Leveraging GPT-4 for identifying cancer phenotypes in electronic health records: a performance comparison between GPT-4, GPT-3.5-turbo, Flan-T5, Llama-3-8B, and spaCy's rule-based and machine learning-based methods.

Autor: Bhattarai, Kriti, Oh, Inez Y, Sierra, Jonathan Moran, Tang, Jonathan, Payne, Philip R O, Abrams, Zach, Lai, Albert M
Zdroj: JAMIA Open; Oct2024, Vol. 7 Issue 3, p1-9, 9p
Databáze: Complementary Index