Reference-Free large language model agents for physician-guided radiotherapy treatment planning

Area of Science:

• Medical Physics
• Artificial Intelligence
• Radiation Oncology

Background:

• Large language models (LLMs) show promise for diverse tasks but face challenges in specialized domains like radiotherapy.
• Adapting LLMs for clinical radiotherapy treatment planning requires domain-specific knowledge integration.

Purpose of the Study:

• To adapt a general large language model (LLM) into a clinically viable agent for intensity-modulated radiation therapy (IMRT) inverse treatment planning.
• To investigate the architectural and functional components needed for LLM-driven radiotherapy planning.

Main Methods:

• A large language model (LLM) agent was developed to interact with a treatment planning system (TPS) for head-and-neck cancer patients.
• The LLM agent iteratively proposed new constraint values, informed by real-time plan evaluations, in a reference-free setting.
• LLM-generated plans were compared to clinically approved plans using dosimetric endpoints and statistical analysis (Wilcoxon signed-rank test).

Main Results:

• LLM-generated plans demonstrated comparable organ-at-risk (OAR) sparing to clinical plans.
• Improved hot spot control (Dmax: 106.5% vs. 108.8%, p < 0.05) and superior conformity (CI: 1.18 vs. 1.39, p < 0.05) were observed for the boost PTV.
• The LLM agent operated effectively without prior exposure to manual treatment plans or fine-tuning.

Conclusions:

• A reference-free, LLM-driven workflow is feasible for automated IMRT treatment planning within a commercial TPS.
• This approach offers a generalizable solution to reduce planning variability and promote AI adoption in radiotherapy.
• LLMs can be adapted for clinical radiotherapy planning, enhancing efficiency and consistency.