The Evolving Landscape of Generative Large Language Models and Traditional Natural Language Processing in Medicine

TL;DR

This paper compares traditional NLP and generative large language models in medicine, highlighting their respective strengths and emphasizing the importance of ethical considerations as these technologies evolve.

Contribution

It provides a comprehensive analysis of 19,123 studies, revealing the distinct roles of LLMs and traditional NLP in various medical tasks and discussing ethical implications.

Findings

LLMs excel in open-ended medical tasks

Traditional NLP is more effective for information extraction

Ethical use is crucial for medical AI applications

Abstract

Natural language processing (NLP) has been traditionally applied to medicine, and generative large language models (LLMs) have become prominent recently. However, the differences between them across different medical tasks remain underexplored. We analyzed 19,123 studies, finding that generative LLMs demonstrate advantages in open-ended tasks, while traditional NLP dominates in information extraction and analysis tasks. As these technologies advance, ethical use of them is essential to ensure their potential in medical applications.