In the courtyard of a maximum security prison in Maputo, Mozambique, a man with a shaved head, in an orange t-shirt emblazoned with the term “inmate”, waits patiently, his chest facing a large white tablet hung vertically.

Behind him, a nurse presses the button on the portable X-ray machine connected to an artificial intelligence (AI) program: a new device hailed as a breakthrough in the fight against tuberculosis, a disease caused by bacteria that affects most often the lungs. AI makes it possible to read the radio precisely and instantly, without requiring the expertise of a doctor.

“It’s real time, we have the results in less than five minutes,” explains the caregiver. The image soon appears on the screen of a technician, installed a few meters away, accompanied by a diagnosis: “Radiological signs suggestive of tuberculosis: negative”, displays the computer. This test carried out at the beginning of November is part of a pilot project aimed at examining inmates in three prisons in the Mozambican capital and managed by Stop TB, an organization supported by the United Nations.

Overcrowded prisons are a hotbed of tuberculosis, the world’s second deadliest disease after Covid, which has infected more than 10 million people in 2022 and killed 1.3 million, according to the World Health Organization. health (WHO). Nearly one in four people who contracted the disease last year were in Africa. Mozambique, which has a population of 32 million, has recorded around 120,000 cases.

Early diagnosis helps save lives and stop the spread, because while chronic cough is a hallmark of TB, some carriers have no symptoms. This is particularly true in prison, where tuberculosis spreads through the air and crowded cells provide breeding ground. Mozambique’s prisons were approximately 50% over capacity in 2022, according to the United Nations.

« Science-fiction »

The portable X-ray machine, aided by AI, improves diagnosis through its speed without comparison to skin or blood tests which must be analyzed in the laboratory. In addition, it does not require patients to travel and does without radiologists, who may be rare in rural areas or poor countries, explains Suvanand Sahu, deputy director of Stop TB. “It’s a big technological step forward,” he enthuses.

At the Maputo provincial penitentiary, inmates who test positive are quarantined behind a rusty metal door. Inside, a dozen men wearing masks sit on mattresses on the floor, while clothes, blankets and other personal belongings hang from a rope fixed between two faded blue pillars. Serious cases are treated in the infirmary.

“It’s not easy to see your comrades stretching their legs outside, playing, but you have to accept that I’m sick,” confides Kennet Fortune, detained for ten years on a drug charge, pointing the trees in the prison yard. Testing positive for tuberculosis, he is currently undergoing treatment which may take months. “When the time comes, I can come out again. »

Earlier this month, a WHO report found that global deaths from tuberculosis fell in 2022, a sign of progress in eradicating the disease. And 7.5 million people were newly diagnosed in the same period, the highest figure since the UN agency began monitoring tuberculosis in 1995.

Stop TB’s Sahu hopes the success of pilot projects like Mozambique’s will help secure funding to scale up the use of AI and wearable radios to beat the disease. “Only a few years ago, if I had said at a meeting that we could bring X-rays everywhere that would be read by a machine without the use of radiologists, I would have been told to go write a science fiction,” he smiles.