Unitaid and the Clinton Health Access Initiative (CHAI) have secured agreements with AI screening providers DeepTek and Delft Imaging that will cut the cost of computer-aided detection (CAD) software for tuberculosis programs by up to 65% across 138 low- and middle-income countries. The new pricing model, effective immediately, replaces upfront costs that could reach $12,000 per X-ray unit per year with a single annual fee of $2,500 to $2,850 that covers unlimited scans, software, installation, training, and maintenance.
Tuberculosis remains the world's deadliest infectious disease, and chest X-rays are a critical screening tool. But in countries with the highest TB burden, radiologists are scarce - sometimes as few as one per million people. Without specialists to interpret images, screening programs stall and people can go undetected for months, spreading the disease through their communities.
How AI removes the bottleneck
CAD software uses artificial intelligence to read chest X-ray images and flag signs of TB, performing on par with a trained radiologist. A chest X-ray taken in a village or mobile clinic can be analyzed instantly, giving people screening results the same day. This addresses a core problem: in many settings, images simply go unread due to the lack of qualified personnel.
Until now, programs faced tens of thousands of dollars in upfront costs for software licenses, computing hardware, installation, training, and warranty extensions. Recent donor funding disruptions and the expiry of previously purchased access licenses made those costs unsustainable. Many programs that scaled screening in recent years now face renewal prices their domestic budgets cannot absorb.
What the new pricing model covers
The all-inclusive annual fee of $2,500 to $2,850 per connected X-ray unit covers unlimited CAD scans, software updates, installation, training, and maintenance with no hidden add-on costs. Programs can also choose a three-year term at $7,500 per unit for greater procurement flexibility. The agreements set transparent pricing for additional indications the same software can detect, including cardiomegaly, silicosis, and lung cancer, broadening CAD's use as a general chest X-ray interpretation tool.
Combined with price reductions on portable X-ray devices negotiated by Unitaid and CHAI earlier in 2025, programs can now deploy a complete AI-powered TB screening solution - X-ray hardware and CAD software together - at a fraction of the previous cost.
"TB is curable, but only if you catch it. Today, millions of people go undiagnosed simply because there is no expert around to read their X-ray," said Dr. Neil Buddy Shah, CEO of CHAI. "AI-powered software can now do that in seconds, and this agreement makes it far more affordable, dramatically increasing the number of people who can be diagnosed and put on treatment."
Dr. Philippe Duneton, Executive Director at Unitaid, said, "By reducing the cost of AI interpretation software by up to 65%, these agreements demonstrate how market shaping can make proven innovations more accessible. At a time when funding constraints are threatening progress in TB detection, lowering the cost of these tools is critical to helping countries expand access to timely TB diagnosis and care, and make a difference for millions of people."
Scope and product availability
The agreements cover public sector and not-for-profit buyers across all 138 World Bank-defined low-, lower-middle-, and upper-middle-income countries. Both DeepTek's Genki and Delft Imaging's CAD4TB products are available with immediate effect directly from the suppliers. The World Health Organization recommends CAD for TB screening and triage in individuals 15 and older, and both products meet WHO performance standards. Neither agreement restricts use in age groups where products gain regulatory approval, and both suppliers have committed to future pediatric access at no additional cost.
For government procurement teams and health ministries, the shift from variable upfront licensing to a predictable annual fee removes a major budget planning obstacle. The model aligns with the growing use of AI for Government applications in public health, where transparent pricing and scalability are essential for national deployment.
Why this matters for healthcare and development professionals
This pricing shift changes the economics of TB screening at scale. A program that previously budgeted $70,000 to $120,000 annually for CAD software across ten X-ray units can now cover the same capacity for $25,000 to $28,500. For professionals in global health, radiology, and development, the agreement removes a cost barrier that has limited the reach of AI for Healthcare in the settings where it is needed most. The explicit inclusion of additional indications - heart, lung, and occupational disease screening - also means the same software investment can support broader diagnostic programs beyond TB alone.
Your membership also unlocks:
