The POCUS4TB project seeks to develop a simple, cheap imaging solution to screen for tuberculosis (TB). We will conduct clinical research to determine whether POCUS is an effective and practicable tool for TB screening, can be easily taught, and does not overburden health workers or patients.
We are a multi-disciplinary team of researchers from the UK, Bolivia, Mozambique, Peru, The Netherlands and Zambia aiming to develop and test an artificial intelligence-supported point-of-care ultrasound (AI-POCUS) to improve the reach and quality of tuberculosis screening in vulnerable adults and children.
Who we are
We are a multidisciplinary Group made up of the London School of Hygiene & Tropical Medicine (LSHTM) (UK), Centro de Investigação em Saúde de Manhiça (CISM) (Mozambique), Delft Imaging Systems (DIS) (The Netherlands), Imperial College London (UK), Asociación Benéfica PRISMA (Peru and Bolivia) and Zambart (Zambia). We also collaborate with colleagues from Community Initiative for Tuberculosis, HIV/AIDS and Malaria plus related diseases (CITAM+), Asociación de personas afectadas con tuberculosis de Santa Cruz (ASPACONT-SC), Barcelona Institute for Global Health (ISGlobal), Ministries of Health (MoH) and expert collaborators.
What we do
Can AI and ultrasound help find the missing cases of TB?
Tuberculosis (TB) is one of the world’s deadliest infectious diseases—and yet, in many places, it’s still flying under the radar. Too many people are diagnosed too late, or not at all, despite the knowledge that early TB detection not only improves and saves lives, but also stops the disease from spreading in the community. In fact, early diagnosis and treatment is a critical step if we want to meet the United Nations goal of ending the TB epidemic by 2030. But to do that, we need to get better—and faster—at finding people with TB. That’s why the World Health Organization (WHO) recommends to systematically screening the general population for TB in high-burden settings, in order to catch more cases and catch them earlier.
The problem: Great tools, but hard to access
One of the best screening tools we currently have is the chest X-ray (CXR). But here’s the downside: CXR machines are expensive, need infrastructure, and require trained radiologists to read the images—something many high-burden countries simply don’t have in every clinic or village.
This is where artificial intelligence (AI) has already stepped in. Several computer aided detection (CAD) software programmes designed to interpret CXR images for screening of TB have been developed, and what’s more, been shown to be as accurate as human readers. This has led them to be endorsed by WHO, with several CAD software products available on the market and being scaled up globally, filling the gap where radiologists are unavailable.
Still, the access barrier remains, as X-ray equipment may be financially and logistically unaffordable in remote or rural areas in high burden countries.
A new competitor: Ultrasound
Ultrasound is a familiar technology, widely used for procedures relating to pregnancy and many other medical conditions. It's portable, safe, and relatively affordable. Today, with just a handheld probe and a smartphone or tablet, clinicians can carry out point-of-care ultrasound (POCUS) in the most remote corners of the world.
And now, evidence is building that POCUS might help detect lung disease—including TB. Sounds promising, right? But there’s still one big challenge; interpreting ultrasound images isn’t easy, and trained experts aren’t always on hand in certain settings.
So, what if AI could help with that too?
AI + POCUS = A game-changer for TB
This is exactly what POCUS4TB, an ambitious, four-year research project, is aiming to solve. Funded by NIHR and led by LSHTM, POCUS4TB is bringing together experts from all over the world—academics, researchers, non-profits, an industry partner in the Netherlands and study sites in four high TB burden countries under a Global Health Research Group (GHRG).
Their goal? To develop a portable, AI-assisted point-of-care ultrasound (AI-POCUS) tool that can help screen TB anywhere—even in places with no X-ray machines and no radiologists.
Still many questions to answer
While the vision is exciting, there’s still a lot we don’t know. For example:
- What’s the best way to capture ultrasound images of the lungs for TB screening?
- Which parts of the lungs should be scanned?
- What specific findings in those images are the most predictive of TB?
