AI and blood test combination could transform TB diagnosis in Africa
The three-year AddiCAD project, officially launched in May 2026, is funded by the Global Health European and Developing Countries Clinical Trials Partnership 3 (Global Health EDCTP3). The initiative seeks to develop and validate a non-sputum-based diagnostic approach that could make TB detection faster, more accurate and more accessible in resource-limited settings.
TB remains a major global health challenge. Of an estimated 10.7 million new cases each year, approximately 2.5 million people remain undiagnosed. Limited access to affordable diagnostic services, laboratory infrastructure and suitable sputum samples continues to contribute to delays in detection and treatment.
Combining AI and biomarker testing
AddiCAD combines CAD4TB, an AI-based system that analyses digital chest X-rays for signs of TB, with a biomarker test that measures the body's immune response to infection.
Preliminary findings indicate that the combined approach improved specificity by 20% compared with CAD4TB alone, without compromising sensitivity. If confirmed through clinical validation, the technology could reduce false-positive results while continuing to identify people who are likely to have TB. The approach is designed to complement, rather than simply replace, existing diagnostic pathways. By rapidly identifying individuals who are most likely to have TB, healthcare workers could prioritise those requiring confirmatory testing and treatment.
Validation across three African countries
The AddiCAD consortium will develop a novel biosensor and companion mobile application before conducting a clinical study involving approximately 1,000 adults with presumptive TB in South Africa, Namibia and The Gambia.
The project brings together six partners: Delft Imaging Systems in the Netherlands, Life SADX in South Africa, LINQ Management GmbH in Germany, the London School of Hygiene & Tropical Medicine in the United Kingdom and The Gambia, Stellenbosch University and the University of Namibia.
The consortium will also engage healthcare providers, patient representatives, regulators and commercial partners to assess how the technology could be implemented and scaled beyond the research setting.
Prof Stephanus Malherbe, SU associate professor of immunology and AddiCAD project coordinator, said the technology could help address the persistent gap between the availability of effective TB treatment and timely diagnosis.
“If the initial findings are validated, AddiCAD could provide healthcare workers with a practical tool for identifying TB more efficiently, particularly in communities where conventional diagnostic services are difficult to access,” he said.
AddiCAD is one of two Global Health EDCTP3-funded projects currently coordinated by Stellenbosch University to advance TB diagnostics. The second, PRECISE-TBM, focuses on improving the diagnosis of childhood tuberculous meningitis.
Together, the projects reflect a growing effort to develop diagnostic technologies that can deliver faster and more accessible healthcare solutions in high-burden and resource-limited settings.