Early detection of epidemics in livestock and wildlife using deep learning

David Duchene Garzon, University of Copenhagen
Grant amount: DKK 10,886,002

David Duchene Garzon says: “Identifying infected animals as early as possible allows us to minimize the spread of a pathogen and even prevent a pandemic. At the moment, we can only make sure that an

animal is infected by costly laboratory analysis. This is problematic for livestock and wildlife given the limited funds that can be spent on each animal, yet these settings are the most common source of dangerous pathogens to humans. Surprisingly, video data is not yet being used for identifying infected animals, despite great strides in video analysis in recent years.

This project will cover this gap and improve our ability to halt epidemics in their tracks. A broad range of animals will be filmed, and their behavior will be compared with their blood tests. Whether infected or not, each recording will help train computers, which will inform us about how pathogens can drive behaviour. A free app will then be developed for companies, governments, and lay people to detect infected animals at a minimal cost.”

Project participants
David Duchene Garzon
Postdoc, Department of Public Health, Faculty of Health and Medical Sciences, University of Copenhagen