The Randles Lab develops high performance computational platforms and algorithms to advance biomedical research, uncover new insights into complex biological systems, and improve human health. Our work centers on multiscale, image-based digital twins of the human vasculature that combine physics-based modeling, AI, and clinical data to deliver accurate, personalized predictions of disease progression and treatment outcomes.
We are extending both the spatial and temporal reach of high-fidelity fluid–structure–interaction simulations through innovations such as the Adaptive Physics Refinement (APR) algorithm, which couples subcellular-scale modeling to organ-level blood flow, and the Longitudinal Hemodynamic Mapping (LHM) algorithm, which enables simulations spanning millions of heartbeats. These methods are powered by HARVEY, our massively parallel blood flow solver, and run efficiently on some of the world’s largest supercomputers and scalable cloud infrastructure.
Applications range from guiding coronary interventions, monitoring carotid plaque, and predicting heart failure risk to modeling rare cancer cell transport in realistic blood environments. We work closely with physicians and experimental collaborators in a continuous feedback loop—using clinical and in vitro measurements to refine models, and using simulations to generate actionable insights for care and discovery.
We are committed to fostering a collaborative, inclusive environment where researchers from diverse backgrounds can tackle challenging problems at the interface of computation and medicine. Through innovative algorithms, scalable platforms, and close clinical integration, our goal is to educate the next generation of biomedical engineers and make a lasting impact on patient care.
Lab photo: William Ladd, Emily Rakestraw, Amanda Randles, Samreen Mahmud, Cyrus Tanade, Junyu Nan, Daniel Puleri, David Urick, Aristotle Martin, Sayan Roychowdhury, Christopher Jensen, Ayman Yousef, Simba Chidyagwai, Saquan Anthony
*Hero image created by Joseph Insley, Argonne National Laboratory.
