A new study has used artificial intelligence (AI) modelling to effectively identify areas with a high burden of tuberculosis in Bangui, Central African Republic, that were previously unknown – an approach that could be scaled and used in other low-resource settings.
Tuberculosis (TB) is one of the oldest known infectious diseases, yet it remains a leading cause of death in many low-income countries. Now, in a pioneering collaboration between the International Union Against Tuberculosis and Lung Disease (The Union), EPCON, and the National Tuberculosis Programme (NTP) of the Central African Republic, artificial intelligence is being used to confront this ancient disease with modern precision.
To support the NTP’s goal of reducing missed cases and expanding access to care, The Union collaborated with EPCON’s AI specialists to develop a high-resolution predictive model. The team created a digital map of the city, dividing it into 100 by 100 metre grids. They trained the AI model to predict TB positivity rates in each area by using publicly available data, such as population density, access to clean water, and distance from TB clinics.
The model helped identify TB hotspots that would otherwise remain undetected enabling more targeted interventions. Some neighbourhoods with high predicted risk had not registered many cases at all, suggesting under-diagnosis and gaps in access to testing and treatment.
In Bangui, the capital city of the Central African Republic (CAR), TB continues to present a major public health challenge. Traditional surveillance methods often struggle to keep up with the true spread of the disease, particularly in underserved communities.
Dr Kobto Ghislain Koura, lead author of the study and Director of the TB Department at The Union said: “For too long, communities have suffered because we’ve been reacting to TB rather than anticipating it. AI allows us to shift from passive surveillance to proactive intervention.”
The paper was recently published in the Tropical Medicine and Infectious Disease journal.
Dr Gando, CAR NTP Manager said: “It's fantastic that the Central African Republic has been able to explore the potential of AI in public health through this innovative study.
“The AI-generated mapping of high-burden TB areas has allowed us to strengthen active TB screening activities in these areas and raise awareness among local communities. All of this has improved the identification of people with TB.”
Crucially, the model does not rely on personal health records or high-tech infrastructure. It works with the data available and adapts where data is limited making it well-suited to settings like Bangui, and potentially to other cities facing similar challenges.
Dr Koura added: "This tool has the potential to reshape how we approach TB surveillance in countries where the disease is most prevalent. It helps us go beyond what is visible and detect where the disease may be silently spreading.
“This isn’t about replacing existing systems, the goal is to give TB programmes a sharper lens to see what might otherwise be missed, so that resources can be used where they’re needed most. It also opens the door for future innovation.”
Table 2. Comparison of predicted and notified TB positivity in the vicinity of TB clinics.
| TB Clinic | Predicted TB Positivity | Notified TB Positivity |
|---|---|---|
| Petevo Centre de Santé | Medium | Low |
| Lakounanga Urbain Centre de Santé | High | High |
| Centre de Santé Saint Joseph | Low | No data available |
| Complexe pédiatrique | Medium | No data available |
| CNRISTAR CTA | Low | No data available |
| Castors CSU | High | High |
| CNHUB HN | Medium | Low |
| Mamadou Mbaiki Centre de Santé | High | High |
| Hospital Communautaire | Low | High |
| Obrou Fidel Camp Centre de Santé | Medium | High |
| Amis Afrique ONG | Medium | Low |
| Malimaka | Medium | Medium |
| Hospital Amite | Medium | Medium |
| Bédé Combattant CSU | Medium | High |
The table shows seven discrepancies between predicted and notified TB positivity rates across 14 health facilities in Bangui. While some clinics, such as Castors CSU and Mamadou Mbaiki Centre de Santé, showed alignment between predicted and reported data, others including Hospital Communautaire and Petevo Centre de Santé, had lower notified rates than predicted. These differences may indicate underdiagnosis, limited access to services, or inconsistencies in reporting that obscure the true burden of TB in certain areas.
Dr Koura explained: “These mismatches could stem from a range of factors. Some clinics may be under-reporting or facing delays in case notification, while in other cases, people with TB may bypass nearby facilities due to stigma, limited services, or poor perceptions of care quality. Differences in diagnostic capacity between clinics and unaccounted local factors, such as informal settlements or recent displacements, may also contribute. Together, these issues highlight both the limitations of passive case detection and the potential for AI models to uncover hidden risks that traditional systems may miss.”
The Union and its partners plan to work closely with the Ministry of Health in CAR to validate and refine the model through real-time data. The goal is to embed this technology into national strategies and expand its use across other high-burden regions.
Dr Cassandra Kelly-Cirino, Executive Director of The Union, commented: “By integrating innovative technologies into national TB strategies, this approach demonstrates how digital tools can support equitable, data-driven public health while ensuring that even the most entrenched diseases are not beyond the reach of modern solutions.”
ENDS
Notes to editor
About The Union
Established in 1920 as the world’s first global health organisation, the International Union Against Tuberculosis and Lung Disease (The Union) is committed to a healthier world for all, free of tuberculosis and lung disease. Its members, staff, and consultants work in more than 140 countries globally.
The Union is a global membership, technical and scientific organisation, striving to end suffering due to tuberculosis and lung diseases by advancing better prevention and care. It seeks to achieve this by the generation, dissemination, and implementation of knowledge into policy and practice.
The Union’s approach to tackling global health problems is unique – KNOW. SHARE. ACT. We start with developing knowledge through global research (KNOW), which we then share as widely as possible (SHARE) and turn that into the real action to save lives on a local level (ACT).
The Union aims to ensure that no one is left behind, people are treated equally and we have a focus on vulnerable and marginalised populations and communities.
The Union’s work is exemplified by its core values of quality, transparency, accountability, respect, and independence.
Website: www.theunion.org
Social Media:
LinkedIn: International Union Against Tuberculosis and Lung Disease (The Union)
Bluesky: @theunion-tblh.bsky.social
Twitter: @TheUnion_TBLH
Facebook: The Union Lung Health
Instagram: @theunion_tblh