Frontiers in digital health · 2026
Automated voxel-based prostate volumetry improves PSA density-based detection of clinically significant prostate cancer in the PSA gray zone: comparison with manual segmentation and the ellipsoid formula.
СтатьяНазвание и аннотация: Перевод пока не подготовлен
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.
Han Jing, Qiu Yi, Liu Houdong, Wen Jing
Аннотация
PURPOSE: To investigate whether voxel-based prostate volume can improve the diagnostic performance of prostate-specific antigen density (PSAD) for detecting clinically significant prostate cancer (csPCa) in patients within the prostate-specific antigen (PSA) gray zone (4-10 ng/mL), and to compare it with conventional formula-based estimation.
METHODS: In this retrospective study, 489 patients with PSA levels of 4-10 ng/mL who underwent prostate MRI were included. Prostate volumes were calculated using three methods ellipsoid formula (PVformula ), manual voxel-based segmentation (PVmask ), and automated voxel-based segmentation (PVauto_seg ) based on nnU-Net. Agreement between volumetric measurements was assessed using Bland-Altman analysis, and PSAD was calculated according to three prostate volume estimations. Diagnostic performance for the detection of csPCa was evaluated using receiver operating characteristic (ROC) analysis.
RESULTS: PVformula systematically overestimated prostate volume compared with PVmask , with mean volumes of 58.8 ± 28.1 mL and 54.5 ± 26.6 mL (bias 4.4 mL); whereas PVauto_seg (55.8 ± 20.8 mL) showed close agreement with PVmask and reduced bias of 1.4 mL. In ROC analysis, voxel-based PSADmask (AUC 0.762, 95% CI 0.717-0.808) outperformed the formula-based PSADformula (AUC 0.719, 95% CI 0.671-0.767; P = 0.01). PSADauto_seg calculated from PVauto_seg yielded an AUC similar (0.754, 95% CI 0.707-0.800) to that of with PSADmask , and also outperformed PSADformula (P = 0.02).
CONCLUSION: Voxel-based prostate volumetry provides more accurate and consistent volume estimation than the conventional ellipsoid formula, leading to improved PSAD performance for detecting csPCa in the PSA gray zone. Moreover, automated segmentation enables reliable and scalable PSAD calculation with diagnostic performance comparable to manual delineation, supporting its potential for clinical implementation in risk stratification of PCa.
METHODS: In this retrospective study, 489 patients with PSA levels of 4-10 ng/mL who underwent prostate MRI were included. Prostate volumes were calculated using three methods ellipsoid formula (PVformula ), manual voxel-based segmentation (PVmask ), and automated voxel-based segmentation (PVauto_seg ) based on nnU-Net. Agreement between volumetric measurements was assessed using Bland-Altman analysis, and PSAD was calculated according to three prostate volume estimations. Diagnostic performance for the detection of csPCa was evaluated using receiver operating characteristic (ROC) analysis.
RESULTS: PVformula systematically overestimated prostate volume compared with PVmask , with mean volumes of 58.8 ± 28.1 mL and 54.5 ± 26.6 mL (bias 4.4 mL); whereas PVauto_seg (55.8 ± 20.8 mL) showed close agreement with PVmask and reduced bias of 1.4 mL. In ROC analysis, voxel-based PSADmask (AUC 0.762, 95% CI 0.717-0.808) outperformed the formula-based PSADformula (AUC 0.719, 95% CI 0.671-0.767; P = 0.01). PSADauto_seg calculated from PVauto_seg yielded an AUC similar (0.754, 95% CI 0.707-0.800) to that of with PSADmask , and also outperformed PSADformula (P = 0.02).
CONCLUSION: Voxel-based prostate volumetry provides more accurate and consistent volume estimation than the conventional ellipsoid formula, leading to improved PSAD performance for detecting csPCa in the PSA gray zone. Moreover, automated segmentation enables reliable and scalable PSAD calculation with diagnostic performance comparable to manual delineation, supporting its potential for clinical implementation in risk stratification of PCa.