AI ultrasound tool counters TB pandemic in Africa

The tool, named ULTR-AI, connects to a smartphone and utilises artificial intelligence to interpret lung ultrasounds. It identifies signs of tuberculosis without requiring chest X-rays, which are often expensive or unavailable in remote locations.

By eliminating the need for trained radiologists and large equipment, ULTR-AI facilitates quicker TB detection and aids in diagnosing conditions such as pneumonia. Its portability improves access in remote areas and complements telemedicine efforts in under-resourced or climate-vulnerable regions.

The $11 million grant for this project was awarded by the Global Health European and Developing Countries Clinical Trials Partnership 3 Joint Undertaking, a European Union-funded programme. The initiative, called Computer Assisted Diagnosis Lung Ultrasound for Tuberculosis (CAD LUS4TB), will run for five years and aims to generate clinical data to refine the artificial intelligence model through open-access training.

ULTR-AI was developed by the Swiss Federal Institute of Technology in Lausanne and Lausanne University Hospital. The Swiss teams behind the project highlight that it addresses a critical diagnostic gap: interpreting ultrasound images typically requires advanced training, which many frontline workers do not have. The tool simplifies this process through automated analysis. The CAD LUS4TB consortium comprises ten institutions across Africa and Europe, with key partners including Stellenbosch University, Carnegie Mellon Africa, the Foundation for Innovative New Diagnostics and the Swiss Tropical and Public Health Institute.

“This project was designed for field conditions,” said Dr Mary-Anne Hartley of the Swiss Federal Institute of Technology in Lausanne. Professor Noémie Boillat-Blanco from Lausanne University Hospital noted that the lack of access to ultrasound specialists remains a significant barrier. Dr Véronique Suttels added that the tool enables earlier TB detection, reducing treatment costs and transmission in high-burden settings.

ULTR-AI’s portability helps address the healthcare worker shortage by allowing non-specialist staff to perform diagnostic tasks typically handled by radiologists. In areas where access to imaging infrastructure is limited, the tool facilitates earlier identification of tuberculosis and other respiratory illnesses. Its smartphone-based design makes it suitable for low-resource or mobile clinics, where personnel turnover is high and the need for simple, effective tools is essential for sustaining patient services.

The device’s low power requirements tackle the electricity challenges common in rural and off-grid areas. Its portability enhances patient outreach by bringing diagnostics closer to communities, thereby reducing the need for travel to central facilities. ULTR-AI also enables more consistent data collection at the point of care, improving surveillance and case tracking. Together, these features strengthen health systems and provide actionable insights for both tuberculosis control and broader primary healthcare delivery.

Climate change threatens to accelerate the tuberculosis pandemic by driving both transmission and vulnerability. Extreme weather events, such as floods and cyclones, can displace individuals into overcrowded shelters, thereby increasing the risk of TB spread.

Additionally, these events often disrupt access to healthcare services, including those for HIV, diabetes and TB preventive therapy. A critical pathway through which climate change impacts TB is by worsening food security; undernutrition remains the leading risk factor for TB in many African countries.

This situation increases susceptibility to the disease, heightens its severity and raises mortality rates. As climate change continues to progress rapidly and unpredictably, these combined pressures could significantly undermine global efforts to eliminate TB in already fragile health systems.