Estimation of resilience parameters based on activity measuredwith computer vision following LPS injection

Resilience could be referred to as an animal’s ability to successfully adapt to achallenge, characterized by a relatively quick return to the pre-challenge state,including normal activity levels and behaviours. Pigs have distinct diurnal activitypatterns. The level of activity patterns could be influenced by housing conditions anddeviations from these patterns could be utilised to quantify resilience, and However, humanobservations of these patterns are labour intensive and not feasible in practise. In this studywe show the use of a computer vision tracking algorithm to quantify resilience based onactivity patterns in response to a lipopolysaccharide (LPS) sickness challenge. 144 pigs werehoused in either barren or enriched pens. Four out of six pigs per pen were injected withLPS, the remaining two received a saline injection and served as controls. Results showedenriched housed pigs were more active than barren housed pigs pre-injection ofLPS. LPS injected animals showed a dip in activity followed by a recovery period, asexpected. This was not observed in the saline-control animals. Individual variation inrecovery patterns may provide important information regarding resilience of individualpigs. Although no effects of housing were identified regarding resilience, these resultsdemonstrate the usefulness of a computer vision tracking algorithm to measure resilienceusing proposed resilience parameters, and contributes to future on-farm applications.