RadOnc-GPT: An Autonomous LLM Agent for Real-Time Patient Outcomes Labeling at Scale
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866914199291035648 |
|---|---|
| author | Holmes, Jason Hao, Yuexing Borras-Osorio, Mariana Mastroleo, Federico Brufau, Santiago Romero Carducci, Valentina Van Abel, Katie M Routman, David M Foong, Andrew Y. K. Muller, Liv M Shiraishi, Satomi Ebner, Daniel K Ma, Daniel J Keole, Sameer R Patel, Samir H Fatyga, Mirek Bues, Martin Stish, Brad J Garces, Yolanda I Wittich, Michelle A Neben Foote, Robert L Vora, Sujay A Laack, Nadia N Waddle, Mark R Liu, Wei |
| author_facet | Holmes, Jason Hao, Yuexing Borras-Osorio, Mariana Mastroleo, Federico Brufau, Santiago Romero Carducci, Valentina Van Abel, Katie M Routman, David M Foong, Andrew Y. K. Muller, Liv M Shiraishi, Satomi Ebner, Daniel K Ma, Daniel J Keole, Sameer R Patel, Samir H Fatyga, Mirek Bues, Martin Stish, Brad J Garces, Yolanda I Wittich, Michelle A Neben Foote, Robert L Vora, Sujay A Laack, Nadia N Waddle, Mark R Liu, Wei |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_25540 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | RadOnc-GPT: An Autonomous LLM Agent for Real-Time Patient Outcomes Labeling at Scale Holmes, Jason Hao, Yuexing Borras-Osorio, Mariana Mastroleo, Federico Brufau, Santiago Romero Carducci, Valentina Van Abel, Katie M Routman, David M Foong, Andrew Y. K. Muller, Liv M Shiraishi, Satomi Ebner, Daniel K Ma, Daniel J Keole, Sameer R Patel, Samir H Fatyga, Mirek Bues, Martin Stish, Brad J Garces, Yolanda I Wittich, Michelle A Neben Foote, Robert L Vora, Sujay A Laack, Nadia N Waddle, Mark R Liu, Wei Artificial Intelligence 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. |
| title | RadOnc-GPT: An Autonomous LLM Agent for Real-Time Patient Outcomes Labeling at Scale |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2509.25540 |