Newswise — A study conducted in University Hospitals of Leicester NHS Trust explored the potential of artificial intelligence (AI) in predicting the risk of lethal heart rhythm disturbances, known as ventricular arrhythmias (VA). Led by Dr. Joseph Barker and Professor Andre Ng, the study found that an AI tool correctly identified the condition in 80% of cases.
Published in the European Heart Journal – Digital Health, the study focused on VA, which originates from the heart's bottom chambers and can lead to sudden death if untreated. Dr. Barker, alongside Dr. Xin Li, co-developed an AI tool named VA-ResNet-50. This tool analyzed Holter electrocardiograms (ECGs) of 270 adults taken during their normal routines at home.
These ECGs were part of routine NHS care between 2014 and 2022, with outcomes known for the patients. Sadly, 159 patients experienced lethal ventricular arrhythmias on average 1.6 years after the ECGs were taken.
Professor Ng highlighted the insufficiency of current clinical guidelines in accurately identifying patients at risk of ventricular arrhythmias, leading to preventable deaths. The study, collating the largest dataset of its kind, found that the AI tool outperformed current guidelines, correctly predicting risk in 4 out of 5 cases.
Notably, if the AI tool indicated a person was at risk, their likelihood of a lethal event was three times higher than normal. This underscores the potential of AI in assessing risk and guiding treatment decisions, ultimately saving lives.
Dr. Barker's contributions were acknowledged with prestigious awards, and further research is planned to expand upon this work. The full paper is available for reference.
This groundbreaking research was made possible by the collaborative efforts of an exceptional team at the NIHR Leicester Biomedical Research Centre. The center, part of the NIHR and hosted by the University Hospitals of Leicester NHS Trust, operates in partnership with several academic institutions and healthcare providers.