New artificial intelligence system could be set to revolutionise cardiology diagnosis

Posted: 04/01/2018

An artificial intelligence (AI) system, developed at the John Radcliffe Hospital in Oxford, could be set to revolutionise the diagnosis of heart disease.

When someone is suspected of having a problem with their heart, one of the tools used by cardiologists is a scan of the heart. The new AI system has been designed to review these scans and to pick up tell-tale signs of cardiac problems.

Currently, a consultant cardiologist identifies a cardiology problem by reviewing the timing of the heartbeat as shown on the scan of the patient’s heart. NHS figures show that, of the 60,000 heart scans carried out each year, 12,000 scans are misdiagnosed. The cost to the NHS is estimated at £600 million a year.

The results of the clinical trials of the new AI system are yet to be published and the work is still subject to peer review. However, according to the cardiologist who developed the new system, Professor Paul Leeson, it ‘greatly outperformed his fellow heart specialists’ in clinical trials, picking up details in heart scans ‘much more accurately’ than a cardiologist.

Generally, it is hoped that the introduction of AI could reduce NHS spending. The Government’s healthcare tsar Sir John Bell has been quoted as saying that the introduction of AI might reduce NHS spending on pathology services by 50%.

Camilla Wonnacott, an associate in Penningtons Manches’ clinical negligence team who specialists in cardiology claims, said: “Misdiagnosis of a cardiology problem can result in delays in treatment, leading to permanent heart muscle damage and, in the worse cases, to the patient losing their life. It can also lead to a patient undergoing unnecessary procedures or operations or taking unnecessary medication. A system which increases the accuracy of diagnosis in these circumstances would be very welcome.”

Return to news headlines