Q&A: Duke’s Amanda Randles, Ph.D., on the Future of Digital Twin Innovation

Why It Matters

By delivering high‑resolution, non‑invasive hemodynamic insights, HARVEY can accelerate diagnosis, personalize treatment, and reduce reliance on invasive procedures, reshaping cardiovascular care and broader disease modeling.

Key Takeaways

• HARVEY simulates patient-specific blood flow across entire vasculature
• Cloud‑based engine reduces supercomputer runtime from hours to minutes
• Integration with wearables enables continuous, non‑invasive cardiovascular monitoring
• AI extracts early biomarkers from petabyte‑scale simulation data
• Expanded applications include heart failure, aneurysms, and cancer mechanics

Pulse Analysis

The concept of a digital twin— a virtual replica that mirrors a patient’s physiology— is moving from research labs to clinical decision‑making. At Duke, Dr. Amanda Randles’ team has spent more than a decade refining HARVEY, a physics‑driven engine that models blood flow from the aorta down to the microcirculation. Early versions required the world’s largest supercomputer and six hours to simulate a single heartbeat. Advances in high‑performance cloud computing have compressed that runtime to minutes, making patient‑specific simulations feasible for routine use and opening the door to longitudinal studies.

Beyond raw computation, HARVEY now ingests data from wearable sensors, creating a continuous digital twin that reflects daily physiological fluctuations. By feeding real‑time heart rate, blood pressure, and activity metrics into the fluid‑dynamic model, researchers can predict hemodynamic changes before symptoms appear. Artificial intelligence algorithms sift through the resulting petabytes of simulation output to surface early biomarkers of heart failure or ischemia, dramatically shortening the path from data to actionable insight. Clinicians also benefit from augmented‑reality visualizations that translate complex flow fields into intuitive 3‑D displays, accelerating adoption at the bedside.

The versatility of HARVEY is prompting collaborations that stretch far beyond cardiology. Duke’s team has already modeled cerebral artery dynamics to assess aneurysm rupture risk and is simulating deformable cancer cells navigating the vasculature to uncover novel therapeutic targets. Such cross‑disciplinary projects attract biotech partners, federal grants, and FDA‑cleared pathways, signaling a maturing market for digital‑twin diagnostics. As wearable adoption reaches billions and AI tools become more efficient, the industry is poised for a rapid scale‑up, turning what was once a six‑hour supercomputing exercise into a routine, cloud‑based service that could redefine preventive medicine.