Sylvia K. Plevritis, PhD

William M. Hume Professor and Chair of Biomedical Data Science, Stanford University

Multiscale Modeling of Lung Cancer: Toward Connecting Population Science and Tumor Biology through Multimodal AI

Progress against lung cancer depends on connecting what we learn from populations with an understanding of the tumor biology that shapes each patient’s disease. This presentation explores how multiscale modeling can bridge population, patient, tissue, and cellular perspectives to reveal opportunities for earlier detection and more effective treatment. Models of lung cancer natural history identify potential windows for intervention across the continuum of care. Complementary studies connect medical imaging with molecular features and reveal how tumor plasticity and the spatial organization of the tumor microenvironment relate to progression, treatment resistance, and survival. Building on these approaches, we are developing multimodal AI to integrate longitudinal clinical records with imaging, pathology, and molecular data. Together, multiscale modeling and multimodal AI offer a path toward linking population evidence with individual tumor biology to uncover cancer vulnerabilities and inform when and how to intervene to improve patient outcomes.