Preventing West Nile Disease spread in Italy with EO data and Artificial Intelligence

West Nile virus is one of the most diffused vector-borne diseases in Italy and Europe. The virus is transmitted through birds which are primary hosts and mosquito vectors which transfer the virus to other birds, while humans and horses are dead-end hosts. Identifying large areas characterized by the presence of hosts and vectors and therefore having the highest likelihood to spread the virus can be challenging.

The AIDEO project – Artificial Intelligence and Earth Observation data: innovative methods for monitoring West Nile Disease spread in Italy aims at combining the massive availability of Earth Observation data with the most recent innovative developments in Artificial Intelligence to make highly accurate predictions of where the virus is more likely to spread. More concretely, AIDEO aims at verifying the possibility of producing West Nile disease risk maps by using EO data and AI algorithms. The risk maps will be elaborated by applying AI learning architectures to ground data extrapolated from the National Information System for Animal Disease Notification (SIMAN) and EO data gathered, pre-processed and harmonised from Sentinel -2, -3 and PROBA-V etc. This process aims at studying the temporal evolution of data as well as forecasting future behaviours of the virus. The elaborated risk maps will be then compared to classical statistical methods to assess the degree of improvement in forecasting the disease spread.

Those risk maps appear to be highly relevant for competent authorities such as Ministries of Health to: on the one side create efficient alert systems that could warn them as soon as environmental and climatic conditions become favourable for the spread of the disease, on the other to finetune surveillance activities related to the virus.

The initiative was kicked off in September 2019 and is a joint project contributed by the Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise, AImage Lab, Progressive Systems and ReMedia financed under the ESA EO Science for Society programme.