Doctors have developed a “superhuman” artificial intelligence (AI) tool capable of detecting heart disease in less than two seconds. Trained on data from millions of patients, the innovative technology can extract more information from a routine electrocardiogram (ECG) than the human eye can.
Traditional ECGs have been used for a century to diagnose heart attacks and heart rhythm disorders, but they cannot detect structural heart disease. For this, patients need an echocardiogram (ultrasound examination of the heart), which often means waiting for months at clinics.
The new AI algorithm can rapidly identify signs of heart failure and valvular heart disease — two of the most common forms of heart disease. Details of the technological breakthrough were presented in Munich at the annual congress of the European Society of Cardiology (ESC), the world’s largest cardiology conference.
In a large clinical study involving 67,000 patients in the United States, the AI accurately identified 81% of heart failure cases and 90% of patients with heart valve disease. The study was led by scientists from Imperial College London and funded by the British Heart Foundation (BHF).
According to Professor Fu Siong Ng, Professor of Cardiology at Imperial College London, the technology will not completely replace echocardiography and cannot independently provide a definitive diagnosis. However, it could play a critically important role in identifying and prioritizing high-risk patients. This means that people who need urgent care could receive an ultrasound examination much sooner and begin potentially life-saving treatment without unnecessary delays.
Experts also believe that the model could be integrated into a hospital-wide network, allowing it to detect disease incidentally in patients who undergo an ECG for an entirely different reason. The team’s next goal is to develop handheld, portable AI-powered ECG readers for doctors.
Another interesting innovation was also discussed at the Munich congress. Researchers from the University of Tokyo and the Institute of Science Tokyo presented an AI system capable of rapidly and accurately detecting undiagnosed high blood pressure and type 2 diabetes by analyzing a five-second video of a patient’s face.
