Artificial intelligence can now calculate the risk of someone’s death from COVID-19 by scanning their blood vessels – and people with inflammation are eight times more at risk
- AI detects blood vessel inflammation in CT chest scans of COVID patients
- Inflamed blood vessels have been linked to severe cases of COVID-19
- By applying innovation, patients can quickly get an anti-inflammatory drug
- This has reduced the risk of death sixfold in patients with severe cases
A new artificial intelligence (AI) tool claims to calculate a patient’s risk of dying from COVID-19 and associated variants by scanning for increased blood vessel inflammation.
Scientists from the University of Oxford have trained an algorithm to recognize a COVID-19 signature in chest CT scans.
The technology detects abnormalities in the fat surrounding blood vessels to measure the degree of inflammation caused by cytokines in infected patients.
Those with increased blood vessel inflammation were up to eight times more likely to die in hospital from the virus, but were also shown to respond well to an anti-inflammatory drug that had a six-fold reduction in the risk of death.
The team believes the innovation could personalize treatment and allow specialists to deliver anti-inflammatory drugs more quickly to save the person’s life.
Severe cases of COVID-19 have been linked to a cytokine storm, which is caused by the immune system flooding the bloodstream with specific proteins called cytokines that cause inflammation in response to the virus.
Scroll down for video
Using this information, scientists at the University of Oxford trained machine learning to recognize a COVID-19 signature in chest CT scans (stock). The technology detects abnormalities in the fat surrounding blood vessels to measure the degree of inflammation caused by cytokines in infected patients. Pictured is a CT chest scan of a person with COVID-19
Professor Charalambos Antoniades, professor of cardiovascular medicine at the University of Oxford, told the British Heart Foundation (BHF): ‘By simply adding an extra step to routine care for people hospitalized with Covid-19 who already have a CT scan, we can now detect and treat patients at high risk for life-threatening complications. possibly their treatment to aid long-term recovery.
But the benefits don’t stop there. We know that this exaggerated immune response to the virus can also cause abnormal blood clotting, which is why we are developing this AI platform to specifically identify Covid-19 patients who are most at risk for a future heart attack or stroke.
“We can also easily pivot our platform to develop a new scanning ‘signature’ to better understand future viruses and diseases plaguing our populations.”
To test the AI, scientists applied the COVID-19 signature to chest scans of 435 people in hospitals in Oxford, Leicester and Bath.
Those with increased blood vessel inflammation were up to eight times more likely to die in hospital from the virus, but were also found to respond well to an anti-inflammatory drug that had a six-fold reduction in the risk of death
The team then compared inflammation levels and risk of death with those with the coronavirus and those without the virus.
The results showed that the levels of cytokine-driven inflammation in blood vessels were much higher in patients with COVID-19, and even greater in patients infected with the B.1.1.7 or ‘alpha’ variant, which was used for the was first identified in the UK.
BHF Associate Medical Director, Professor James Leiper said:
“Over the past year, we’ve supported our scientists to focus their expertise on the global effort to understand Covid-19. This research clearly shows that Covid-19 is a powerful virus that can wreak havoc on our circulation, and that different variants come with different levels of risk.
“There are still many unknowns about how the virus could affect our health in the long term, but this AI tool could ultimately help save lives.”