Researchers at Oxford University in the UK have developed what they call the world’s most accurate echocardiography software that improves the diagnosis of coronary heart disease by more than 90%.

The solution, called Ultromics, uses artificial intelligence (AI) to identify potential heart problems in patients. This likelihood is calculated based on the information gathered from the scans of 1,000 patients who were treated over the last seven years, and whether or not they went on to have heart problems.

The introduction of such AI technology could greatly improve the treatment of patients. Cardiologists currently determine whether a patient has a heart problem based on the timing of the heartbeat in scans. However, this method is prone to mistakes; it’s estimated that one in five doctors gets the diagnosis wrong leaving the misdiagnosed patient to either go home and have a heart attack or undergo unnecessary surgery.

According to a BBC article, in the UK 60,000 heart scans are carried out each year, of which 12,000 are misdiagnosed. This costs the National Health Services (NHS) around £600m in unnecessary operations and the treatment of people who were given the all-clear but who went on to have heart attacks.

According to the researchers, their system is far more accurate than doctors because it can pick up more details in the heart scans and sends out an alert if the patient is at risk of having a heart attack.

According to the cardiologist who developed the system, Professor Paul Leeson, Ultromics has undergone clinical testing in six cardiology units where it has outperformed cardiologists. These results, which will be published once the data has been reviewed by experts, indicate that the AI system could save the NHS more than £300m a year.

“As cardiologists, we accept that we don’t always get it right at the moment. But now there is a possibility that we may be able to do better,” said Prof Leeson.

Following the publication of the study, Ultromics will be rolled out to NHS hospitals for free in mid-2018.

For more information contact, like us on Facebook or tweet us @eHealthNewsZA.