# Characterising responses in group-housed pigs to Salmonella typhimurium infection through integrated computer vision–based behavioural monitoring and statistical analyses

## TL;DR

This study uses computer vision to monitor pig behavior and detect Salmonella infection, showing it can track health changes over time but may miss early signs.

## Contribution

Integrates computer vision and statistical analysis to detect Salmonella-induced behavioral changes in group-housed pigs.

## Key findings

- TRT pigs showed significant health deterioration 4 days post-inoculation, confirmed by growth decline and elevated health indicators.
- Automated behavior monitoring revealed subtle behavioral anomalies as early as 0 days post-inoculation using time-specific Z-score analysis.
- Compensatory growth was observed in TRT pigs during the recovery period after antibiotic treatment.

## Abstract

Health monitoring is crucial for early disease detection and prompt intervention to mitigate the disease. Computer vision is one of the novel methods for disease detection, but a significant gap remains in its application for detecting behavioural deviations associated with disease. This study employed YOLOv8s-based behavioural monitoring combined with statistical analysis to evaluate disease detection efficacy in group-housed pigs. Two groups of pigs (Control [CON] and Treatment [TRT]), 9–10 weeks old of a (Large White × Landrace) × Duroc cross, were raised for 21 days. The growing period was divided into three periods (adaptation, challenge, and recovery) and evaluated based on growth performance, health indicators (ear base temperature and faecal score), and behaviour (postures, feeding, and drinking). The TRT group was challenged with Salmonella typhimurium during the challenge period to induce infection, then treated with antibiotics. Two pre-trained YOLOv8s models were employed to quantify postures (Lateral Lying, Sternal Lying, Standing, and Sitting) and nutritive behaviours (Feeding and Drinking). Z-score analyses based on daily data (DZA) and time-specific or 12-h interval (TSZA) data were used to detect behavioural anomalies, with the adaptation period as the baseline.

During the challenge period, TRT pigs exhibited a drastic decline in growth, increased ear base temperature, and elevated faecal scores, confirming successful infection. Compensatory growth was observed during the recovery period. Automated behaviour monitoring enabled detailed temporal analysis of responses to infection, treatment, and environmental fluctuations. Notable behavioural deviations in the TRT group emerged at 4 days post-inoculation (DPI), aligning with significant health deterioration. However, health indicators diverged as early as 1 DPI, suggesting that group-based behavioural monitoring may be less sensitive to early individual responses. TSZA detected subtle behavioural anomalies earlier than DZA, with disruptions in the TRT group beginning at 0 DPI. These included sharp fluctuations in sitting, lying, and feeding behaviours, which gradually stabilised after treatment.

This study highlights the potential of computer vision-based behavioural monitoring as a non-invasive, high-throughput tool for real-time health surveillance. While effective for group assessments, results emphasise the need for more advanced methods to enhance early disease detection and improve precision in pig health management.

The online version contains supplementary material available at 10.1186/s40813-026-00497-2.