Correlation-Informed Residual Learning for Pretherapy Prediction of Voxel-Wise Time-Integrated Activity Maps in [ 177 Lu]Lu-PSMA-617 Radioligand Therapy (EANM 2026) - Sep 29, 2026 - Abstract #OP-686; Pres time: Oct 20, 2026; 03:00 PM - 04:30 PM; Location: Hall B; "The correlation-informed residual DL framework improves pretherapy TIA prediction accuracy in clinically relevant normal organs compared with both standalone linear regression and endto-end DL. By anchoring network predictions to organ-level linear baselines, the approach stabilizes training on limited datasets and generates voxel-level TIA maps with potential utility for pre-treatment dosimetry planning in PSMA-targeted RLT. Further multicenter validation and integration of lesion-level dosimetry are ongoing."
Castration-Resistant Prostate Cancer • Genito-urinary Cancer • Prostate Cancer • Solid Tumor
|
|
Y. Chen; J. Hu; R. Seifert; A. Rominger; A. Afshar-Oromieh; K. Shi
|