Standardized two-hour fasting produced reproducible stomach volumes in only 42% of patients receiving magnetic resonance-guided pancreatic stereotactic radiotherapy.
A macro Monte Carlo framework reproduced reference dose calculations for very high energy electron beams up to 250 megaelectronvolts while improving computational efficiency by as much as twenty-sevenfold.
A hybrid foundation-model framework achieved 94% lesion-wise sensitivity and was preferred over physician-generated brain metastasis contours in blinded, bias-adjusted comparisons.