AI model GatorOnco achieves expert-level performance in colorectal cancer treatment planning

Researchers have developed GatorOnco, an agentic large language model designed for treatment planning in colorectal cancer. Trained on a massive dataset of 282 billion tokens, including extensive clinical text from UF Health, GatorOnco utilizes a domain-adaptation method and an agentic retrieval-augmented generation approach to integrate up-to-date clinical guidelines. In a clinical evaluation, GatorOnco demonstrated expert-level performance comparable to oncologists, receiving higher ratings for readability and completeness, while matching expert performance in correctness, currency, and safety. AI

IMPACT Demonstrates potential for LLMs to assist in complex, high-stakes medical decision-making, improving efficiency and accuracy.

RANK_REASON Publication of a research paper detailing a new LLM for a specific medical application. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• alphaXiv
• arXiv
• CatalyzeX
• colorectal cancer
• DagsHub
• GatorOnco
• Gotit.pub
• Hugging Face
• retrieval-augmented generation
• ScienceCast
• University of Florida Health

• paper
• model release
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →