# Learning to Optimize Radiotherapy Plans via Fluence Maps Diffusion Model Generation and LSTM-based Optimization

arXiv: https://arxiv.org/abs/2605.13713 · HTML: https://mlsanity.com/paper/2605.13713/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Medical imaging and segmentation, one of 14 directions in Diffusion models.
- Builds on 16 of its 34 references that are in the map.
- First posted on arXiv 13 May 2026.

## Placed in

- Diffusion models / [Medical imaging and segmentation](https://mlsanity.com/diffusion/medical-imaging-and-segmentation/)

## Builds on (most cited first, 16 of 16)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [Analyzing and Improving the Image Quality of StyleGAN](https://mlsanity.com/paper/1912.04958/) (2019)
- [Elucidating the Design Space of Diffusion-Based Generative Models](https://mlsanity.com/paper/2206.00364/) (2022)
- [Improved Distribution Matching Distillation for Fast Image Synthesis](https://mlsanity.com/paper/2405.14867/) (2024)
- [Learning to Optimize](https://mlsanity.com/paper/1606.01885/) (2016)
- [MAISI: Medical AI for Synthetic Imaging](https://mlsanity.com/paper/2409.11169/) (2024)
- [Towards Constituting Mathematical Structures for Learning to Optimize](https://mlsanity.com/paper/2305.18577/) (2023)
- [μLO: Compute-Efficient Meta-Generalization of Learned Optimizers](https://mlsanity.com/paper/2406.00153/) (2024)
- [Multi-Agent Reinforcement Learning Meets Leaf Sequencing in Radiotherapy](https://mlsanity.com/paper/2406.01853/) (2024)
- [Revisiting Diffusion Models: From Generative Pre-training to One-Step Generation](https://mlsanity.com/paper/2506.09376/) (2025)
- [HyperAdam: A Learnable Task-Adaptive Adam for Network Training](https://mlsanity.com/paper/1811.08996/) (2018)
- [A Beam's Eye View to Fluence Maps 3D Network for Ultra Fast VMAT Radiotherapy Planning](https://mlsanity.com/paper/2502.03360/) (2025)
- [Demo: Generative AI helps Radiotherapy Planning with User Preference](https://mlsanity.com/paper/2512.08996/) (2025)
- [Radiotherapy Dosimetry: A Review on Open-Source Optimizer](https://mlsanity.com/paper/2305.18014/) (2023)
- [PyLO: Towards Accessible Learned Optimizers in PyTorch](https://mlsanity.com/paper/2506.10315/) (2025)
- [AI End-to-End Radiation Treatment Planning Under One Second](https://mlsanity.com/paper/2603.06338/) (2026)

## Abstract (arXiv)

Volumetric Modulated Arc Therapy (VMAT) is a cornerstone of modern radiation therapy, enabling highly conformal tumor irradiation and healthy-tissue sparing. Yet, its planning solves inverse and nested optimization for multi-leaf collimators, monitor units and dose parameters, while enforcing their consistency to ensure mechanical deliverability. Nevertheless, this process often requires repeated re-optimization when treatment configurations change, resulting in substantial planning time per patient. To address these problems, we present a diffusion-driven Learning-to-Optimize (L2O) method for end-to-end VMAT planning. A distribution-matching distilled diffusion model learns a clinically feasible manifold of fluence maps, enabling their one-shot generation. On top of this, an LSTM-based L2O module learns gradient update dynamics to swiftly refine fluence maps toward prescribed dose objectives during inference. Experimental results on clinical and public prostate cancer cohorts demonstrate improved planning efficiency, flexibility, and machine deliverability over currently available end-to-end VMAT planners.