- Do TB-related ultrasound features differ across populations—like between adults and children, or between people living with HIV and those who are not?
- Can ultrasound images be used also to assess TB severity or monitor treatment response?
To tackle these critical questions, POCUS4TB is launching a series of formative studies, with one in each participating country. Each study will focus on different research gaps, while collectively building a large, diverse library of lung ultrasound images. These images will be used to train the first version of the AI model, ensuring it reflects the wide range of real-world TB presentations across varied populations and settings.
What makes this project special?
POCUS4TB isn’t just about coming up with the tech side of things and hoping it works in the real world. From day one, the team has worked closely with local communities, health workers, and civil society groups to make sure the tool actually fits the needs of the people it’s meant to help.
The research is being carried out with multiple partners across many countries:
- Asociación Benéfica PRISMA (a non-governmental organisation in Peru and Bolivia)
- Centro de Investigação em Saúde de Manhiça (CISM, a research institution in Mozambique)
- Delft Imaging (a technological company in the Netherlands)
- Imperial College London
- Zambart (a research institution in Zambia)
Each country has its own unique TB context:
- Zambia: Adults, high TB/HIV co-infection rates
- Mozambique: Children, high TB/HIV prevalence
- Peru: Adults, high multidrug-resistant TB
- Bolivia: Adults, high TB burden
These diverse settings will help create a rich, varied database of ultrasound images to train the AI model—and sure it works well for different populations.
It’s about ethics, too
Building AI for healthcare is more than just code and algorithms. There are important questions around data privacy, ethics, and fairness. Who owns the data? How do we share it responsibly across countries? What does consent look like in low-resource settings?
POCUS4TB is taking these questions seriously. One part of the project is dedicated to exploring the ethical and cultural dimensions of using AI in healthcare. The team is also piloting something called federated machine learning, an approach that lets sites train the AI model locally while keeping control of their data—before combining these models into a centralized one, with no need to transfer sensitive patient data across borders.
What’s next?
After the formative phase, the final AI-powered ultrasound tool will be tested in the field, across all four countries, with both children and adults being evaluated for TB. Researchers will assess its clinical performance, usability, and cost-effectiveness.
If successful, this could be a game-changing tool for TB screening—especially in the most vulnerable and underserved communities. Follow the journey of POCUS4TB as we turn cutting-edge research into real-world impact—and bring high-quality TB screening to those who need it most.
Aim
To develop and test an AI-supported point-of-care ultrasound (AI-POCUS) to improve the reach and quality of tuberculosis screening in vulnerable adults and children through an international collaboration between researchers, developers, clinicians, communities affected by TB and policy makers.
Objectives
- To develop a network of multi disciplinary researchers, clinicians, communities affected by TB, industry partners and policy makers to co-design and evaluate AI-POCUS for TB screening in children and adults
- To evaluate and optimise different POCUS acquisition protocols in different epidemiological contexts
- To broaden understanding around the ethics of data sharing in AI and pilot a collaborative machine learning approach
- To evaluate the clinical performance (sensitivity, specificity and predictive value), acceptability and costs of implementing AI-POCUS
- To maximise dissemination of research findings and uptake of an AI-POCUS
- To build capacity for inter-disciplinary research in four low and middle-income countries (LMIC)
This research aims to develop a simple, cheap imaging solution to screen for tuberculosis (TB), the world’s leading infectious disease killer. The World Health Organisation has set a target to "End TB” by 2035, and developed a strategy to help countries achieve this. One of the key activities in this strategy is to find and treat more people with TB. Current screening guidelines include chest x-ray, but this is inaccessible to many of the poorest people worldwide who are most at need of screening for TB. Ultrasound may provide a solution. Ultrasound is a simple and safe procedure that can be used for adults and children. Portable ultrasound probes (“Point-of-care” ultrasound, POCUS) that can connect to mobile phones are available and could be placed at the lowest levels of the health care system in countries most affected by TB. Initial research is promising, but highly trained health workers able to interpret ultrasound images are not widely available. Therefore, artificial intelligence (AI) could provide a solution.
