Health Farm

-- Artificial intelligence – computer modelling – is the brains behind driverless cars, chatbots, and product suggestions we never knew we needed – so why not detecting irregular heartbeats.

Atrial fibrillation – an irregular heartbeat – affects millions of people worldwide and increases the risk of stroke and long-term heart problems.

It is relatively easy to diagnose when the heart beats irregularly and the patient is hooked up to an electrocardiograph machine – but not so much when a dodgy ticker returns to normal beating.

Hidden abnormalities no match for AI

Researchers at the Mayo Clinic in the US modelled data from heart testing carried out on nearly 181,000 patients between 1993 and 2017.

The computer model searched for subtle signs of past irregular rhythms, such as scarring of the heart, which are difficult to spot with the human eye. The Mayo Clinic’s model correctly identified the subsequent diagnosis of atrial fibrillation from earlier test results that recorded a normal heartbeat in 83% of cases.

The approach could lead to earlier and easier detection of heart problems.