Scientists have successfully built artificial intelligence (AI) that can diagnose patients with heart disease and lung cancer much earlier than human doctors.
The ability to accurately pick up these problems at a less advanced stage could not only save people from the fatal diseases, but cut huge costs to the health service.
Ross Upton, CEO of Ultromics who created the AI, tells HuffPost UK: “Currently 1 in 5 patients are misdiagnosed, we’ll be able to reduce the number of patients being misdiagnosed by more than 50% which will save thousands of lives.”
“We plan to give the technology to the NHS for free this year. ”
Currently cardiologists use scans to look for irregularities in patient’s heartbeats as an indicator that something isn’t right, but this method has a 20% margin of error.
Meaning of the 60,000 patients who have a scan every year, 12,000 are either sent home when they are at risk, or are given a false positive and undergo unnecessary surgery, something which the BBC says creates a £600 million bill every year.
But the AI, developed by a team at John Radcliffe hospital in Oxford and tested in six different clinical units, is able to detect details in the scans that doctors can’t see.
It then gives a recommendation which means it believes there is a risk.
Initial feedback indicates the system is the most accurate echocardiography software available anywhere in the world, and improves diagnosis by more than 90%.
Hugely outperforming the rates achievable to human eyes alone.
The team said: “Making a diagnosis from echo relies on experienced clinicians having to make qualitative judgements based on only a fraction of the data that is potentially available to them from a typical scan.
“But our technology extracts more than 80,000 data points from a single echocardiogram image to overcome subjectivity and increase diagnostic accuracy.”.
If these results are confirmed by a secondary round of testing, the system will be rolled out in medical units across the country, for free.
The potential impact has been hailed by Oxford University’s Professor of Medicine, John Bell, as something that could “save” the NHS.
“There is about £2.2bn spent on pathology services in the NHS. You may be able to reduce that by 50%,” by using Ultromics.
Another AI system is looking for signs of lung cancer. It searches for large clumps of cells called nodules.
Doctors can’t tell whether these clumps are harmless or will go on to become cancerous.