This Global Health Research Group (GHRG) will bring together researchers from four LMIC with different TB epidemiology contexts; Mozambique and Zambia (high burden of TB/HIV), Bolivia and Peru (high burden of MDR-TB), alongside a medical imaging partner to develop an AI-supported POCUS tool.
The project will start with co-design of the AI-POCUS with communities affected by TB, health care providers, and regulators in these settings to understand what they would like from such a product. We will conduct clinical research to determine whether POCUS is an effective and practicable tool for TB screening including which scanning protocol provides the best images, yet can be easily taught and does not overburden health workers or patients, and to start collecting clinical, microbiological and imaging data.
To develop the AI reading, large numbers of images and data are needed, however there can be challenges with data privacy and ownership if data is transferred to a central source. The GHRG will therefore conduct research around the broader ethics of data privacy and transfer, and develop capacity to pilot a “federated machine learning approach”, which involves all sites retaining full ownership of their data and concurrently training a general machine learning model on-site, before combining these models into a centralised one.
The finalised AI-supported POCUS product will then be tested across the four epidemiologically different sites in adults and children being investigated for TB. We will assess the clinical performance and the costs of the system.
The GHRG will finally convene a workshop with communities affected by TB and policy makers to present the findings and discuss the roadmap to the further development of the product. We will also produce a training package for health workers to facilitate adoption beyond the facilities and countries in this GHRG. If successful, this product has the potential to be a game-changer for screening of TB in high-burden countries by dramatically increasing access to better quality and co-ordinated TB diagnosis amongst the world’s population that are both the most vulnerable to TB and face the greatest barriers to being diagnosed.
Work packages
- WP1 - Workshop consortium
The first activity of the consortium will be a workshop of all consortium members, who will then be joined by other researchers, communities affected by TB, plus international and national policy makers to discuss the current knowledge base around POCUS for TB screening and ensure research and policy gaps are identified. Throughout the project coordination between the different work packages and across the sites will be achieved by regular virtual group meetings and cross-site interaction. WP1 will ensure high standards of research governance are maintained including ethical and regulatory compliance and monitor progress against milestones.
- WP2 - Formative research on POCUS
The first study will be qualitative research with TB patients Health Care Workers (HCW), and regulators to understand how POCUS tools could best be deployed, including the optimal user and patient experience. Clinical research with both adults and children will identify optimal POCUS acquisition protocols, including combining lung, FASH and mediastinal views, to obtain and compare images with clinical data and microbiological results. Several acquisition protocols have been proposed but a balance between image quality, training complexity, and HCW and patient time needs to be evaluated. The research will identify whether protocols should differ in children and adults, and based on HIV or other clinical factors.
- WP3 - Artificial Intelligence
Machine learning, and in particular deep learning, have become a major trend in medical image analysis. In order to have good clinical applicability large medical datasets derived from various sites are needed. However, the current centralised learning techniques raise concerns about patient privacy, confidentiality, security, and data ownership. Even in anonymized datasets, an individual could be re-identified correctly with a high likelihood. Prof. Bond, a social anthropologist, will supervise a PhD student to conduct research around the broader ethics of data ownership and privacy in medical research. Ir. Swinkels, a deep learning engineer, will lead a pilot of federated learning, a potential solution to develop the AI model. Federated learning may solve data privacy and ownership issues by training a consensus model collaboratively between different decentralized datasets. The sites will be capacitated to manage their own data to train the model locally. Thereafter, updated model parameters will be sent to a central server that aggregates the separately trained local models to generate a new global model. This new model could be personalized for each client to include data heterogeneity.
