'Superhuman' AI Tool Spots Heart Disease in Less Than 2 Seconds

Doctors have created an AI tool that can find signs of heart disease in less than two seconds. The tool was shown off at the world's largest heart health conference, the European Society of Cardiology annual congress, held in Munich, Germany. Scientists say the technology could help millions of people get diagnosed and treated much sooner than they do today.

The AI tool works by reading results from a test called an electrocardiogram, or ECG. An ECG records the electrical activity of the heart, including how fast it beats and whether the rhythm is normal. Doctors have used ECGs for about 100 years, and around one billion of these tests are done every year around the world. The test is quick, common, and easy to do.

However, a regular ECG cannot detect heart disease on its own. To find heart disease, patients usually need a different test called an echocardiogram, which is a type of ultrasound scan. Many patients have to wait several months just to get that scan, which can be a dangerous delay for people who are seriously ill. That long wait is exactly what scientists are hoping to fix.

The new AI tool was trained using data from millions of patients. It can spot two of the most common types of heart disease — heart failure and heart valve disease — by finding patterns in ECG results that human eyes often miss. In a study of 67,000 patients in the United States, the tool correctly identified up to 81% of people with heart failure and up to 90% of those with heart valve disease. Those are very strong results for such a fast test.

Dr. Sonya Babu-Narayan works for the British Heart Foundation, which helped pay for the research. She said the technology is exciting because it can read an ECG almost instantly. She explained that while the tool won't catch every single case of heart disease, it can help doctors find the patients who are most at risk. Those high-risk patients can then be sent for more tests right away, instead of waiting months in line.

Professor Fu Siong Ng from Imperial College London helped lead the research. He said patients can often wait several months for a heart ultrasound after their doctor refers them for one. With this AI tool, the riskiest patients could be moved to the front of the line and get their scans faster. That faster path to diagnosis could mean getting life-saving treatment much sooner.

The tool could also help in another big way. Sometimes people get an ECG for a completely different reason, and they don't even know they might have heart disease. The AI could be set up to scan all ECGs done in a hospital and flag anyone who might have heart failure or heart valve disease. This means some people could be diagnosed with a serious condition before they ever feel seriously sick.

Dr. Ahmed El-Medany, a researcher at Imperial College London who led the analysis, called it a 'superhuman AI.' He said the next big goal is to build small, handheld ECG readers powered by AI that doctors and nurses could carry with them. At the same conference in Munich, researchers also shared news about a different AI tool. Scientists from Japan showed that AI can analyze a five-second video of a person's face and help detect high blood pressure and type 2 diabetes, two conditions that millions of people have without knowing it.