Projects

DSTN > Projects > AIRFARE EWS

Responsible Researchers:
● ACE MITIC, Senegal
○ Dr Dame Diongue, Gaston Berger University, Senegal.
○ Prof. Maissa Mbaye, Gaston Berger University, Senegal.
○ Dr. Nicolas Djighnoum Diouf, Université Gaston Berger, Senegal.

● ACE CApIC, Nigeria
○ Prof. E. F. Adebiyi, Covenant University, Ota, Nigeria
○ Prof. E. Adetiba, CApIC-ACE, Covenant University, Ota, Nigeria
○ Dr. Joke A. Badejo, Covenant University, Ota, Nigeria


Other partners:

● Dr Assane G. Fall, Institut sénégalais de recherches agricoles, Senegal.
● Dr Mamadou Ciss, Institut sénégalais de recherches agricoles, Senegal.
● Dr Modou Moustapha Lo, Institut Sénégalais de Recherches Agricoles, Senegal.
● Dr Ibrahima Ndao, Direction des parcs et réserves, Saint-Louis, Senegal.
● Dr. Seynabou Diack Sy, Veterinary Services, Saint-Louis, Senegal.

DSTN > Projects > AIRFARE EWS

Title: Artificial intelligence-based early warning system for the detection of Rift Valley fever

Type of project: Development project (2 years)

Summary:
Rift Valley Fever (RVF) is a viral zoonosis that can cause severe disease in small ruminants and humans. It extends from East Africa to Southern Africa and the Middle East. It has caused massive epidemics in Egypt, Mauritania (2010, 2012), Senegal (2012, 2020) and Madagascar. It can be transmitted between humans and animals, and its main manifestation is an abnormal rate of abortions in animals, which can be fatal for humans. Epidemic warning systems in African countries such as Senegal face the problem of early detection of outbreaks if they are to be effective (data collection and forecasting). Both breeders and laboratory technicians are at risk due to probable handling prior to confirmed diagnosis. However, research such as the NASA studies show the relationship between VFR epidemics and environmental parameters such as the presence of Nina to be able to predict the next epidemic. 

The aim of this project is to develop an early warning system for Rift Valley fever based on AI for prediction and detection. It consists of a data collection system based on IoT/mobile applications and a Cloud AI/machine learning service (classification, reinforcement deep learning, CNN, computer vision, ...). The expected achievements are the incremental construction of datasets that can be used to provide Cloud AI/ML services with spatial and temporal correlations to rapidly recognize the onset of epidemics. Data are heterogeneous (abortion images, vectors/mosquitoes, official datasets, ...), from various sources (WHO, ISRA, on-site collection, as well as breeders), of different levels of validation (declaration of suspicion by breeders, confirmed analysis by laboratory, ...), bi-dimensional (special and temporal). The mobile application will support on-site data collection. It is therefore necessary to find ML models that adapt well to data analysis, taking all these characteristics into account, and that learn incrementally. The consortium involves MITIC for AI and IoT services and CaPIC for Cloud service developments in FedGenHealth. We also have ISRA (Institut Sénégalais de Recherche Agronomique) as a technical partner for medical and vector-related issues. 

Keywords:
● Digital science research themes: Emerging diseases, artificial intelligence, early warning system, One Health, TinyML.
● Other research themes and application areas: Rift Valley fever, zoonoses, epidemic surveillance system, deep machine learning, cloud services, IoT, FEDGEN platform.