Health Care Implications of Large Language Models like ChatGPT

Video Transcription

Video Summary

The Radiological Society of North America (RSNA) hosted a session on large language models (LLMs) moderated by Linda Moy from NYU, featuring a panel of experts discussing the advancements, applications, and challenges of LLMs, like ChatGPT, in radiology. Key points included the integration of LLMs into clinical workflows for tasks such as summarizing clinical notes, reducing documentation burdens, and enhancing decision support. Panelists explored the regulatory aspects, noting that all healthcare-related AI applications are considered high-risk, especially under European regulations. The responsibility of educating patients about potential inaccuracies from AI models was emphasized. The session also tackled the medical-legal risks and transparency required when implementing these tools. Experts suggested that while AI holds immense potential to transform radiology, ensuring accurate and effective patient communication remains critical. The session concluded by addressing concerns about AI's impact on radiology careers, with panelists expressing optimism about AI enhancing the field, creating more efficient and patient-centered workflows.

Keywords

Radiological Society of North America

large language models

ChatGPT

radiology

clinical workflows

AI regulation

patient communication

medical-legal risks

radiology careers