Digital Twin of the Aorta

Scientists at Queen Mary University of London have created TAVR-AID, a digital twin of the aorta specific to each patient, which simulates blood flow and makes predictions using artificial intelligence.

Replacing a heart valve is fraught with risk. The valve you use and how you position it directly affects blood flow and pressure – which in turn determines how efficient your heart will be. Get it wrong, and the changed blood flow can degrade the structural integrity of the implanted valve.

Clinicians treating stenosis must weigh up numerous factors including valve selection; positioning based on the size, shape, and angle of your aorta where it meets the heart; and orientation to optimise blood flow and pressure to ensure long-term valve durability.

TAVR-AID is not designed to replace human decision-making but to enhance it.  CT scans are used to build a digital twin of the patient’s aorta, and echocardiography to determine the severity of the stenosis being treated. This allows an individual patient’s blood flow to be simulated.

The AI model will use this simulation and the patient’s health records to predict which valve to use, what size it should be, how to position it, and the risk of complications such as stroke, leaks, or device failure.

To achieve this, Queen Mary assembled an inter-disciplinary team of interventional cardiologists, surgeons, engineers, research nurses, computer scientists, and a patient group – with the work led jointly by Professor Anthony Mathur, Prof Greg Slabaugh, and Dr Caroline Roney. The blend of clinical insight and technological skills is what makes the models sophisticated and state-of-the-art but also grounded in practical, real-world clinical requirements.

We are looking for a commercial partner to help bring the technology to clinicians.