Topic: "AI for Healthcare: How Can Large Language Models Help Physicians at the Bedside?"
Presenter: Yanjun Gao, Ph.D.
Research Associate, University of Wisconsin
School of Medicine and Public Health
Abstract: The rise of large language models has revolutionized the field of natural language processing, reshaping the landscape of information processing and decision-making. LLMs have shown great promise in the realm of healthcare, for instance, synthesizing electronic health records (EHR) data, yet they also face unique hurdles due to the critical need for precise domain knowledge and the potential consequences of errors in medical applications.
This talk will explore both the potential advantages and challenges of leveraging LLMs within clinical applications. The first part of the talk will examine the progress of clinical NLP in the past decades. Following that, the talk will present a novel knowledge-intensive EHR summarization task designed to optimize clinicians’ diagnostic workflow. It will then assess the performance of various LLMs, including knowledge-graph prompted models, in tackling this critical challenge.
Friday, December 1 at 10:00am to 11:00amVirtual Event