Manual labeling limits the scale, accuracy, and timeliness of patient outcomes research in radiation oncology. We present RadOnc-GPT, an autonomous large language model (LLM)-based agent capable of independently retrieving patient-specific information, iteratively assessing evidence, and returning structured outcomes. Our evaluation explicitly validates RadOnc-GPT across two clearly defined tiers of increasing complexity: (1) a structured quality assurance (QA) tier, assessing the accurate retrieval of demographic and radiotherapy treatment plan details, followed by (2) a complex clinical outcomes labeling tier involving determination of mandibular osteoradionecrosis (ORN) in head-and-neck cancer patients and detection of cancer recurrence in independent prostate and head-and-neck cancer cohorts requiring combined interpretation of structured and unstructured patient data. The QA tier establishes foundational trust in structured-data retrieval, a critical prerequisite for successful complex clinical outcome labeling.
@article{arxiv.2509.25540,
title = {RadOnc-GPT: An Autonomous LLM Agent for Real-Time Patient Outcomes Labeling at Scale},
author = {Jason Holmes and Yuexing Hao and Mariana Borras-Osorio and Federico Mastroleo and Santiago Romero Brufau and Valentina Carducci and Katie M Van Abel and David M Routman and Andrew Y. K. Foong and Liv M Muller and Satomi Shiraishi and Daniel K Ebner and Daniel J Ma and Sameer R Keole and Samir H Patel and Mirek Fatyga and Martin Bues and Brad J Stish and Yolanda I Garces and Michelle A Neben Wittich and Robert L Foote and Sujay A Vora and Nadia N Laack and Mark R Waddle and Wei Liu},
journal= {arXiv preprint arXiv:2509.25540},
year = {2025}
}