Predicting the fine-scale spatial distribution of zoonotic reservoirs using computer vision.

Autor: Layman NC; EcoHealth Alliance, New York, New York, USA.; Institute for Interdisciplinary Data Sciences, University of Idaho, Moscow, Idaho, USA., Basinski AJ; Institute for Interdisciplinary Data Sciences, University of Idaho, Moscow, Idaho, USA., Zhang B; Institute for Interdisciplinary Data Sciences, University of Idaho, Moscow, Idaho, USA., Eskew EA; Institute for Interdisciplinary Data Sciences, University of Idaho, Moscow, Idaho, USA., Bird BH; One Health Institute, School of Veterinary Medicine, University of California-Davis, Davis, California, USA., Ghersi BM; One Health Institute, School of Veterinary Medicine, University of California-Davis, Davis, California, USA.; Tufts University, Medford, Massachusetts, USA., Bangura J; University of Makeni and University of California, Davis One Health Program, Makeni, Sierra Leone., Fichet-Calvet E; Bernhard Nocht Institute for Tropical Medicine, Hamburg, Germany., Remien CH; Department of Mathematics and Statistical Science, University of Idaho, Moscow, Idaho, USA., Vandi M; Ministry of Health and Sanitation, Freetown, Sierra Leone., Bah M; Ministry of Agriculture and Forestry, Freetown, Sierra Leone., Nuismer SL; Department of Biological Sciences, University of Idaho, Moscow, Idaho, USA.
Jazyk: angličtina
Zdroj: Ecology letters [Ecol Lett] 2023 Nov; Vol. 26 (11), pp. 1974-1986. Date of Electronic Publication: 2023 Sep 22.
DOI: 10.1111/ele.14307
Abstrakt: Zoonotic diseases threaten human health worldwide and are often associated with anthropogenic disturbance. Predicting how disturbance influences spillover risk is critical for effective disease intervention but difficult to achieve at fine spatial scales. Here, we develop a method that learns the spatial distribution of a reservoir species from aerial imagery. Our approach uses neural networks to extract features of known or hypothesized importance from images. The spatial distribution of these features is then summarized and linked to spatially explicit reservoir presence/absence data using boosted regression trees. We demonstrate the utility of our method by applying it to the reservoir of Lassa virus, Mastomys natalensis, within the West African nations of Sierra Leone and Guinea. We show that, when trained using reservoir trapping data and publicly available aerial imagery, our framework learns relationships between environmental features and reservoir occurrence and accurately ranks areas according to the likelihood of reservoir presence.
(© 2023 John Wiley & Sons Ltd.)
Databáze: MEDLINE
Externí odkaz: