Creating Non Invasive, Accurate, Personalized Models
Our cardiovascular research develops patient specific digital twins that combine medical imaging, physics based simulation, AI, and clinical data to provide accurate, non invasive assessments of cardiovascular disease. Our goal is to give clinicians richer physiological information while reducing the need for invasive procedures.
HARVEY: Non Invasive FFR and iFR Prediction
HARVEY combines high fidelity blood flow simulation with AI to estimate fractional flow reserve (FFR) and instantaneous wave free ratio (iFR) without the need for invasive pressure wires. Starting from standard 2D coronary angiograms, HARVEY reconstructs patient specific 3D coronary artery models and simulates blood flow to evaluate whether a narrowing is limiting blood supply to the heart.
Working closely with clinical collaborators, we have shown that HARVEY can closely match invasive FFR measurements while providing a more complete picture of patient specific coronary physiology. By reducing the need for invasive testing, HARVEY has the potential to lower patient risk, shorten procedures, and improve treatment planning.
Longitudinal Hemodynamic Mapping (LHM)
Longitudinal Hemodynamic Mapping (LHM) extends cardiovascular digital twins from a single snapshot to a continuously evolving representation of patient physiology. Built on HARVEY, LHM enables simulations spanning millions of heartbeats, allowing us to study how cardiovascular function changes over days, weeks, and months.
By integrating wearable sensor data, physics based modeling, and AI, LHM continuously updates a patient's digital twin to reflect changes in physiology during everyday life. This approach creates the foundation for continuous cardiovascular monitoring, enabling earlier detection of disease progression, more personalized treatment strategies, and a shift from reactive care to proactive disease management.
HarVI: Virtual Intervention Planning
HarVI is an AI powered virtual intervention planning platform that allows cardiologists to explore patient specific treatment strategies in an interactive 3D digital twin before entering the catheterization laboratory. By combining extended reality, machine learning, and physics based blood flow simulation, HarVI provides rapid fractional flow reserve (FFR) predictions in minutes rather than days.
Clinicians can virtually evaluate treatment options, including stent placement, and immediately assess their physiological impact on coronary blood flow. This real time feedback supports more informed decision making, reduces procedural uncertainty, and enables more personalized cardiovascular care.