Invited Talks

Daniel McDuff,

Google, UW

Title: Foundation Models & Agentic AI that Supports Healthy Living

Abstract: A new era of wearable foundation models that capture rich information about behavior and physiology present the opportunity to discover novel biomarkers of disease and forecast health states into the future. These models, combined with Personal Health Agents, built using agentic AI, will provide new opportunities for people to interface with these data, learn about them in a personalized way, and make healthy choices. In this talk I will present several of our latest projects that bridge these topics.

Speaker Bio: Daniel McDuff is a Staff Research Scientist and Manager at Google and Affiliate Professor at the University of Washington. Daniel completed his PhD at the MIT Media Lab in 2014 and has a B.A. and Masters from Cambridge University. Previously, Daniel worked at the UK MoD, was Director of Research at MIT Media Lab spin-out Affectiva. His work has received nominations and awards from Popular Science magazine as one of the top inventions in 2011, South-by-South-West Interactive (SXSWi), The Webby Awards, ESOMAR and the Center for Integrated Medicine and Innovative Technology (CIMIT). His projects have been reported in many publications including The Times, the New York Times, The Wall Street Journal, BBC News, New Scientist, Scientific American and Forbes magazine. Daniel was named a 2015 WIRED Innovation Fellow, an ACM Future of Computing Academy member and has spoken at TEDx and SXSW. Daniel has published over 150 peer-reviewed papers on health AI (Nature, Nature Medicine), multimodal machine learning (NeurIPS, ICLR, ICCV, ECCV, ACM TOG), human-computer interaction (CHI, CSCW, IUI) and biomedical engineering (TBME, EMBC).

Anna C. Schaar,

Bioptimus

Title: Forming cellular niches - towards foundation models for single-cell and spatial omics

Abstract: Cells form complex microenvironments and understanding their spatial dependencies in health and disease paves our way of identifying novel cellular mechanisms. To guide and accelerate the tidysome study of spatial niches, we built Nicheformer, a transformer-based foundation model trained on over 110 million human and mouse cells from both dissociated and spatial transcriptomics data. Nicheformer learns spatially informed cell representations and outperforms models trained only on dissociated data in tasks designed to capture spatial dependencies in cells without the need of measuring them with cost- and time-intensive assays. This sets a new foundation for scalable, multiscale spatial single-cell analysis.

Speaker Bio: Anna Schaar is a Senior Research Scientist at Bioptimus working on multimodal and multiscale foundation models in oncology and other diseases. At Bioptimus, Anna aims together with her colleagues to unlock the first universal AI foundation models to understand biology and accelerate innovations in biomedicine. Prior to her work at Bioptimus, she worked with Fabian Theis at Helmholtz Munich on machine learning and deep learning methods designed to understand cellular dependencies in spatial omics technologies which capture molecular features at their physical location. Additionally, she built the Single Cell Best Practices Book across modalities and is a community manager at the scverse consortium which develops foundational tools for omics data in life sciences.

Mingu Kang,

Lunit

Title: Foundation Model Development: How we built and Where we go next

Abstract: Building a foundation model in computational pathology is profoundly impactful and challenging. In this talk, I will share our journey at Lunit in creating foundation models for computational pathology—why they are needed, how we built them, and the lessons we learned along the way. I will also showcase how these models are now driving real-world impact through our products, bridging the gap between research and clinical application. Finally, I will offer a glimpse into what lies ahead as we continue to shape the future of AI in this field.

Speaker Bio: Mingu Kang is a Senior Research Scientist at Lunit, where he is part of the AI Foundation team developing large-scale foundation models to advance computational pathology. His research focuses on building generalizable AI systems that bridge the gap between academic innovation and real-world clinical impact. At Lunit, Mingu has contributed to several major AI product developments and has led the creation of multiple foundation models, many of which have been shared and published externally. He received his M.S. in Computer Science from KAIST in 2019.

Faisal Mahmood,

Harvard Uni.

Title: TBA

Abstract: TBA

Speaker Bio: TBA

Imon Banerjee,

Mayo Clinic

Title: Bias in Multimodal foundation model: causes, evaluation and mitigation

Abstract: As multimodal foundation models become central to healthcare AI, their hidden biases threaten to widen existing health disparities. This talk will explore how bias arises in data and model design, how it can be measured rigorously, and what strategies can promote fairness and trust in multimodal learning for cancer detection and risk estimation.

Speaker Bio: Imon Banerjee is an Associate Professor and Director of the AI Innovation Hub at Mayo Clinic Arizona. She serves as the AI Course Director at the Mayo Clinic Alix School of Medicine and is also a graduate faculty member at Arizona State University. Her research focuses on fairness in artificial intelligence, multimodal deep learning, and the clinical translation of AI for digital healthcare. With a strong commitment to ethical AI, Dr. Banerjee works at the intersection of adversarial, causal, and contrastive learning methods to develop models that are not only accurate but also equitable. Her work addresses critical issues of bias related to race, gender, and other demographic factors, aiming to build AI systems that are both innovative and socially responsible—particularly within healthcare. Dr. Banerjee has authored over 200 scientific publications and currently leads several multi-institutional research projects funded by both federal and non-federal agencies, all aimed at advancing AI-driven digital healthcare solutions. She completed her postdoctoral training at Stanford University. Prior to joining Mayo Clinic, she was an Assistant Professor at the Emory University School of Medicine. She earned her Ph.D. in Computer Science from the University of Genova, Italy, and was a Marie Curie Fellow at the Center for National Research. She also completed her master's thesis at CERN, Switzerland.

Report abuse