DIDIDI

The challenge

Key neglected tropical diseases, soil-transmitted helminths (parasitic worms) and schistosomes (flukes) affect around 1.5 billion people, most of them in Africa. They are targeted for elimination, with control relying mainly on mass drug administration campaigns.

Monitoring progress towards elimination is essential but challenging. In addition to the presence of parasites, it is important to track the intensity of infection, or individual worm burdens. This generally relies on manual microscopic analysis of stool samples to quantify the number of parasite eggs present. This is labour-intensive and time-consuming. 

The project

The DIDIDI project is assessing whether automation and AI technologies can accelerate the analysis of stool samples to characterise parasitic worm infections. 

A prototype system has been developed based on a very low-cost (below €220) 3D-printed microscope that can be transported in a backpack to field locations. It has a modified stage that automatically scans stool samples and is controlled by a simple mobile phone app, which also enables AI software to interpret the image and provide a numerical count of eggs, distinguishing between those from five different types of parasitic worms and schistosomes, in faecal samples.

The DIDIDI project is working with the microscope manufacturer to refine its design to make it more robust for use in African field settings. In addition, automatic scanning will be extended to the third dimension so the full volume of stool samples can be analysed. The accompanying app interface and decision-support module will be co-designed with users. 

The new version will then be evaluated at six schools in Kenya and Uganda, in areas where parasitic worms and schistosome infections are common. The aim is for each microscope to process 100 samples a day, or 400 a week (with a day for associated support and transport activities). Results from several microscopes may be processed by networking information through a single mobile phone. Findings will be compared with the results of parallel testing using gold-standard microscopy techniques analysed by local health workers, as well as by lateral flow assay test results and molecular testing.

As well as testing individuals, the project is exploring the use of community-wide sampling to track parasite burdens. Latrine pits will be tested to determine whether the information on village-level parasite prevalence over the period of mass drug administration is useful to policymakers.

Individuals’ and community results will be delivered directly to national electronic medical records and to standard health information systems in both Uganda and Kenya, with digests presented to policymakers via a digital dashboard. Health economic analyses will be carried out to provide insights into the cost-effectiveness and budget impact of both individual-level and community-focused approaches. 

Finally, the project is also collating local meteorological data from a range of sources to determine whether there are correlations between meteorological determinants and parasite prevalence. This could provide valuable insights into how disease burdens might be affected by recent changes in weather patterns, associated with climate change, which could help to optimise community treatment campaigns.

Impact

The DIDIDI project could support the more effective control of key neglected tropical diseases in sub-Saharan Africa. It will: 

• Show whether a new portable, higher-throughput microscopy technology can reliably determine the intensity of infection of schistosomes and soil-transmitted helminths in individuals or communities, in low-resource settings.
• Provide data that could both support the treatment of individuals and guide local and national control efforts.
• Generate insights into how worm infections are being impacted by changes in meteorological patterns in East Africa. 

By refining microscopy tools and integrating AI image analysis, the DIDIDI project could greatly increase the amount of information on parasitic worm burdens available to policymakers, helping them to assess the effectiveness of control efforts and prioritise areas for more intensive intervention.