Study uses AI to create cost-effective early detector of heart disease
A study by the Mayo Clinic in Minnesota, US has found that applying artificial intelligence (AI) to electrocardiogram (ECG/EKG) results is a simple and affordable way to indicate the early symptoms of asymptomatic left ventricular dysfunction, a precursor to heart failure.
The Mayo Clinic research team revealed in a recent paper that AI/ECG test accuracy compares favourably with a range of common screening tests, such as mammography tests for breast cancer and cervical cytology for cervical cancer. The findings have been published in Nature Medicine.
Asymptomatic left ventricular dysfunction affects seven million patients in the US and is known to reduce quality of life and longevity. The condition is associated with the presence of a weak heart pump. Although it can lead to heart failure, it is treatable when identified. However, there is currently no inexpensive, non-invasive or painless way to diagnose the condition.
Senior author of the study and chair of the Midwest Department of Cardiovascular Medicine at the Mayo Clinic Dr Paul Friedman said: “Congestive heart failure afflicts more than 5 million people and consumes more than $30bn in health care expenditures in the US alone.
“The ability to acquire a ubiquitous, easily accessible, inexpensive recording in 10 seconds – the EKG – and to digitally process it with AI to extract new information about previously hidden heart disease holds great promise for saving lives and improving health.”
The researchers initially hypothesised that asymptomatic left ventricular dysfunction could be detected in an ECG by a properly trained neural network. They then used the Mayo Clinic’s stored digital data, consisting of 625,326 paired ECG and transthoracic echocardiograms, to identify the best population to analyse. To test their hypothesis, the researchers created, trained, validated and then tested a neural network.
The study concluded that applying AI to a standard ECG can reliably detect asymptomatic left ventricular dysfunction. It also found that, in patients without ventricular dysfunction, those with a positive AI screen were four times more likely to develop future ventricular dysfunction, compared to patients with a negative screen.
Friedman added: “The test not only identified asymptomatic disease, but also predicted risk of future disease, presumably by identifying very early, subtle EKG changes that occur before heart muscle weakness.”