The artificial intelligence tool can detect changes in the heart from MRI images with 40% better accuracy than a doctor manually reviewing the scans, according to a study.
Researchers estimate it could save clinicians 3,000 days of time if adopted for all MRI scans in the UK.
It is initially deployed at three hospitals in London – University College London (UCL) Hospital, St Bartholomew’s Hospital and the Royal Free Hospital. The rollout is expected to expand to 40 sites across the UK and globally later this year.
Dr Rhodri Davies, Associate Clinical Professor at UCL who led the research, said: “Our new AI reads complex heart scans at record speed, analyzing the structure and function of a patient’s heart with more precision than ever.
“The beauty of the technology is that it replaces the need for a doctor to spend countless hours analyzing scans by hand.
“After this initial roll-out to the NHS, we will collect the data, then train and refine the AI so that it is accessible to more heart patients in the UK and around the world.”
Around 120,000 cardiac MRIs are performed each year in the UK.
AI is designed to diagnose new conditions when a patient is first evaluated for heart disease. It can also help doctors check whether patients are responding well to treatment.
The team trained the AI to measure the size of the left ventricle (the main pumping chamber of the heart), the thickness of the heart muscle, and the ability of the left ventricle to pump blood throughout the body.
They did this using heart MRIs from 1,923 people – including patients with seven different heart conditions at 13 different hospitals and using 10 different models of MRI scanner.
The AI was then validated on 109 other patients who were scanned twice. AI technology has been found to analyze scans more accurately than three doctors.
Dr Sonya Babu-Narayan, Associate Medical Director at the British Heart Foundation, which funded the research, said: “This is a huge step forward for doctors and patients, revolutionizing the way we can analyze MRI images of a person’s heart to determine if they have faster heart disease.
“The pandemic has resulted in a backlog of hundreds of thousands of people waiting for life-saving cardiac exams, treatment and care. Despite the delay in cardiac care, as people remain on waiting lists, they risk preventable disability and death.
“That’s why it’s encouraging to see innovations like this, which together could help speed up cardiac diagnostics and lighten the workload so that in the future, we can give more patients NHS heart patients the best possible care much sooner.”
The research was published in the Journal of Cardiovascular Magnetic Resonance.