A new AI model is being used to help diagnose one of the world's deadliest diseases.

According to NPR, a mother went to a community health center in Mali because she felt ill. After taking an X-ray, doctors fed it to an AI algorithm. Professionals soon found she was positive for tuberculosis, also known as TB.

Typically, getting a TB test result in regions like Mali could take weeks, as it requires submitting a sample and waiting for lab results. However, mobile X-rays and trained AI algorithms can provide same-day results, per NPR.

Here's an example of how these systems work: the patient is X-rayed. The image is sent to a computer running the AI model, which scores it based on how closely it resembles TB. 

For people suspected of having the disease, NPR said the health worker can collect a sputum sample for lab confirmation.

"You have an X-ray. You have a label which says whether they have it or not — and you just train the model," said Professor Regina Barzilay, per NPR. 

She's one of the experts who built a TB-detecting AI model. "It's straightforward, very cheap, very fast to develop."

TB is one of the world's deadliest diseases, with up to 3,500 people dying from it per day, according to NPR. But in regions with limited access to top-quality medical care, X-ray machines and AI models can speed up diagnosis and treatment.

Over 80 low- and middle-income countries have begun using AI models for TB screening, NPR reported. 

"It is revolutionary," said Dr. Lucica Ditiu, the executive director of the Stop TB Partnership. Her organization was one of the first to develop this technology over the past eight years.

Would you swallow a robot pill to improve your health? Absolutely 💯 Maybe once it's more common 🤔 Only if I were sick 🤒 Only as a last resort ☠️ Click your choice to see results and speak your mind.

Additionally, using AI screenings could help reduce the number of sputum samples needed. Fewer samples help save time and resources.

Many other industries have also begun adapting AI for lifesaving measures, too, including weather forecasting and invasive species detection.

That said, some people are more wary about adding AI models to the mix. Without regulations, patients from these areas could face safety issues, as NPR observed. Plus, the AI's quality could deteriorate over time, putting patients at risk of incorrect diagnoses.

There is also concern about AI's impact on humans and the environment. 

AI data centers consume significant amounts of water, require rare elements, and use large amounts of electricity, according to the UN Environment Programme. Using these resources could drain drinking water supplies and increase planet-warming air pollution.

Barzilay, on the other hand, sees AI as vital to the health and safety of underserved populations. 

"AI is going to be adopted much faster in developing countries because they have serious unmet needs," she said, per NPR.

With that in mind, while AI models can provide much-needed access to health screenings, more progress is necessary. But with adequate healthcare infrastructure and AI safety regulations, time will tell how this technology may change the world.

Get TCD's free newsletters for easy tips to save more, waste less, and make smarter choices — and earn up to $5,000 toward clean upgrades in TCD's exclusive Rewards Club.