- WP4 - Evaluation of clinical performance, acceptability and economic effectiveness of AI-POCUS
Data from WP2 and WP3 will be used to develop an AI-POCUS product that will be evaluated in the different epidemiological situations provided by the sites (TBHIV adults Zambia, children Mozambique, non-HIV TB adults Peru, children Bolivia). WP4 will develop unified tools and standardised outcomes for the protocols across the sites to enable evaluation within and across countries. Clinical performance will be assessed by measuring the sensitivity, specificity and predictive value of the AI-POCUS compared to standardised endpoints, in the hands of intended users. Qualitative research will assess usability and acceptability, with an economic evaluation to assess costs.
- WP5 - Dissemination and uptake
To ensure uptake of the proposed AI-POCUS, tools and training materials will be needed for HCW to ensure that the system can be deployed. We will work with communities and HCW to co-develop a training package based on the final acquisition protocol. WP5 will engage with regulators and policy makers to communicate findings and understand pathways to inclusion of the tool in screening guidelines.
- WP6 - Research capacity development
Underpinning the research activities of the Global Health Research Group (GHRG) will be development of research capacity including one PhD in each country engaged to conduct different research studies. The PhDs will include one social scientist to look at broader issues of ethics in AI, a clinician who will focus on the evidence needed for the AI-POCUS model, a computer scientist and an implementation researcher. Institutional capacity development will include development of infrastructure and data skills for federated learning as well as broader research and finance governance tasks.
E. Swinkels,1 C. Bresges,2 K. Mateyo,3,4,5 M. Fentress,6 S. Lungu,7 R. Chimfwembe,7 K. Shanaube,3 H. Ayles,3,5
1Eindhoven University of Technology, Eindhoven, Netherlands, 2University Hospitals Sussex NHS Trust, Brighton, United Kingdom of Great Britain and Northern Ireland, 3Zambart, Lusaka, Zambia, 4University Teaching Hospital, Lusaka, Zambia, 5London School of Hygiene and Tropical Medicine, London, United Kingdom of Great Britain and Northern Ireland, 6University of California, San Francisco, United States of America, 7Kanyama General Hospital, Lusaka, Zambia
56th Union World Conference on Lung Health, Copenhagen, 2025. Int J Tuberc Lung Dis 2025: 29 (11 Suppl 1): S1 – S963, TBSCIENCE 2025 Oral Abstracts S791
Zambart’s POCUS4TB team held a knowledge-sharing session aimed at simplifying and making practical the concepts of artificial intelligence in healthcare, particularly for TB screening. It focused on sharing key lessons from an AI and digital transformation course, demonstrating how AI can support ongoing health research, and highlighting its challenges and opportunities. The session also encouraged open discussion on how these technologies could be applied within current projects, helping staff better understand and explore the role of AI in improving their work.
Several investigators from across the POCUS4TB project sites recently attended a short course at the University of Cambridge titled “How are Big Data, AI, and Digital Transformation shaping the future of healthcare?”, held from 15–17 December at Madingley Hall, Cambridge. The course was delivered by Dr Fatemeh Torabi, Assistant Professor of Health Data Science at the University of Cambridge.
The course explored how big data and artificial intelligence are transforming healthcare systems and research, with a strong emphasis on integrating patient data into data-driven decision-making. It introduced participants to the fundamentals of health data science and recent advances in the field, as well as core concepts of artificial intelligence and digital transformation in healthcare. Key discussions focused on the distinctive governance frameworks required to protect individuals’ health records while enabling research, alongside ethical considerations such as data privacy, bias, transparency, and trust.
Through a series of sessions covering AI in health data research, digital transformation of health systems, operationalising data-driven approaches, and evaluating AI-based decision tools, the course provided practical examples of real-world implementation scenarios. It also equipped participants with the tools, techniques, and terminology needed to assess the robustness and trustworthiness of data-driven and AI-enabled tools in healthcare settings, while considering future challenges and opportunities in this rapidly evolving field.
Attendance at the course was supported by the NIHR Global Health Research Cohort Academic Development Award scheme (GHR CADA), which aims to strengthen research capacity and skills development among global health researchers. Participation in this training has enhanced investigators’ understanding of responsible and effective use of AI in healthcare and will inform the project’s ongoing work in data-driven health innovation.
