Loading Events

« All Events

  • This event has passed.
Event Series: Colloquium Series

“Multimodal AI for Biomedical Discovery: Encoding, Fusion, and Simulation”

November 19 @ 2:00 pm – 3:00 pm

College of Arts & Sciences AI Umbrella Colloquium

Presented by Tianlong Chen, Ph.D.
Assistant Professor, Department of Computer Science
University of North Carolina at Chapel Hill

Abstract: Clinical decisions draw on many data streams at once: waveforms, electronic health records, images, notes, and omics. Most medical AI models still read a single modality and assume it is complete and clean.

This talk presents a path from multimodal representation to world-model simulation, in three parts. First, I explain why multimodal modeling is necessary in medicine, then focus on physiological waveforms, the densest and most time-critical signals in the hospital. I describe how to encode waveforms that are irregularly sampled, artifact-laden, and loosely aligned with the rest of the patient record.

Second, I address multimodal fusion under realistic conditions, where almost every patient is missing some modalities and the ones present differ widely in quality. I present mixture-of-experts architectures that accept arbitrary modality combinations and expose the interactions behind each prediction, along with adaptive token re-balancing that weights each modality by its contribution.

Finally, I discuss how to turn a multimodal foundation model into a multimodal world model, one that predicts how a patient’s state evolves under intervention, as a step toward virtual clinical trial simulation.

Location: Petty Building, Room 313
Organizers: The UNCG Computer Science Department and
The UNCG College of Arts & Sciences AI Umbrella

Tianlong Chen, Ph.D.

Tianlong Chen is an Assistant Professor and Chief Scientist at Foci Labs. Previously, he was a postdoctoral researcher at MIT and Harvard (2024). He received his Ph.D. in ECE at UT Austin (2023), a B.S. in Applied Mathematics, and a B.Eng. in Computer Science from the University of Science and Technology of China (2017). His primary research focuses are multimodal learning for biomedical science, large language model agents, and responsible AI. He is the recipient of the Amazon Research Award, Cisco Faculty Award, Meta Research Award, and several best/outstanding paper awards from various conferences and workshops like ACM SIGBio, NAACL, LoG, AMIA, ACM MM, AAAI, NeurIPS, etc. His research is gratefully supported by NIH, NSF, DOE, as well as industry and university grants.

336.256.1112

317 College Avenue
Greensboro, North Carolina 27412 United States
+ Google Map