Makerere AI Health Lab develops AI-powered tools, datasets, and research partnerships that support diagnosis, disease surveillance, and health systems strengthening in low-resource settings.
6
Research projects
Diagnostics, maternal health, surveillance and environmental health
3
Dataset initiatives
Open, locally-curated health data for African contexts
22
Researchers & staff
Scientists, engineers, clinicians, laboratory and field team
2
Countries in data collection
Blood smear datasets curated across Uganda and Ghana
The Makerere AI Health Lab sits within the College of Computing and Information Sciences at Makerere University. We work at the point where machine learning meets frontline healthcare — building diagnostic tools, curating the datasets those tools depend on, and validating both in the health facilities where they will actually be used.
Our work is grounded in the realities of low-resource settings: limited specialist availability, intermittent connectivity, and equipment that has to work in a district health centre, not only in a teaching hospital.
From malaria microscopy to maternal health screening, our projects pair machine learning with the clinical and laboratory workflows they have to fit into.
Ocular's 3D printable adapter connects smartphones to microscopes, allowing for easy image capture and alignment in health centers. This technology enhances pathogen identification for diseases like malaria and tuberculosis, supported by a mobile and web application for diagnostic assistance.
We are developing a rapid, low-cost screening system for malaria diagnosis using a 3D printed smartphone adapter for microscopes and deep learning models, aimed at enhancing real-time diagnostics and surveillance in Uganda to improve public health outcomes.
The Lacuna Fund project aims to enhance malaria microscopy diagnosis in low and middle-income settings by generating high-quality, open-labeled blood smear datasets from Uganda and Ghana, using smartphone-mounted microscopes to support machine learning models for detection.
This project aims to enhance cervical cancer detection in Sub-Saharan Africa by curating a labeled digital dataset of 1,000 Pap smear images using a smartphone-mounted microscope, supporting machine learning tasks for early and reliable screening.
We propose an AI-powered ultrasound screening tool to detect high-risk pregnancies early, aiming to reduce maternal mortality in Uganda and similar settings. The solution will integrate clinical, demographic, and imaging data to support timely and accurate diagnosis at the primary healthcare level.
This project aims to support natural resource management in Uganda by using AI to analyze satellite and forestry inventory data for predicting land cover changes and improving decision-making in response to deforestation and climate change.
Automated microscopy and image analysis for malaria, tuberculosis and cervical cancer screening.
Point-of-care risk assessment from 2D ultrasound to identify high-risk pregnancies earlier.
Curating open, labelled, locally-owned datasets so African health data trains African models.
Privacy, dataset governance, bias and fairness, and human-in-the-loop clinical decision-making.
Using diagnostic data to support real-time disease surveillance and public health response.
Designing tools around existing clinical workflows, staffing realities and infrastructure.
Models are only as good as the data behind them. A significant part of our work is curating high-quality, labelled, openly-licensed health datasets from African clinical settings.
Computer scientists, engineers, clinicians, laboratory technicians and students working together across research, development and deployment.
We collaborate with universities, hospitals, ministries, funders and technology partners on research, dataset development, clinical validation and deployment.
