Nicholas Mills is the BHF Professor of Cardiology at the Centre for Cardiovascular Science, University of Edinburgh.
Professor Mills and his team are at the forefront of cardiovascular diagnostics and healthcare data research - harnessing the power of digital healthcare data to learn about heart disease and its consequences for our health.
The power of data
Much of what is known about heart disease has been learnt through careful observation over centuries. Now, the wealth of information made available through the digitisation of healthcare records allows us to make these observations more reliably and faster.
Using this data, Professor Mills and his team will provide new insights into the tests we use to diagnose heart conditions. Improving the diagnosis of heart disease can lead to better outcomes as it ensures individual patients receive the most effective treatment for them.
Streamlining diagnosis
Diagnostic tests may account for only a small part of the healthcare budget, but their results go on to influence most treatment decisions. Given their power over treatment options, it is essential that tests are constantly evaluated so that they can be improved.
Professor Mills’ team are working to evaluate and improve diagnostic tests for patients with symptoms such as chest pain or breathlessness by using routinely collected healthcare data.
One way to improve these tests is to combine information from several sources. This could include looking at images of the heart as well as looking for blood ‘biomarkers’ which can give an early indication of potential issues. Coupled with risk factors, this can help in the approach to screening for coronary heart disease and identify the best care to prevent future problems.
Artificial intelligence
Looking for patterns across many different kinds of data is difficult for the human brain however, but artificial intelligence has the potential to transform the way we practice medicine. Its ability to spot patterns in massive amounts of data means that we can improve diagnosis and risk prediction.
Already, Professor Mills’ team have explored a range of approaches to improve the diagnosis of heart attacks and heart failure using artificial intelligence, resulting in clinical decision support tools based on machine learning, which have been taught to look for particular patterns. Within 3 hours, these tools can accurately predict the likelihood that a patient has had a heart attack based on an individual's age, sex and blood test results alone.
Clinical trials
Data is also being used to support the conduct of clinical trials. Professor Mills has established multiple trials of diagnostic tests, which benefit from identifying patients using real-time data as they are examined and tested in hospital. This means trials can be set up faster and the findings are more representative.
Taking this further, Professor Mills intends to use app-based recruitment strategies to reach potential patients. Participants can then record trial outcomes on the app, which can then be measured by researchers. This will not only help in selecting the right patients for the right trials, but will also bring down the cost of clinical trials. Together with the BHF data science centre, these models will lead and contribute to national and international trials in heart disease.
Find out more about our research into heart disease