AI Breakthrough for Faster, Cheaper and Injection-free Heart Scans

New artificial intelligence (AI) technology could lead to the next generation of heart MRI scans to enable doctors better diagnose people with a range of heart muscle diseases, according to research BHF part-funded and published in Circulation.

The technology would offer doctors and patients a faster, cheaper, safer, and needle-free alternative to the current way of detecting heart disease.

The current gold-standard for imaging and diagnosing heart muscle conditions is performing a cardiac magnetic resonance (CMR) scan that uses a method called ‘late gadolinium enhancement’. This requires the patient to be injected with a contrast agent – a dye-like substance – so any scarring and damage to the heart muscle can be identified. The injection of the contrast substance means that the time to perform the scan is prolonged, increases the cost, and is cautioned in some patients, such as those with severe kidney failure.

Virtual dye to spot heart damage

Now, researchers at the University of Oxford have developed a new AI technique called ‘virtual native enhancement’ by using AI to train machines to predict what a contrast-enhanced image would look like. This new technology was developed using CMR heart scan images from over 1,300 people with hypertrophic cardiomyopathy, a disease where the heart muscle wall becomes thickened.

The AI-enhanced scan images were as clear, and even better quality than the traditional contrast-enhanced images, providing doctors with the same information about a patient’s heart. The new technology is estimated to slash the scan time from the standard 30-45 minutes to 15 minutes, and save more than half of the scan cost.

The researchers now plan to further develop their technology on a range of heart conditions, including in people who’ve had a heart attack.

They aim to test this new AI technology in a large multinational human clinical trial in the near future.

Easier patient experience

Dr Qiang Zhang, BHF-funded researcher at the University of Oxford and lead author of the study, said:

“We’re excited by our new concept where AI could work as a ‘virtual’ dye to replace the injections currently required when someone has a heart MRI scan. Our focus is to develop deep learning solutions for immediate patient benefit, and this AI tool could help to make heart MRI more accessible to patients and hospitals in the coming years.”

Dr Sonya Babu-Narayan, BHF Associate Medical Director, said:

“If successful in clinical trials, this research could have important medical benefits as the same heart diagnosis could be made faster and without an injection, giving heart patients an easier experience.”