End‐to‐end deep learning pipeline for automated IMRT planning in lymphoma: Feasibility and limitations
Denis Kutnár, Ivan Richter Vogelius, Hristo Atanasov Georgiev, Katrin Elisabet Håkansson, Lena Specht, Jens Petersen · 2026
Read End‐to‐end deep learning pipeline for automated IMRT planning in lymphoma: Feasibility and limitations by Denis Kutnár, Ivan Richter Vogelius, Hristo Atanasov Georgiev, Katrin Elisabet Håkansson, Lena Specht, Jens Petersen (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13438194.txt.
Description
Kutnár and colleagues test fully automated deep learning planning for lymphoma radiotherapy in this 2026 study. Intensity-modulated treatment shapes radiation around irregular tumors, but planning is slow and hard to standardize, and lymphoma is an especially varied test case across the body. The team replanned a single-institution cohort of 571 patients. The plans became a consistent fifteen-field reference. They then trained a two-stage pipeline. It predicts dose first, then the beam fluence that delivers it.
Variants trained to copy existing plans predicted dose less accurately than the references, while variants trained with a clinically shaped loss produced better target dose distributions. After the final conversion step, every configuration produced machine-deliverable plans, but all had lower target coverage and hotter spots than human-made plans. The authors conclude the approach is feasible and fast, yet currently inferior for this difficult setting.
Catalog record
| CREATOR | Denis Kutnár, Ivan Richter Vogelius, Hristo Atanasov Georgiev, Katrin Elisabet Håkansson, Lena Specht, Jens Petersen |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 41 KB |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |