Equine Facial Action Coding System for determination of pain-related facial responses in videos of horses.
Authors: Rashid Maheen, Silventoinen Alina, Gleerup Karina Bech, Andersen Pia Haubro
Journal: PloS one
Summary
# Equine Facial Action Coding System: Standardising Pain Assessment Through Video Analysis Inconsistency across existing equine pain assessment tools—particularly regarding which facial muscles to monitor and how to interpret them—prompted development of the Equine Facial Action Coding System (EquiFACS), a standardised methodology adapted from human FACS research to provide objective, repeatable analysis of pain-related facial responses in horses. Researchers coded video recordings from six horses undergoing acute experimental pain and 21 clinical cases (with and without suspected pain), using both traditional frequency-based statistical analysis and a novel co-occurrence method that examined combinations of facial actions within varying time windows. Ear rotation, nostril dilation and lower face actions—particularly chin raising—emerged as the most consistent pain indicators, whilst ear and eye changes showed less reliable results; notably, the co-occurrence approach identified chewing and increased half-blinking frequency as pain-specific behaviours often missed by frequency analysis alone. This refined methodology demonstrates superior cross-validation between experimental and clinical settings, offering practitioners a scientifically robust framework for recognising pain in live horses and archived video footage. For farriers, vets and physiotherapists, standardised FACS-based assessment could substantially improve pain recognition accuracy during routine practice and client communication, whilst the identification of lower facial behaviours provides additional diagnostic markers beyond the widely recognised ear and eye changes.
Read the full abstract on PubMed
Practical Takeaways
- •Watch for ear pinning, nostril flaring, and chin tightening as primary signs of pain—these are the most reliable visual indicators across different pain scenarios
- •Pay attention to lower facial changes like chewing patterns and increased blinking, which may indicate pain even when other signs are subtle or inconsistent
- •Use systematic video observation to assess multiple facial regions together rather than relying on single indicators, as combined facial actions provide better pain assessment than isolated features
Key Findings
- •Ear rotation (EAD104), nostril dilation (AD38), and chin raising (AU17) were the most reliable facial indicators of pain in horses
- •Lower face behaviors, particularly chewing (AD81), were identified as pain-specific indicators with better generalization between experimental and clinical cases
- •Half blink (AU47) frequency increased during pain states and represents a previously undocumented pain indicator
- •Co-occurrence analysis of facial actions over time slots provided better discrimination between pain and non-pain states than frequency-based methods alone