Multimodal Data Analysis in Healthcare: Opportunities and Challenges

Presentation Time: 01:45 PM - 03:15 PM

Abstract Keywords: Large Language Models (LLMs), Deep Learning, Data Mining
Primary Track: Applications
Programmatic Theme: Clinical Research Informatics
Moderator: Yifan Peng, PhD
Weill Cornell Medicine; Dept of Population Health Sciences; Div of Health Informatics

The recent advances in multimodal foundation models have made a significant shift in research and clinical practices. However, to fully realize the potential of multimodal data analysis, there are various scientific and social challenges that need to be addressed, such as how to ensure models’ trustworthiness and scalability, and how to maintain data quality and integration. The objective of this panel is to introduce the audience to the opportunities and challenges, as well as the development and responsible employment of such technology in research and healthcare. It will specifically focus on the development of multimodal foundation models in healthcare, issues of model transparency, accountability, and fairness, and multimodal data de-identification and sharing. After participating in this session, attendees should be able to understand the most important challenges facing multimodal data analysis and some of the possible solutions.

Speaker(s):
Hoifung Poon, PhD
Microsoft Research

Zhiyong Lu, PhD
National Library of Medicine, NIH

Kevin Johnson, MD, MS
University of Pennsylvania

Imon Banerjee, PhD
Arizona State U, Mayo Clinic