Healthcare Implications of Large Language Models like ChatGPT (2023)-WEB37-2023

Video Transcription

Video Summary

The transcript outlines a webinar featuring experts discussing the implications of large language models (LLMs) like ChatGPT in healthcare, particularly radiology. Hosted by Linda Moy from New York University, the session includes panelists Dr. Sam Biswas, Dr. Jonathan Elias, Dr. Keith Hentel, Felipe Kitamura, and Dr. George Hsi. Dr. Hsi provides an overview of ChatGPT and GPT-4, highlighting improvements such as the ability to handle complex inputs and incorporating multimodal capabilities. The discussion covers numerous potential applications for ChatGPT in healthcare, such as streamlining radiology report generation, with the possibility of enhancing clinical decision support by synthesizing information from patient charts. A key topic is the need for AI and large language models to maintain patient privacy, meeting compliance requirements like HIPAA. The discussion also touches on the importance of radiologists remaining involved directly with patient care to complement technology, which is unlikely to replace physicians entirely. Questions from attendees include the future role of AI in radiology education and the regulatory responsibilities towards these technologies, emphasizing the need to validate these tools scientifically while navigating ethical considerations.

Keywords

Large Language Models

ChatGPT

Healthcare

Radiology

Patient Privacy

AI Compliance

Radiology Education

Clinical Decision Support

Ethical Considerations