Large language models in medicine

Area of Science:

• Artificial Intelligence in Medicine
• Natural Language Processing
• Clinical Informatics

Background:

• Large language models (LLMs) demonstrate capacity for unprompted query response, sparking interest and apprehension regarding their healthcare integration.
• Generative artificial intelligence (AI) chatbots, like ChatGPT, are advanced LLM applications increasingly relevant to medical fields.

Purpose of the Study:

• To provide a foundational understanding of LLM development and application in clinical settings.
• To critically evaluate the strengths, limitations, and potential impact of LLMs on medical practice, education, and research.
• To serve as a primer for clinicians navigating the evolving landscape of AI in healthcare.

Main Methods:

• Review of LLM development processes, focusing on generative AI chatbots such as ChatGPT.
• Analysis of current deployments and applications of LLM chatbots in biomedical contexts.
• Discussion of the potential benefits and drawbacks of LLM integration into healthcare workflows.

Main Results:

• LLM applications, including ChatGPT, are being actively explored and deployed across various biomedical domains.
• Initial results from LLM chatbot deployment in healthcare show promise but are mixed, highlighting the need for further evaluation.
• LLMs possess the potential to enhance efficiency and effectiveness in clinical, educational, and research activities.

Conclusions:

• LLM technology presents significant opportunities for advancing medical efficiency and effectiveness.
• Careful consideration of LLM strengths and limitations is crucial for successful integration into healthcare.
• Clinicians play a pivotal role in determining the appropriate and beneficial use of LLM technology in patient care and medical practice.