Elze Swinkels delivering an oral presentation at the 56th Union World Conference on Lung Health, held in Copenhagen in 2025. Her session showcased emerging evidence on the potential of handheld, AI-supported lung ultrasound as a scalable tool to improve tuberculosis (TB) screening in community and primary care settings.
The POCUS4TB team made use of a spare afternoon during the 56th World Conference on TB and Lung Health to engage in structured discussions on prospective PhD research projects. These conversations provided a valuable opportunity to align scientific priorities, strengthen collaborative efforts within the programme and gather constructive feedback from the wider study team.
Zambart, has taken a significant step forward in the fight against tuberculosis (TB) with the launch of The Lung Ultrasound for Tuberculosis Diagnosis: An Evaluation of Image-Acquisition Protocols and Probes (POCUS4TB). This groundbreaking initiative, funded by the National Institute for Health Research (NIHR) Global Health Research Groups, aims to improve early and accurate TB diagnosis using point-of-care lung ultrasound in high-burden settings.
The study is being implemented in collaboration with Kanyama Level One Hospital and is part of a global consortium working together to advance TB diagnostics. The POCUS4TB study seeks to evaluate the diagnostic accuracy of lung ultrasound, focusing on different acquisition protocols and probes.
@Ministry of Health, Permanent Secretary, for Technical Services @Dr. Kennedy Lishimpi, commended Zambart for this on point-of-care lung ultrasound technology research, which will incorporate AI-assisted diagnostics, stating that it aligns with the Ministry ‘s broader strategy to leverage technological advancements for more efficient and accessible healthcare delivery.
“We applaud the efforts of Zambart and Kanyama Level One Hospital, and their partners involved in this initiative. Their collective expertise and dedication will ensure that research translates into impactful health interventions”.
He added that strengthening collaboration across government agencies, health institutions, research bodies, and community stakeholders is essential for achieving sustainable health improvements.
He further stated that beyond the scope of the study, the Ministry remains committed to integrating innovative diagnostics within national TB programs, improving health infrastructure, and training healthcare personnel to effectively implement new technologies.
“This approach is aligned with our national TB elimination strategy, which emphasizes early detection, targeted treatment, and improved patient outcomes” he said.
Dr. Lishimpi, further reaffirmed the Ministry’s unwavering support for efforts that drive public health progress, enhance research-driven healthcare solutions, and, most importantly, improve the lives of all Zambians.
And Zambart Executive Director, Dr. Kwame Shanaube, said Tuberculosis remains one of the most pressing public health challenges, affecting millions worldwide, adding that the POCUS4TB study does not only have the potential to transform TB diagnostics in Zambia and beyond, but also marks a significant step towards reducing TB-related morbidity and mortality.
“Zambart's commitment to finding innovative solutions for TB diagnosis and treatment is demonstrated through its ongoing research studies, including:
- “Ubumi Bwandi” Integrated wellness screening and prevention in Zambia
- TB HEART (The burden of cardiac disease in TB infection in an HIV-prevalent setting)
- EPiTB (Addressing an unmet need: same-day diagnosis of extra-pulmonary TB in a high-burden setting)
- CommuniTB (GATES MRI) vaccine study
- XACT – 19 - Evaluating the impact of computer-assisted x-ray diagnosis and other triage tools
- Yaba Guy Che (For the men): Implementation and evaluation of a co-developed multi-disease intervention for men
- ZAMSA - Advancing epidemiological methods for surveillance of Mycobacterium Tuberculosis infection.
And Dr. Kondwelani Mateyo, CREATE PhD Fellow at Zambart and Consultant Pulmonologist at the University Teaching Hospital (UTH), said that the findings will help determine whether lung ultrasounds with AI integration can be incorporated into Zambia's health systems.
The study will enroll 2,700 adult participants, at Kanyama Level One Hospital which has been selected as the study site in Zambia, representing one of the country’s highest TB burden areas. The study is being conducted in collaboration with partners from Bolivia, Mozambique, Peru, and the UK.
