Photo By: Etatics Inc. from Pexels / 2022
AI is now helping doctors detect heart disease even before visible symptoms appear. While some people are cautious about this advancement, one company–Caristo Diagnostics, an Oxford University spin-out–has shown how AI can be used responsibly to pre-empt heart conditions. Their technology enables the identification of individuals with a higher risk of heart disease at an earlier stage, allowing for preventive measures to be implemented and potentially avoiding more serious conditions.
Currently, heart obstructions can be viewed using CT scanners, which provide an image of the heart using X-rays converted to computer signals. CT scans are often the first diagnostic tool used when people experience chest pain. For some, this kickstarts a treatment plan to avoid more serious complications in the future. However, 80% of people are dismissed without a plan put in place, according to a recent study in The Lancet. Caristo Diagnostics have developed another diagnostic tool called the CaRi-Heart AI platform, which can detect inflammation in the heart at an early stage by measuring fat deposition in arteries, something which cannot be viewed on a CT scan. Heart inflammation is one of the first warning signs of serious cardiac diseases–it promotes plaque build-up, making arteries more likely to rupture, hence significantly increasing heart attack risk. The study suggests that inflammation in coronary arteries indicates a 20 to 30 times higher chance of dying from heart disease in the next 10 years. Early diagnosis is imperative to implementing medication or lifestyle changes to lower the possibility of more serious cardiac conditions.
Preventative medicine could be game-changing in the NHS, which currently spends an estimated £7.4bn a year on cardiology. The British Heart Foundation predicts that 7.6 million people are living with heart disease in the UK and 350,000 patients are referred for a cardiac CT scan every year. Once a patient is identified as high risk, doctors can recommend lifestyle changes in diet, exercise and sleep routines, or prescribe treatments such as statins to lower blood pressure and cholesterol, helping prevent serious cardiac diseases.
The technology was rolled out in a pilot project involving 5 NHS hospitals in the UK. 40,000 patients were studied for 2-7 years following a CT scan for suspected heart issues. Of those, 45% were provided with a preventative treatment plan which aimed to reduce their risk of serious heart disease in the coming years. These management changes were made possible by the early detection of coronary inflammation, which was previously undetectable on CT scans. Using AI complements traditional heart disease predictors such as obesity, smoking and diabetes by giving concrete medical evidence to a patient that they should make lifestyle changes and providing a more complete image of a person’s overall heart health. It is hoped that the technology will convince patients to take their health seriously and start making tangible changes for the good of their hearts. The developers hope that AI technology will soon be applied to detect not only heart disease, but also stroke and diabetes, giving patients a head start on prevention. This early diagnosis will not just save lives, but will also reduce the strain on health services by reducing costly emergency admissions and long-term treatments for heart failure.
Article written by Phoebe Lewis, a PhD student at The University of Edinburgh studying supramolecular chemistry for PET imaging of cardiac collagen.
Article edited by Priscilla Wong, a Fourth-Year Biological Sciences (Immunology) student at the University of Edinburgh, and an Online News Editor for EUSci.
Resources:
Chan, K., Wahome, E., Tsiachristas, A., Antonopoulos, A.S., Patel, P., Lyasheva, M., Kingham, L., West, H., Oikonomou, E.K., Volpe, L. and Mavrogiannis, M.C., 2024. Inflammatory risk and cardiovascular events in patients without obstructive coronary artery disease: the ORFAN multicentre, longitudinal cohort study. The Lancet, 403(10444), pp.2606-2618.
Leave a Reply