A team from Oxford University claims that technology that can identify hidden data in CT images could save thousands of lives.
In Britain, an artificial intelligence system may soon be in use that can predict a person’s risk of having a heart attack up to ten years from now.
The National Institute for Health and Care Excellence (Nice) is evaluating the technology, which has the potential to save thousands of lives annually. A decision regarding its implementation in the National Health Service (NHS) is anticipated by the end of the year.
The researchers behind the experiment have also disclosed that they are developing analogous artificial intelligence systems to identify individuals at risk of developing diabetes or a stroke.
Prof. Charalambos Antoniades, the head of the Orfan (Oxford Risk Factors And Non-Invasive Imaging) project, stated, “This technology has now been tested at several hospitals across the UK and results have been tremendously encouraging.” “It would prevent thousands of people from dying from heart disease or having early heart attacks if it were implemented nationwide.”
According to Antoniades, more than 300,000 people in Britain receive CT scans to rule out cardiac anomalies including artery blockages in cases of severe chest discomfort. On the other hand, less than 20% of individuals scanned had coronary artery blockages or hazardous narrowing. More than 80% of them exhibit no anomalies. Many times without any prescription, they are reassured and sent home, according to Antoniades, the University of Oxford’s professor in cardiovascular medicine. “But these assurances are frequently misplaced.”
Actually, almost two-thirds of this “safe” group go on to experience serious cardiac events, such as heart attacks, which can occasionally be fatal. “It is evident that we have been overlooking signals from our scans that could alert us to individuals who are truly in danger,” he continued. “We think AI is the ideal technology to deal with this enormous healthcare problem.”
The study, which was conducted by a group at Oxford University’s Radcliffe Department of Medicine and published last week in the Lancet, is intended to identify the anomalies that conventional CT (computed tomography) scans miss. With this knowledge, physicians could prescribe anti-inflammatory medications to patients as a preventative measure.
A significant portion of the issue, according to Antoniades, was that a CT scan was unable to detect arterial damage brought on by inflammation. “By applying AI to improve our CT scan images, we were able to uncover hidden information and demonstrate the extent of damage.” We could not have imagined this in the past, but we can now.
The method provides important information on the condition of our heart arteries by utilizing data on the features of coronary plaques and alterations in the fat around inflamed arteries. According to him, these measurements essentially indicate the patient’s absolute risk of suffering a fatal cardiac episode during the following ten years.
These risk factors were initially determined through US case studies; however, 40,000 patients at UK hospitals were used to assess the data.
“What we discovered was unambiguous. It was shown that patients with significant levels of inflammation in their coronary arteries were also far more likely to experience serious cardiac conditions including heart attacks. We now know how to identify the unseen causes of heart attacks.
The British Heart Foundation supported the study, which found that in 45% of cases, doctors changed a patient’s course of treatment based on information gleaned from AI analysis. High dosages of statins or medications like colchicine, which are known to lower the risk of cardiovascular disease, are part of these treatments.
The Food and Drug Administration is currently evaluating this UK-made technology in the US, and Antoniades continued, “We are also planning to expand delivery of this technology in Europe, where it is already approved for clinical use.”
