Creating Non-Invasive, Accurate, Personalized Models
Our cardiovascular modeling program focuses on developing predictive tools that deliver clinically relevant insights without the need for invasive procedures.
HARVEY: FFR and iFR Prediction
We combine high-fidelity vascular simulations with machine learning to estimate fractional flow reserve (FFR) and instantaneous wave-free ratio (iFR) without invasive procedures. By reconstructing detailed 3D vascular models from standard 2D angiograms and capturing complex flow dynamics, our HARVEY digital twin platform provides richer physiological insight than traditional diagnostics. In collaboration with clinicians, we have shown that HARVEY’s predictions can closely match invasive FFR measurements, helping identify functionally significant lesions earlier, while reducing patient risk, procedure time, and unnecessary interventions.
Longitudinal Hemodynamic Mapping (LHM)
The Randles Lab has developed Longitudinal Hemodynamic Mapping (LHM), a first-of-its-kind framework for simulating millions of heartbeats to track cardiovascular function over extended periods—from weeks to months. Building on our HARVEY software for patient-specific blood flow modeling, LHM uses parallel simulation of Hemodynamic Units (HUs) to achieve unprecedented efficiency and scale. The framework seamlessly integrates continuous physiological data from wearable biosensors, allowing digital twin models to reflect a patient’s evolving state across real-world activities and physiological conditions. This approach enables earlier detection of subtle changes in cardiovascular health, supports more personalized treatment strategies, and moves cardiac care from reactive to truly proactive monitoring.
HarVI - Virtual Intervention Planning
HarVI is a machine learning–powered planning tool that gives cardiologists a patient-specific, interactive 3D environment for exploring coronary intervention strategies. By combining extended reality for intuitive visualization with real-time hemodynamic analysis, HarVI delivers fractional flow reserve (FFR) predictions in minutes instead of days. This rapid, accurate feedback enables clinicians to virtually test treatment options—such as stent placement—before entering the operating room, reducing risk and supporting more personalized, data-driven care.