In 2021-22, over 90,000 stents were implanted in the UK, saving thousands of lives, and reducing heart damage and long-term disability. However, over 30% of stents are implanted in suboptimal positions, which can lead to their own complications such as acute artery blockage or re-narrowing. Intracoronary imaging can be used to visualise the inside of coronary arteries in real time and improve stent positioning, but it is only used in 25% of all procedures in the UK as interpreting images is complex and time consuming.

Intracoronary imaging offers unparalleled resolution, but it is also an invasive and expensive technology. The goal is therefore to get as much information out of the procedure as possible, as well as to optimise the size and position of stents. CHIMiRA mathematicians have aided the interpretability of images by using deep-learning AI techniques to reduce artifacts and identify healthy and unhealthy artery walls. Collaborating with clinicians, they have also used the images of the multiple coronary vessels around the heart to construct biomechanical models. These coupled fluid and solid mechanical simulations help identify areas of high stress and predict whether diseased areas are at high risk of rupture, and determine if any preventative treatment is needed.

To translate this cutting-edge interdisciplinary research into a user-friendly decision support tool for clinicians, in 2021 CHIMiRA mathematician, Michael Roberts, co-founded a spin-out company, Octiocor. The team’s long-term goal is that their algorithm can be run in clinics in real time, to inform diagnosis and personalised treatment, ultimately improving patient outcomes by reducing heart attacks and stent failure. The software is in beta testing in three expert laboratories around the world.