Virtual contrast-enhanced MRI matched gadolinium-enhanced imaging for diagnosis, staging, and tumor delineation.
98.98% of virtual contrast-enhanced MRI scans were rated suitable for diagnosis, and 92.33% for tumor delineation.
Tag
AI
Neural model compresses proton Monte Carlo phase-space data while preserving dose accuracy
The neural model reduced phase-space storage from 3 gigabytes to 600 kilobytes while maintaining at least 89.9% gamma passing at 1%/1 mm.
Large language model framework extracted cardiac events with up to 85.5% accuracy
The framework achieved 85.5% external-validation accuracy and reduced estimated review time from 822 hours to 2.3–4.8 hours.
Body composition adds little to radiotherapy toxicity prediction
Computed tomography-derived muscle measures ranked highly for toxicity and quality of life, but added little predictive value beyond established clinical factors.
Hybrid foundation model produces expert-preferred brain metastasis contours
A hybrid foundation-model framework achieved 94% lesion-wise sensitivity and was preferred over physician-generated brain metastasis contours in blinded, bias-adjusted comparisons.