In remote communities across the globe, a new ally has emerged in the battle against tuberculosis, the world's leading infectious killer. Artificial intelligence is now being used to screen for the disease in seconds, reaching populations that have long been underserved by traditional healthcare systems.
This technological shift is addressing a critical shortage of radiologists in over 80 low- and middle-income countries, offering a rapid and accessible diagnostic tool. By analyzing chest X-rays with remarkable speed, AI is helping healthcare workers identify potential TB cases on the spot, a process that previously could take weeks.
Key Takeaways
- Artificial intelligence is being deployed in over 80 low- and middle-income countries to screen for tuberculosis (TB).
- The technology analyzes mobile X-ray images in seconds, addressing a severe global shortage of radiologists.
- Tuberculosis is the world's top infectious disease killer, responsible for over 1.2 million deaths annually.
- While AI accelerates diagnosis, experts raise concerns about regulation, accuracy over time, and ensuring patient access to treatment after diagnosis.
A New Frontline in Global Health
Tuberculosis claims the lives of approximately 3,500 people every day. The bacterial infection, which primarily affects the lungs, has seen a rise in cases, with 10.8 million new infections reported in 2023. A significant barrier to controlling its spread has been the difficulty of diagnosis, especially in rural areas.
"There are countries in which there are less than five radiologists. It's like a disaster," said Dr. Lucica Ditiu, executive director of the Stop TB Partnership. She described the new technology as "revolutionary," enabling screenings in places previously thought impossible, such as among nomadic groups in Nigeria.
Mobile health units equipped with portable X-ray machines and AI software are changing this reality. A trained technician can now capture an image, have it analyzed by an algorithm, and receive a risk score almost instantly. The system often produces a visual heat map of the lungs, highlighting areas of concern for immediate follow-up.
TB by the Numbers
- 1.2 million+ deaths annually worldwide.
- 10.8 million new cases in 2023.
- 80+ low- and middle-income countries using AI for screening.
- $200 million invested by The Global Fund in AI-enabled TB screenings over four years.
How AI Works in the Field
In a community health center in Mali, the impact is tangible. A local health worker, not a doctor, can manage the screening process. After taking an X-ray, the image is sent to a computer where the AI model provides an immediate analysis. Patients with high-risk scores are then asked for a sputum sample for laboratory confirmation.
This two-step process has proven highly efficient. Bassy Keita, a program officer with the local nonprofit ARCAD Santé PLUS, noted that since incorporating AI, the number of sputum samples they need to process has been cut by about half. This is particularly beneficial for children, who often struggle to produce the required sample.
"Having AI makes a big difference," explained Keita. It allows his team to quickly clear individuals with no signs of the disease and focus resources on those who need them most.
The technology's effectiveness in remote locations is a key advantage. Peter Sands, executive director of the Global Fund to Fight AIDS, TB and Malaria, highlighted its use in refugee camps. "There are no radiologists. So who gets to look at the [X-ray] and say: 'Is there a problem here or not?' Well, actually, AI does," he stated.
The Promise and the Peril
The development of these AI models has been surprisingly rapid and cost-effective. Regina Barzilay, a professor at MIT, built a TB detection model in a few months for under $50,000. She believes TB is an ideal candidate for AI because the signs are visual on an X-ray, making it straightforward to train an algorithm.
"AI is going to be adopted much faster in developing countries because they have serious unmet needs and the clinician understands they need other help," Barzilay predicted. She compared the adoption to how many parts of Africa leapfrogged landlines and went directly to mobile phones.
Concerns Over Regulation and Reliability
Despite the optimism, some healthcare professionals urge caution. The rapid deployment of AI in regions with limited regulatory oversight raises important questions about patient safety and accountability. Without established frameworks, it is unclear what happens if an AI model makes a mistake.
Dr. Erwin John Carpio, a radiologist in the Philippines who helped draft AI guidelines for his country, pointed to a significant challenge. "It is a real challenge for the developing nations, because usually the technology is offered to us for free," he said, emphasizing the need to avoid potential problems from unregulated tools.
One major concern is a phenomenon known as "model drift," where an AI's performance deteriorates over time. "They fail silently. They don't tell you that they're making a mistake," Carpio warned. Preventing this requires continuous quality control from a team of experts, including radiologists, data scientists, and AI engineers, which adds complexity and cost.
Balancing Innovation with Responsibility
Proponents argue that the benefits of AI in these settings far outweigh the risks, especially when the alternative is often no screening at all. The World Health Organization (WHO) endorsed the use of this technology in 2021, providing guidance on how to calibrate it for local populations.
The ultimate goal is to connect a diagnosis with effective treatment. The AI tool is just the first step. After a positive screening, patients must have access to the six-month course of antibiotics required to cure the disease.
As AI becomes more integrated into global health initiatives, the focus is shifting from simply identifying the sick to ensuring they receive the care they need. The technology provides a powerful new way to find patients, but the human element of healthcare remains as critical as ever.