Human-centred AI for Disease Modelling and Policy
Human-centred AI for Disease Modelling and Policy
Informing decision-making around plans for the elimination of African sleeping sickness with data, visualisation and AI
For details, contact:
, University of Warwick
HAT MEPP Lead: Prof Kat RockLink opens in a new window
Scientific Project Manager: Dr Emily Crowley emily.crowley@warwick.ac.uk
led by researchers at the Zeeman Institute at Warwick and supported by the Bill & Melinda Gates Foundation. The project currently provides decision support to national programmes in Democratic Republic of Congo, Chad, Cote d'Ivoire, Guinea and Uganda regarding intervention strategies and cost effectiveness, taking into account resource constraints. In this third phase of the initiative, our plan is to extend the support to a range of new geographies.
Within CIM, our focus is on designing and developing a next-generation, data and model intensive decision support tool that combines innovative data visualisation methods, large language models and automated model fitting techniques within a multi-modal, interactive infrastructure. The research is at the intersection of AI, human-computer interaction and design and is aiming to contribute to these areas through methodological and technological innovations as well as having real-world impact in supporting the elimination of a neglected tropical disease.
Further context
Gambiense human African trypanosomiasis (gHAT, African sleeping sickness) is a neglected tropical disease (NTD) and one of the few targeted for elimination of transmission (EoT) under the World Health Organization’s (WHO’s) 2030 roadmap. Whilst global trends have seen substantial reductions in reported case incidence in the past two decades and many countries have validated elimination as a public health problem (<1 case per 10,000 population averaged over five years at a health district level), recent global cases (2020–2022) have not continued to fall at the same rate. It is well recognised that progressing from very low to zero cases is especially challenging, with a large proportion of current case burden in regions with poor infrastructure and health care systems and/or ongoing conflict. Modelling work to date is optimistic that the current gHAT toolkit could successfully interrupt transmission if included as part of appropriate interventions. This is, however, contingent on the successful operationalisation and application of appropriate tools.
This proposal extends previous modelling and health economic work of the HAT Modelling and Economic Predictions for Policy (HAT MEPP) 1 and 2 projects to now incorporate other endemic regions, in addition to providing continued support for the original five countries as their strategies evolve and the question of durable elimination becomes increasingly important. Vital to this work is close collaboration with national programmes and communication of the modelling results by the HAT MEPP team. To support this, the HAT MEPP graphical user interface (GUI) will continue to be refined and improved based on updated data, analysis and user feedback in order to provide granular information to support on-the-ground implementation.