Heart failure detection faster with Us2.ai, results interchangeable with humans: the OPERA study

Using AI to interpret images from a handheld ultrasound device could help speed up detection of heart failure, new research suggests.

Dr Ross Campbell, from the University of Glasgow, said: “Our breaking new OPERA results show that investing in AI in healthcare could offer remarkable benefits to both patients and the NHS.

“We have shown that AI can interpret echocardiogram images accurately, and given AI can produce a report in a fraction of the time, this could really make a difference in allowing us to make early diagnosis of heart failure possible.”

The OPERA results were presented in a Late Breaker session at the European Society of Cardiology (ESC) Conference in Amsterdam. The results showed that Us2.ai assisted echocardiography is a fast and convenient tool that enables point of care testing for early diagnosis of heart failure.

Despite evidence of the benefits to early treatment of heart failure, 75% of cases are diagnosed far too late in acute hospital settings which are already overburdened. The OPERA study was conducted during the COVID pandemic and led to dramatic reductions in hospital wait times, demonstrating the potential to streamline the diagnosis of heart failure so that patients receive the care they need earlier, to improve their outcomes.

OPERA results show that Us2.ai interpreted heart ultrasound images, including those taken with a handheld device, are as effective as using a typical ultrasound machine operated by an expert in measuring the pumping action of the heart. However, while a standard analysis of an echocardiogram with a human operator takes around 30 minutes, the AI is able to interpret the images in just one minute – a huge reduction in clinical time which has the potential to significantly speed up heart failure diagnosis waiting times.

Download the full ESC Late Breaker presentation and OPERA slides.