Machine learning algorithms help predict tick-borne pathogen risk across Europe
Climate and vegetation factors were stronger predictors of ticks than hosts.
A new study published in Annals of Medicine explores the potential for machine learning algorithms to predict the risk of tick-borne diseases around Europe. Ticks, known carriers of many pathogens, are influenced by a combination of factors, including climate, landscape, and host availability. This research uses machine learning to predict where tick concentrations are highest.
This study used five different tools including Random Forest and Gradient Boosting to model the distribution of four tick species across the continent. By analyzing over 19,000 grid cells of 20 kilometres in diameter, they incorporated data on vegetation, water source, and host animals to create detailed maps.
Findings revealed that climate and vegetation factors were the strongest predictors of tick presence. Surprisingly, data on distribution of host animals such as deer and rodents did not add much value to the models. This suggests that environmental conditions play a larger role in where ticks thrive.
These new models could assist with planning to prevent tick-borne diseases in Europe and potentially beyond. By identifying high risk areas, public health authorities can better focus prevention efforts to protect communities from the increasing threat of tick-borne illnesses.
For more information, visit the original article in Annals of Medicine.
Citation
Agustín Estrada-Peña, & de, J. (2024). Machine learning algorithms for the evaluation of risk by tick-borne pathogens in Europe. Annals of Medicine, 56(1). https://doi.org/10.1080/07853890.2024.2405074