LLM GatorTron-3.9B outperforms BERT in predicting heart failure risk for cancer patients

Researchers from the University of Florida Health have developed a study using large language models (LLMs) to predict heart failure risk in cancer patients. The study utilized electronic health records from over 12,000 patients, with a specific focus on those diagnosed with lung, breast, and colorectal cancers. The LLM, named GatorTron-3.9B, demonstrated superior performance by achieving higher F1 scores compared to traditional machine learning models and a BERT transformer model, indicating the potential of LLMs in improving patient outcomes and treatment safety. AI

IMPACT Demonstrates LLMs' potential in clinical decision support, improving patient safety in oncology.

RANK_REASON Academic paper detailing a study on LLM performance for a specific medical task. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• alphaXiv
• arXiv
• Bert
• CatalyzeX
• DagsHub
• GatorTron-3.9B
• Hugging Face
• time-aware long short-term memory
• University of Florida Health
• Ziyi Chen

• paper
• model release

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