‘Superhuman’ AI tool spots heart disease in less than 2 seconds

A "superhuman" AI tool has been created to detect heart disease from routine ECGs in less than two seconds. The AI, trained on millions of patient ECGs, extracts more information from these tests than the human eye typically can. While traditional ECGs have long been a vital diagnostic tool for heart attacks and abnormal heart rhythms, they cannot identify heart disease.

An echocardiogram, a type of ultrasound scan, is typically used to diagnose heart disease, but patients often have to wait months to receive this test. The new AI tool can quickly identify signs of heart failure and heart valve disease, two of the most common forms of heart disease. Early diagnosis of these conditions is crucial, as it enables those who need lifesaving medications to receive treatment sooner.

The AI tool was tested on 67,000 patients in the US and identified 81% of those with heart failure and 90% with heart valve disease. Dr. Sonya Babu-Narayan, a consultant cardiologist, noted that AI can provide a readout from an ECG in "the blink of an eye," potentially identifying high-risk patients early. However, the AI cannot definitively diagnose or rule out heart failure or heart valve disease, but it can give a strong indication that someone may have these conditions.

Patients deemed likely to have these issues could then be prioritized for an echocardiogram, potentially leading to faster diagnoses and earlier treatment.