177Lu]Lu-PSMA-617 therapy. SimpleDose was used to generate dose-rate maps (mGy/h) from a single SPECT/CT scan. The dosimetry relies on the CCS approach, which adjusts dose-point kernels according to tissue densities. Organ and lesion delineation were automated using the nnU-Net V2 neural network. MC simulations were performed with GATE 10 for 108 events. To assess the impact of density-scaled DPK on the accuracy of the dosimetry, we implement a simplified version of CCS, denoted as C C S S T $CC{S}_{ST}$ , which assumes a homogeneous soft tissue medium without incorporating the patient-specific density information derived from the CT image. The comparison between CCS, C C S S T $CC{S}_{ST}$ and MC was conducted at the organ, lesion, and voxel levels.
RESULTS: Absolute percentage errors (APE) between CCS and MC were < 5% for all organs and lesions. Compared to CCS, C C S S T $CC{S}_{ST}$ exhibited higher APE with respect to MC in the liver, lungs, salivary glands, and lesions, while lower errors were observed in the bone marrow, kidneys, and pancreas, with comparable performance in the spleen. Voxel-level errors were mostly < 2% for both methods CCS and C C S S T $CC{S}_{ST}$ . Median computation time was, respectively, 24.5 s, 46.45 s, and 6.8 h for CCS, C C S S T $CC{S}_{ST}$ , and MC.
CONCLUSION: CCS showed high agreement with MC with greater computational efficiency, demonstrating its clinical potential for whole-body dosimetry.
PURPOSE: This study proposes to validate the collapsed-cone superposition (CCS) approach against Monte Carlo (MC) simulations for whole-body dosimetry of [177Lu]Lu-PSMA-617 therapy in patients with metastatic castration resistant prostate cancer (mCRPC).
MATERIALS AND METHODS: Thirty patients with mCRPC were retrospectively included in this study. SPECT/CT images were acquired after the infusion of [177Lu]Lu-PSMA-617 therapy. SimpleDose was used to generate dose-rate maps (mGy/h) from a single SPECT/CT scan. The dosimetry relies on the CCS approach, which adjusts dose-point kernels according to tissue densities. Organ and lesion delineation were automated using the nnU-Net V2 neural network. MC simulations were performed with GATE 10 for 108 events. To assess the impact of density-scaled DPK on the accuracy of the dosimetry, we implement a simplified version of CCS, denoted as C C S S T $CC{S}_{ST}$ , which assumes a homogeneous soft tissue medium without incorporating the patient-specific density information derived from the CT image. The comparison between CCS, C C S S T $CC{S}_{ST}$ and MC was conducted at the organ, lesion, and voxel levels.
RESULTS: Absolute percentage errors (APE) between CCS and MC were < 5% for all organs and lesions. Compared to CCS, C C S S T $CC{S}_{ST}$ exhibited higher APE with respect to MC in the liver, lungs, salivary glands, and lesions, while lower errors were observed in the bone marrow, kidneys, and pancreas, with comparable performance in the spleen. Voxel-level errors were mostly < 2% for both methods CCS and C C S S T $CC{S}_{ST}$ . Median computation time was, respectively, 24.5 s, 46.45 s, and 6.8 h for CCS, C C S S T $CC{S}_{ST}$ , and MC.
CONCLUSION: CCS showed high agreement with MC with greater computational efficiency, demonstrating its clinical potential for whole-body dosimetry.