Moulay Akhloufi developed this tool using a database of X-rays of healthy lungs and diseased lungs. At the start of the pandemic, it used data from Europe to build its software.
And so came the idea of working with medical imaging, because at that time we already had some images that came mainly from Italy and Spain, because these were countries where there were a lot of infections and then X-ray images that were made public, there weren’t many, explains Mr. Akhloufi.
In the field, the software would be used to facilitate triage in emergency rooms and in COVID-19 detection centers, and this would strengthen prevention and patient isolation measures.
This tool could allow doctors to separate patients with COVID-19 from those with pneumonia, for example.
Moulay Akhloufi is working in partnership with Dr. El Mostafa Bouattane, of the Montfort Knowledge Institute, in Ottawa, to test whether the tool makes a good diagnosis on real patients with COVID-19.
Dr Bouattane, however, cautions and explains that the new tool is not a replacement for current tests like the testPCR and rapid tests.
Now we have rapid tests, but it could always be another test that could help, often these patients who come with mainly pneumological symptoms, we ask them for an X-ray, so if we integrate the algorithm into these X-rays, that can at least automatically give the result of COVID, and our algorithm also provides images with colors that show the intensity of the damage caused by the virus, he says.
Based on a report by Maria-Isabelle Noël