AI End-to-End Radiation Treatment Planning Under One Second

2026-03-06

AI Summary

This study addresses the inefficiency and inter-planner variability inherent in conventional radiotherapy planning, which compromise clinical throughput and consistency. The authors propose AIRT, a novel framework that, for the first time, enables sub-second, end-to-end generation of deliverable single-arc VMAT plans for prostate cancer directly from CT images and structure contours by predicting multileaf collimator sequences. Integrating differentiable dose calculation, adversarial fluence map shaping, and plan augmentation strategies—accelerated via GPU inference—the model was trained on over 10,000 clinical cases. The generated plans demonstrate non-inferiority to RapidPlan Eclipse in target coverage and organ-at-risk sparing, achieving a target homogeneity index (HI) of 0.10 ± 0.01, thereby significantly enhancing plan quality, robustness, and computational efficiency.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Machine Learning: Efficient ML / Green AI

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Artificial intelligence-based radiation therapy (RT) planning has the potential to reduce planning time and inter-planner variability, improving efficiency and consistency in clinical workflows. Most existing automated approaches rely on multiple dose evaluations and corrections, resulting in plan generation times of several minutes. We introduce AIRT (Artificial Intelligence-based Radiotherapy), an end-to-end deep-learning framework that directly infers deliverable treatment plans from CT images and structure contours. AIRT generates single-arc VMAT prostate plans, from imaging and anatomical inputs to leaf sequencing, in under one second on a single Nvidia A100 GPU. The framework includes a differentiable dose feedback, an adversarial fluence map shaping, and a plan generation augmentation to improve plan quality and robustness. The model was trained on more than 10,000 intact prostate cases. Non-inferiority to RapidPlan Eclipse was demonstrated across target coverage and OAR sparing metrics. Target homogeneity (HI = 0.10 $\pm$ 0.01) and OAR sparing were similar to reference plans when evaluated using AcurosXB. These results represent a significant step toward ultra-fast standardized RT planning and a streamlined clinical workflow.

Problem

• radiation therapy planning
• AI end-to-end
• treatment planning time
• inter-planner variability
• ultra-fast planning

Innovation

• end-to-end deep learning
• differentiable dose feedback
• adversarial fluence shaping
• ultra-fast radiotherapy planning
• VMAT automation