Novel AI model can predict health risks from ECGs

Advancements in artificial intelligence (AI) are transforming healthcare, offering innovative ways to predict and manage diseases. One such breakthrough comes from researchers at Imperial College London and Imperial College Healthcare NHS Trust, who have developed an AI model capable of predicting health risks — including early death — using data from routine electrocardiograms (ECGs).

Published in Lancet Digital Health, the study explores how the technology could reshape patient care by identifying risks early and prioritizing urgent cases for treatment.

How does it work?

The AI model, known as AI-ECG risk estimation (AIRE), leverages large datasets comprising millions of ECGs from routine care. ECGs, which record the heart's electrical activity, provide detailed insights into the functioning of the heart's chambers. AIRE analyzes these electrical patterns with precision beyond the human eye, detecting subtle abnormalities that may signal underlying or future health issues.

Dr. Arunashis Sau, who led the study, noted: "The AI model detects much more subtle detail, so it can 'spot' problems in ECGs that would appear normal to us, potentially long before the disease develops fully."

AIRE was able to predict patients' risk of death within ten years with 78% accuracy. It can also assess risks for various conditions, including heart rhythm problems, heart attacks, and even non-heart-related causes of death, by capturing systemic health information reflected in ECGs.

Why does it matter?

The implications of AIRE are profound. Currently, ECGs are a standard diagnostic tool for heart diseases, but this AI-enhanced approach could revolutionize their use by predicting a range of future health risks.

According to Dr. Fu Siong Ng, senior author of the study, "This AI model could have major benefits for the NHS and globally, improving patient longevity and quality of life while reducing waiting lists and optimizing resource allocation."

Beyond heart health, AIRE offers insights into conditions like diabetes and kidney disease, which subtly impact the heart. The potential to detect accelerated aging or structural changes in the heart makes this tool invaluable for preventive care.

The context

AI's integration into healthcare is not new, but its use in routine diagnostics like ECGs is a significant step forward. Historically, ECGs have provided a snapshot of heart health for over a century. However, as Professor Bryan Williams of the British Heart Foundation explained,
"This research has demonstrated the extraordinary power of AI to gain important health insights from a routine test, taking ECGs beyond their traditional use."

The research, supported by the British Heart Foundation and other organizations, underscores the potential of AI to enhance clinical decision-making. Trials are set to begin by mid-2025 across NHS hospitals, aiming to evaluate how AIRE can improve real-world patient outcomes.