Learning the Language of Patients - Multimodal Generative AI for Precision Health

Explore the cutting-edge intersection of artificial intelligence and precision healthcare in this distinguished lecture from the Paul G. Allen School. Discover how multimodal generative AI is revolutionizing precision health by learning to understand and synthesize the complex language of patient data. Learn about the development of powerful multimodal patient embeddings that serve as digital twins, enabling researchers to harness real-world data from electronic health records, medical imaging, and multiomics to optimize care delivery and accelerate biomedical discovery. Examine groundbreaking research on large language models and multimodal AI applications in healthcare, including popular open-source foundation models like PubMedBERT, BioGPT, BiomedCLIP, LLaVA-Med, and BiomedParse. Understand how population-scale real-world evidence from millions of cancer patients is being applied to advance precision oncology through partnerships with major health systems and life sciences companies. Gain insights into the latest developments in digital pathology, including GigaPath, the first whole-slide foundation model pretrained on over one billion pathology image tiles, as featured in Nature. Learn about practical AI implementations currently in daily use for molecular tumor boards and clinical trial matching, and explore the future vision of continuous learning systems that instantly incorporate new health information to transform healthcare delivery and biomedical research.