UAEU and IIT Madras Zanzibar develop AI-based framework to forecast malaria outbreaks

In a significant step forward for global public health, researchers from United Arab Emirates University (UAEU) and the Indian Institute of Technology Madras’ Zanzibar campus have introduced a cutting-edge, data-driven framework that accurately models and forecasts malaria transmission. By integrating artificial intelligence with mathematical modelling, this new approach aims to support early intervention and improve disease control strategies in malaria-prone regions.

A new approach to malaria modelling
A collaborative research team led by Adithya Rajnarayanan, Manoj Kumar, and Professor Abdessamad Tridane has introduced a novel methodology that enhances how malaria outbreaks can be forecasted. Their work, published in Scientific Reports by Nature, presents a comprehensive model that combines artificial intelligence (AI) with classical epidemiological frameworks to simulate malaria dynamics with higher precision.

The study, titled “Analysis of a Mathematical Model for Malaria Using a Data-Driven Approach”, brings a fresh perspective to disease modelling. It incorporates temperature- and altitude-dependent variables into compartmental disease models, a method that makes the simulations more realistic and region-specific. This is particularly crucial for climate-sensitive and vulnerable areas where environmental factors heavily influence malaria transmission patterns.

Technologies at the core: AI and dynamic systems
To boost the predictive capability of their model, the researchers employed a suite of advanced AI tools. These included:

• Artificial Neural Networks (ANNs)
  
• Recurrent Neural Networks (RNNs)
  
• Physics-Informed Neural Networks (PINNs)
  
  
  

Each of these tools was used to improve the accuracy of disease forecasting, enabling the model to detect patterns in the complex interplay between environmental conditions and malaria spread.

Additionally, the study introduced Dynamic Mode Decomposition (DMD), a mathematical technique that helps break down complex systems into simpler, understandable components. This was used to create a real-time infection risk metric, offering public health officials a powerful resource for early detection and targeted response.

Implications for global health
Professor Abdessamad Tridane of UAEU emphasized the importance of this integration of AI with epidemiological modelling, stating:

This research demonstrates the power of AI when combined with classical epidemiological models,” said Prof. Abdessamad Tridane of UAEU. “By embedding environmental dependencies directly into the transmission functions, our model captures the complex, real-world behaviour of malaria spread, providing a more accurate and timely method for disease tracking.”

The implications of this research are especially relevant for regions like sub-Saharan Africa, which accounts for 94% of global malaria cases. With over half a million malaria-related deaths each year, the need for accurate forecasting models is critical. This work offers a valuable step towards improved surveillance, early warning systems, and data-driven policymaking in the fight against malaria.

Institutional collaboration and background
This study represents a collaboration between two institutions that are expanding their global health research footprint:

• United Arab Emirates University (UAEU), established in 1976 in Al Ain, is the UAE’s oldest public research university. Founded by Sheikh Zayed bin Sultan Al Nahyan, it offers a wide range of undergraduate and postgraduate programs across multiple disciplines.
  
  
• IIT Madras Zanzibar Campus, inaugurated in November 2023, is the first international campus of the Indian Institute of Technology Madras. Located in the Bweleo district of Zanzibar, Tanzania, the campus currently offers programs in Data Science and Artificial Intelligence. It aims to cater to a diverse student population from India, Tanzania, and other African nations, with plans to broaden its academic scope in the coming years.