The World Health Organization has granted prequalification status to CAD4TB, a computer-aided detection tool that uses artificial intelligence to analyse chest X-rays for signs of tuberculosis. The prequalification — WHO’s mechanism for assuring the quality, safety, and efficacy of health products for procurement by UN agencies and national health programs — makes CAD4TB the first AI-powered diagnostic tool to receive this designation for use in low- and middle-income countries.
CAD4TB is developed by Delft Imaging, a Netherlands-based medical imaging company. The software analyses digital chest X-rays and assigns a probability score for TB-related abnormalities, enabling health workers without specialist radiology training to prioritise which patients require sputum testing or further clinical review.
Why this matters for global TB control
Tuberculosis remains the world’s leading infectious disease killer after COVID-19, responsible for approximately 1.3 million deaths in 2023 according to WHO data. A significant proportion of TB cases globally go undetected because access to radiologists and laboratory diagnostics is limited in high-burden countries, particularly in sub-Saharan Africa and South and Southeast Asia.
AI-assisted X-ray reading tools can operate without a specialist present, require only a digital X-ray machine and an internet connection or local compute, and can process images in under a minute. Field studies in Uganda, Zambia, and India have shown CAD4TB achieving sensitivity comparable to trained radiologists at the threshold settings WHO recommends for screening contexts.
What prequalification enables
WHO prequalification opens the door to procurement through the Stop TB Partnership’s Global Drug Facility and UNICEF supply chains, significantly reducing the commercial barrier for adoption in countries that rely on international procurement mechanisms. Countries that previously could not justify purchasing CAD4TB through national budgets — due to the absence of WHO endorsement required by domestic procurement rules — can now do so.
Delft Imaging has said it will offer CAD4TB at tiered pricing for low-income countries under the prequalification framework, though specific pricing terms have not been published.
Broader context for AI in global health
The WHO prequalification comes as regulatory bodies globally are developing frameworks for AI-based medical devices. The US FDA has cleared over 900 AI/ML-enabled medical devices as of 2026, but most are designed for high-resource healthcare settings. WHO’s prequalification mechanism is specifically intended for products that need to function in settings with limited infrastructure — a different and arguably harder problem than AI tools designed for well-equipped hospitals.
The WHO’s full prequalification assessment for CAD4TB is available on the WHO prequalification database.
