Using machine-learning for personalised risk prediction of heart and circulatory disease

The BHF-Turing Cardiovascular Data Science Awards (First Call): Using machine learning for personalised CVD risk management (joint funding with The Alan Turing Institute)

Heart and circulatory disease causes more than a quarter of all deaths in the UK. In GP practices around the country, people over the age of 40 have their risk factors measured to estimate how likely they are to develop heart and circulatory disease over the next ten years. These risk factors include age, smoking, high blood pressure and high cholesterol. The health check aims to enable doctors to identify and help people that are greatest risk of heart disease, type 2 diabetes and stroke. Scientists want to improve the accuracy and efficiency of NHS health checks. These researchers, based at the University of Cambridge, believe that by using data more intelligently, the health checks could be improved. For instance, some of the measures taken in the health check may already exist in a person’s medical records. And the health check could be better at distinguishing the risk of different types of heart condition, so that a person can receive the most suitable treatment. To address these issues, a team of data scientists and health researchers will adapt cutting-edge statistical and machine learning methods, which use computers to learn from data. They will apply these techniques to existing sources of medical data to learn how to identify who is most at risk. If this new way of using health data can improve the accuracy and efficiency of NHS health checks, it could save both money and lives.