If there’s anything worse than experiencing a heart attack, it’s probably waiting for one you’ve been told is coming. But that’s exactly what new AI software that will soon be available on the NHS does – up to a decade in advance. The good news is the early warning system warns of heart attack risk rather than the inevitable, allowing for evasive action.
It is believed the new algorithm-based AI screening could potentially save tens of thousands of lives in the near future. The technology breakthrough behind the method is able to spot heart attack risk in patients that wouldn’t be picked up by scanning technology currently in use. At present, CT scans are used to assess if the conditions which could signal a future heart attack are present. Doctors examining CT scans look for signs of narrowing blood vessels but around 20% of patients given the all clear still go on to be struck by a heart attack within a decade.
But the new AI tool developed by Oxford University researchers is able to read the same CT scans in more detail, picking up early warning signs such as very slight alterations to fat cells, tissue scarring and the appearance of small blood vessels. AI analysis of these early developments that can indicate future heart attack risk has been found in clinical trials to predict cardiac arrest up to 9 years in advance with 85%-90% accuracy.
The research has been published in the European Heart Journal and was also yesterday presented at the European Society of Cardiology congress being held in Paris this week. The study has been led by Professor Charalambos Antoniades, who stated:
“What we can see with CT is the tip of the iceberg. We have developed an extremely powerful artificial intelligence system which [also] looks at . . . what is under the surface, which is the big part of the risk.”
The British Heart Foundation’s associate medical director, Professor Metin Avkiran, also described the research as a “significant advance”. UK government funding has now been secured with a view to further developing the technology with a view to it being available through the NHS within two years.