A new artificial intelligence tool could turn a routine heart test into an effective early warning system for serious cardiac emergencies. Doctors have developed an AI tool for analysing an electrocardiogram (ECG). It could study ECG in less than two seconds and identify signs of heart valve disease that may not be visible to the trained human eye.Developed by researchers at Imperial College London, the AI tool was trained on millions of ECGs and presented at the European Society of Cardiology Congress 2026 in Munich. In a human trial involving around 67,000 US patients, it was able to detect signs of heart valve disease with up to an astonishing 90% accuracy and heart failure with up to 81% accuracy.How Can An ECG Reveal A Valve Problem?A standard ECG records the heart's electrical activity. Traditionally, it has been used to identify heart problems like irregular heart rhythms, but these electrical patterns, according to researchers, can contain signs about the heart's overall structure and function.The AI looks for patterns across the ECG that may be too complex or subtle for a clinician to recognise visually.That means the machine isn't simply reading the obvious peaks and dips on an ECG. It is looking for hidden patterns buried within the electrical signal.Researchers say this could allow routine ECGs, which are already performed millions of times, to flag people who may have an underlying structural heart problem.Also read: World’s First Live AI-Assisted Brain Surgery Saves Patient’s Sight: How Does The Technology Work?AI Tool Delivered Accurate Results Within Two Seconds The results are particularly striking for heart valve disease as the AI tool achieved accuracy of up to 90% in identifying cases.Heart valve disease occurs when one or more of the heart's valves become narrowed or fail to close properly. This can disrupt the normal flow of blood through the heart. Some people may have few or no symptoms initially. As the disease progresses, however, it can cause breathlessness, fatigue, chest discomfort, dizziness and swelling, and severe disease can eventually lead to heart failure.A patient suspected of a heart emergency often requires an ECG, an ultrasound scan that allows doctors to see the heart's chambers and valves. Waiting times for such tests can be long, meaning some patients may remain undiagnosed. The AI could therefore work as a effective tool, identifying people who should be prioritised for echocardiogram. The AI can analyse an ECG in under two seconds, compared to the much longer process of patients being referred for additional cardiac tests.Dr Ahmed El-Medany, the British Heart Foundation clinical research fellow who led the Imperial College London analysis, described the tool as a "superhuman AI". The phrase has become shorthand for the project. He also said the next challenge is designing portable AI-based ECG readers.Dr Sonya Babu-Narayan, consultant cardiologist and BHF clinical director, said she found it exciting to see AI deliver results from an ECG in "what feels like the blink of an eye."Also read: AI-Designed Viruses Have Been Successfully Tested In Lab But Scientists Are Worried; Here's WhyThe technology is being developed as an extra layer of screening. A positive AI result would not mean that someone definitely has heart valve disease. Instead, it could signal that the patient needs a more definitive investigation, such as an echocardiogram.Heart disease can remain hidden for years before symptoms become obvious. A routine ECG is already a familiar, relatively quick and inexpensive test. If AI can study additional data from that same test, it could help doctors in identifying people who otherwise might not have been investigated and treated further.