A veterinarian familiarised with a catalogue of equine facial expressions was able to identify ridden horses with lameness issues through studying pictures that included only the head and neck.
The findings of the British study have been reported in the Journal of Veterinary Behavior: Clinical Applications and Research. The researchers expressed the hope that horse owners could be taught to use a simpler version in order to potentially detect lameness issues earlier.
Poor horse performances were often explained away by rider issues, training problems or behavioural abnormalities. However, riders often failed to recognise lameness, the study team noted.
For their study, they used an ethogram – a catalogue of observed behaviours – based around equine facial expressions which they had described in a previous paper.
In that study, the ethogram was tested by 13 individuals, from horse-owning amateurs to veterinarians, with their level of agreement being assessed at 87 percent. Their level of horse experience had little influence on the accuracy of their assessments.
It was concluded that the ethogram could reliably be used by people from different professional backgrounds to describe equine facial expressions, but further work was needed to determine if non-lame and lame horses could be differentiated through its use.
The study team, led by equine orthopaedics specialist Sue Dyson, set out to determine if the ethogram could be adapted to a pain scoring system for ridden horses.
For this second phase, an equine veterinarian was trained in the use of the ethogram and was provided with a customised instruction catalogue.
She was asked to assess a total of 519 head and neck photographs of 101 horses under saddle at both the trot and canter. Seventy-six of the horses had previously been assessed as lame and 25 were considered free of gait abnormalities. The pictures included 30 photographs of seven lame horses, together with 22 photographs of the same horses after local analgesia had temporarily abolished their lameness issues.
The number of pictures for each horse ranged from 3 to 13.
The vet analysed the images without any prior knowledge of the soundness of the horses involved.
She applied a pain score – where 0 was normal and 1-3 were abnormal – for each feature identified in the ethogram. In effect, she was applying a facial pain score.
In the end, the vet identified a total of 27,407 facial markers across all the photographs.
Dyson and her colleagues reported that the pain scores were significantly higher for lame horses than non-lame horses. The lame horses returned lower scores after the local anaesthetic drugs had kicked in, which the study team had hypothesised would happen.
The facial markers showing the greatest difference between lame and sound horses included ears back, a twisted nose, being above the bit, eyes partially or fully closed, signs of tension around the eyes, intense staring, and an open mouth with teeth showing.
The study team concluded that the facial expression ethogram in combination with a pain score can be used to differentiate sound and lame horses.
“It is likely,” they said, “that the ethogram can be simplified to enable its use by owners, trainers, paraprofessionals and veterinarians to enable pain to be detected without recognition of changes in gait, or in association with lameness.
“This is a significant development towards the assessment of welfare in ridden horses.”
The study team comprised Dyson, Jeannine Berger, Andrea Ellis and Jessica Mullard, who was the veterinarian who assessed the images.
The study was supported by World Horse Welfare and the Saddle Research Trust.
Can the presence of musculoskeletal pain be determined from the facial expressions of ridden horses (FEReq)?
Sue Dyson, Jeannine M. Berger, Andrea D. Ellis, Jessica Mullard
The abstract can be read here.