Back

RADIANT: A fully configurable radiotherapy dose prediction framework

Netherton, T. J.; Tchuindjang, J.; Celaya, A.; Gay, S.; Fuentes, D.; Court, L. E.

2025-06-22 oncology
10.1101/2025.06.20.25330029 medRxiv
Show abstract

In this work, we present the Radiotherapy Dose Inference and Analysis Toolkit (RADIANT), an open-source, fully configurable framework for 3D radiotherapy dose prediction. Built upon the Medical Imaging Segmentation Toolkit, RADIANT supports a wide range of network architectures, loss functions, and training strategies. We demonstrate its capabilities using a cervical cancer dataset generated with the Radiation Planning Assistant, consisting of 158 treatment plans. Dose prediction models were trained using five architectures - nnUNet, FMG-Net, W-Net, DDUNet, and Swin UNETR - and evaluated across clinical metrics such as dose score, homogeneity index, and percent errors in D95, D98, and D99. The best-performing model (MAE loss, nnUNet, polynomial scheduler) achieved the lowest average errors across all key dose metrics while also generalizing well to an independent test set. Our results indicate that RADIANT provides a scalable foundation for the rapid development and benchmarking of dose prediction models.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.