Subtests remain a huge challenge in the vast South American country, however, AI is helping to fill the gap, thanks to a formula called RadVid-19 developed algorithms from the German company Siemens and the Chinese company Huawei.
Brazil has been more affected by the pandemic than any other country in the United States, with nearly 2.8 million infections and 95,000 deaths.
Experts say the numbers would be much higher if there was more widespread evidence.
RadVid-19 seeks to fill this gap and help doctors make a decision about the right remedy for their patients.
Analyze chest x-rays and CT scans to locate points in the lungs of patients likely to be markers of infection with the new coronavirus.
“The software identifies these spaces and evaluates the probability of a COVID-19 case,” explains Marcio Sawamura, deputy director of the radiology center at the University of Sao Paulo Clinical Hospital.
The program shows doctors on a PC screen how their patients’ lungs evolve over time and analyze the white and yellow circles that mark a possible infection.
The software is used through 43 Brazilian hospitals, 60% of which are public, thanks in component to the investment of the Inter-American Development Bank.
Update a lab diagnosis with a doctor.
But it can help doctors make a decision about which remedy to stick to the insuffering in short until the laboratory effects return, in a country where large-scale testing campaigns have not been introduced and where President Jair Bolsonaro has been criticized for downplaying the pandemic.
Creativity in crisis
“With Brazil doing fewer tests than it should, CT scans and X-rays end up being used as diagnostic tools,” says Arthur Lobo, radiologist in the northern city of Belem.
“He helped us make diagnoses when we had doubts.”
It was a welcome relief in the early stages of the pandemic in Brazil, when doctors rushed to stick to an explosive infection curve, says Claudia Leite, professor in the Department of Radiology and Oncology at the University of Sao Paulo.
“At first we were really anxious, because the lab effects took a long time to reappear and the patient started to get worse, breathing difficulties appeared and we still didn’t have a diagnosis,” he says.
Using AI, he says, “We were able to conclude that it is a COVID and take the right steps. And the verification effects showed later.”