@LSHTM @Delft imagining @CISM @Prisma @MOH
#Pocus4TB #Zambart #TogetherWeCanEndTB.
Point-of-Care ultrasound (POCUS) training for the Peru study team at the Hospital de Ventanilla, Municipalidad de Ventanilla, Callao (May 2025)
Point-of-Care ultrasound (POCUS) training for the Bolivia study team at the Centro de Salud 18 de Marzo, Santa Cruz de la Sierra (May 2025)
The POCUSTB kick-off meeting took place in Zambia from the 21st to the 25th of October, 2024. Partners of the POCUS4TB project from Peru, Bolivia, Mozambique and Zambia, as well as from DELFT attended the meeting which was hosted by Zambart, at Chaminuka Lodge in Lusaka. During the three-day meeting various aspects of the POCUS4TB project were covered, from specific Work Package-related (WP) activities and procedures to broader overarching topics on how to best work together as a group, such as safeguarding and equity.
Specific WP sessions were led by Dr Maria Ruperez (WP1 Coordination), Dr Kondwelani Mateyo and Dr Matthew Fentress (WP2 Formative Research) and by Elze Swinkels and Professor Ginny Bond (WP3 Artificial Intelligence). PhD candidates Agostinho Lima and Luz Quevedo also presented and discussed their plans for their PhD’s with the group. Community representatives from Bolivia, Peru and Zambia held several meetings focusing on the community led work in the POCUS4TB project. A specific orientation session on safeguarding was led by Dr Musonda Simwinga, in which several case studies were discussed in groups. Dr Ginny Bond also presented a session on equitable partnerships where she introduced the Equipar tool to the partners – a tool to help support equiptable partnerships for research projects - and further discussions about the use of this tool were then held in small break-out groups. After the three-day meeting, several activities were organised by Zambart, including a group visit to Kanyama community, one of the Zambian study sites in the centre of Lusaka, and the Zambart Office.
The LSHTM team visited our partners in Peru and Bolivia in September with the objective of gaining a better understanding of the sites, getting to know the in-country teams and planning the work ahead.
Prof Helen Ayles, Alex Miller and Dr Maria Ruperez from LSHTM visited the PRISMA offices in Lima to meet many of the PRISMA team face-to-face. Several meetings took place but in particular we met with the CEO of Prisma, Dr Marilu Chiang, and their administrative and financial teams, to discuss the practical side of implementing the POCUS4TB Project in both Peru and Bolivia.
The LSHTM team also visited the Peru research team, who are based at the Cayetano Heredia University, to discuss plans for the research with Luz Quevedo, the POCUS4TB PhD candidate from Peru. We were also give an tour of one of the health facilities in Lima (Health Center Peru-Korea Bellavista-Callao) one of the proposed study sites, and the met with the TB Programme Team at the Ministry of Health Regional Health Department office (DIRESA) in the Callao district. We also had a meeting with one of the community representatives, to help improve our understanding of the role and priorities of patients and communities within the TB landscape in Peru, and the best way for us to incorporate them in the POCUS4TB project.
During this trip, the LSHTM team also visited Santa Cruz, in Bolivia, where the PRISMA Bolivia team are based. In Santa Cruz they met the project lead their, Dr Freddy Tinajeros, the PhD candidate for Bolivia, as well as other laboratory and administrative members of the team.
The LSHTM team also visited the “18 de Marzo” Health Center and met the TB services team who showed them around, introduced them to the staff, and explained the TB-related activities that take place in the center. A call was also conducted with the Director of the TB programme in Santa Cruz where the challenges and priorities of the TB programme were shared and discussed.
Finally, the team also met with a community representative, who is member of ASPACONT, the Bolivian Association for People Affected by TB, and also the Observatorio de TB which is a group of civil society organisations who work to influence policy to improve the rights and health care of TB affected populations.
