Approved Research

Artificial intelligence for deep phenotyping and target discovery in cardiovascular disease including heart failure

Lay summary

Cardiovascular disease (CVD) is a broad group of diseases affecting the heart and blood vessels. Heart failure (HF) is a disease where the heart is unable to carry out its normal function of pumping blood around the body. HF is a long-term condition that often gets worse, and although the symptoms can be managed, there is currently no cure. There are many different causes of CVD, and the severity of the disease and its progression varies across patients. There can be differences in how CVD patients are treated, and how patients respond to treatment. Currently in HF, a measure of the heart's ability to pump blood, called left ventricular ejection fraction (LVEF), is used to assign HF patients into groups. However, we know that there are still high levels of variation in patient groups based on LVEF. 

We aim to better understand the variation across patients in CVD, including HF. This will allow us to improve the management and treatment of CVD patients, and to develop new drugs for specific groups of CVD patients. We will use clinical records from CVD patients, combined with genetic information, medical images of the heart and information about physical activity, to model CVD and its progression to other diseases. We will use advanced methods such as artificial intelligence and computational simulations, to understand the risk factors and mechanisms of CVD. We aim to improve our understanding of why some patients progress rapidly and develop other diseases, and why some patients may not respond to treatment. We hope to identify new sub-groups of CVD patients that are clinically and biologically similar to each other, and clinical markers of these groups. Using these improved CVD sub-groups, we aim to discover new drug targets for CVD that may treat specific causes of the disease.