AI heart disease tool spots hidden risks in under 2 seconds

A routine heart test may soon reveal much more. Researchers at Imperial College London have developed an AI heart disease tool that can identify signs of serious cardiac conditions in less than two seconds, using data from a standard electrocardiogram, or ECG.

The development could help doctors spot high-risk patients earlier and send them for further checks before their condition becomes more dangerous. Here's all that you need to know!

How the AI heart disease tool works

An ECG is one of the most common tests in medicine. It records the electrical activity of the heart, including its rhythm and rate, and has been used for decades to help detect heart attacks and abnormal heart rhythms.

However, a standard ECG is not usually enough to diagnose conditions such as heart failure or heart valve disease. These conditions are typically confirmed using an echocardiogram, which is an ultrasound scan of the heart. In many healthcare systems, patients can wait months for this scan.

The new AI system aims to bridge that gap by analysing routine ECG results and detecting patterns that may be too subtle for the human eye to see.

This research was led by Professor Fu Siong Ng, a consultant cardiologist and researcher at Imperial, with Dr Arunashis Sau also playing a key role in developing the AI approach.

Why early detection matters

The tool has been trained on millions of patient records and is designed to highlight people who may be at high risk of heart failure or heart valve disease. Both are common forms of heart disease, and both can become life-threatening if they are not identified and treated in time.

In a US trial involving 67,000 patients, the AI identified up to 81% of people with heart failure and up to 90% of those with heart valve disease. The research was funded by the British Heart Foundation and presented at the European Society of Cardiology annual congress in Munich.

What it could change for patients

The main promise of the technology is speed. If an ECG can quickly indicate that a patient is likely to have a heart abnormality, doctors could prioritise that person for an echocardiogram instead of placing them on a standard waiting list.

That could mean quicker confirmation, faster treatment and potentially better outcomes. Researchers also believe the tool could be useful in cases where heart disease is not initially suspected. Since ECGs are widely used for different medical reasons, the AI could be run across hospital ECGs to flag hidden risks that might otherwise go unnoticed.

What the researchers are working towards

The research is also moving beyond the laboratory. The Imperial team has helped establish Cardiovolt.ai, a spinout company focused on bringing AI-powered ECG analysis into real-world healthcare. The company is exploring how the technology could eventually be used in hospitals and potentially in portable ECG devices.

Dr Sau has described the approach as looking for things that are effectively “superhuman”, meaning patterns in an ECG that no cardiologist could reliably identify by eye alone. His work combines his clinical training in cardiology with AI research at Imperial's National Heart and Lung Institute.

Why it is not a replacement for doctors

The AI does not provide a final diagnosis and cannot rule out heart disease on its own. Instead, it offers a strong signal that a patient may need further investigation. Doctors would still need to interpret the result alongside symptoms, medical history and follow-up scans.

This AI technology represents a potential step towards faster, more targeted cardiac care. If validated further and adopted safely, AI heart disease screening could turn one of medicine’s most routine tests into an early warning system for hidden heart problems.