AI-Driven ECG Technology Detects Risk of Severe Heart Block Condition

Recent advancements in artificial intelligence (AI) have led to the development of an innovative ECG analysis tool that can predict the likelihood of a patient developing a life-threatening heart condition known as complete heart block. This breakthrough, crafted by researchers from Imperial College London, leverages machine learning to analyze standard electrocardiogram (ECG) recordings, enabling early detection of electrical signal disruptions within the heart.

Heart block occurs when the electrical signals that coordinate the heart's contractions are delayed or blocked, impairing normal heartbeat rhythm. This can result in symptoms like fainting, fatigue, falls, or even sudden cardiac death if left untreated. Currently, clinicians rely on ECG data and international guidelines to diagnose or anticipate heart block; however, these methods often lack sensitivity, especially because early-stage abnormalities can be intermittent and easily missed.

The new AI tool, named AIRE-CHB, significantly improves prediction accuracy. It was trained on over 1.1 million ECGs from nearly 190,000 patients at a hospital in Boston, and tested on an independent dataset comprising more than 50,000 individuals from the UK Biobank. In studies published in JAMA Cardiology, AIRE-CHB achieved an 89% accuracy in predicting future heart block risk, a notable improvement from the 59% accuracy of existing methods. High-risk patients identified by the system were found to be 7 to 12 times more likely to develop the condition.

This advanced prediction capability allows healthcare providers to monitor at-risk individuals more closely, potentially initiating early interventions such as pacemaker implantation to prevent severe complications. Dr. Arunashis Sau, a leading researcher, emphasized the importance of early identification, which can help avoid emergencies, injuries, or fatalities. The tool could also be particularly valuable for evaluating patients with unexplained fainting episodes, often a symptom of underlying electrical conduction problems.

The researchers have plans to trial AIRE-CHB within NHS hospitals later in 2025, aiming to integrate this AI technology into routine clinical practice. Their work builds upon other AI models developed by the team that predict various cardiovascular and health risks, including early death, hypertension, diabetes, and heart valve disease.