Developing AI to detect pain in horses

  

© Lifeontheside Dreamstime.com

Recognising pain in horses is one of the greatest challenges facing veterinarians and owners.
As a prey species, horses instinctively mask signs of discomfort to avoid appearing vulnerable, meaning even experienced clinicians can struggle to identify pain in its early stages. New research from an international team of scientists has developed an artificial intelligence (AI) system that could eventually provide an additional tool to help monitor equine welfare.

The researchers, led by Dr Marcelo Feighelstein of Tel Hai University, Israel, have developed a new AI framework called SHIC-XE (Stable Heatmap Integration and Comparison for Explainable AI). Unlike previous AI systems, SHIC-XE not only detects signs that may indicate pain in horses but also provides a clear explanation of how it reached its conclusions.

Current AI systems often use coloured "heat maps" to show which parts of an image influenced a decision. While these can work reasonably well with still photographs, they become unreliable when analysing video. As the horse or camera moves, the highlighted areas can flicker from one frame to the next, making it difficult to understand exactly what the AI is using to identify pain.

SHIC-XE overcomes this limitation by mapping video images onto a stable three-dimensional model of the horse's face. This allows the system to keep its focus on the same anatomical structures, even when the horse changes head position or is viewed from different angles. Rather than producing inconsistent heat maps, the AI consistently relates its decisions to specific facial regions.

Importantly, the researchers showed that the AI concentrated on areas that experienced veterinarians also regard as important indicators of discomfort, particularly the ears and cheek muscles. By comparing the AI's attention with scores from a validated equine pain scale, they demonstrated meaningful agreement between the model and expert assessments.

This represents one of the first studies to provide a quantitative measure of whether an AI explanation matches established veterinary knowledge, rather than relying solely on subjective visual interpretation. The work also highlights that previous claims about the clinical usefulness of AI explanations may need further validation if they were based only on visual inspection.

The system was evaluated using three separate datasets containing videos of horses showing different levels of pain. Its performance was measured using an F1 score, which combines two important measures: how often the system correctly identified horses in pain and how often it avoided incorrectly classifying horses as being in pain. An F1 score of 1.00 would represent perfect performance, while a score of 0 would indicate very poor performance. The system achieved scores of 0.67, 0.80 and 0.70 across the three datasets, suggesting that it performed reasonably well, although it was not perfect.

The researchers used leave-one-subject-out validation, meaning that the system was tested on horses whose videos had not been used to train it. This provides a stronger indication of whether the AI can work with unfamiliar horses rather than simply memorising the animals in its training data.

The researchers stress that the technology is intended to support, rather than replace, veterinary judgement. In the future, AI could provide continuous monitoring in stables, veterinary hospitals or during transport, alerting staff when a horse's behaviour suggests developing pain, even when no one is actively observing it.

Although the current system simply identifies the presence of pain, the research team hopes future versions may be able to estimate pain severity and perhaps distinguish between different causes, such as orthopaedic injury, inflammation or post-operative discomfort. However, these capabilities remain future goals and have not yet been demonstrated.

While SHIC-XE was developed using horses, the underlying approach could have much wider applications. Similar techniques may eventually assist with pain assessment in other animal species and even in human patients who cannot communicate verbally, such as newborn babies or sedated intensive care patients. For equine medicine, however, the research represents an important step towards AI systems that are not only accurate, but also transparent enough for veterinarians to understand and trust.

 

For more details, see:

Marcelo Feighelstein, Omer Bibi, Ofer Rozenbaum, Nathali Adrielli Agassi De Sales, Guilherme Camargo Ferraz, Ilan Shimshoni, Dirk van der Linden, Emanuela Dalla Costa, Annika Bremhorst, Claudia Spadavecchia & Anna Zamansky 

SHIC-XE: Viewpoint-Invariant Explainability via Dense 2D-3D Correspondences: an Application to Equine Pain Recognition. 

Int J Comput Vis (2026) 134, 342

https://doi.org/10.1007/s11263-026-02910-3

Do pre-ride intravenous fluids benefit endurance horses?

   

© KertuSaarits Dreamstime.com

Endurance riding places exceptional demands on horses, particularly during competitions held in hot weather over
long distances. Maintaining hydration and electrolyte balance is therefore a major concern for riders and veterinarians.

 

As a result, administering intravenous (IV) fluids before competition has become increasingly common, especially at major endurance events where horses are expected to compete in high temperatures, over difficult terrain, or under conditions that increase the risk of dehydration.

 

Despite its popularity, however, there has been little scientific evidence to show whether pre-ride IV fluids actually improve hydration, reduce cardiovascular strain, or enhance performance.

 

To investigate this question, Laura Weintraub and colleagues at Loomis Basin Equine Medical Center, Penryn, California, conducted a study to determine whether administering intravenous fluids before an endurance competition provides measurable physiological or clinical benefits. Their findings have been published in the Journal of Equine Veterinary Science.

 

The researchers hypothesised that horses receiving IV fluids the day before competition would be better hydrated and experience less cardiovascular stress than untreated horses. They expected the treated horses to have lower heart rates during the ride together with laboratory findings consistent with improved hydration.

 

Fourteen client-owned horses entered in a 45 km endurance ride were enrolled in the study. The event was held in extreme heat over challenging terrain, providing an ideal opportunity to assess whether additional fluid support offered any advantage.

 

The horses were randomly assigned to one of two groups. One group received intravenous fluids the day before the ride, while the control group received no fluid treatment.

 

Data were collected at six time points: before travelling from home, at check-in the day before the ride, two to three hours after catheterisation and treatment, one hour before the start, 32 km into the ride, and immediately after finishing.

 

At each assessment, the horses underwent physical examination and blood sampling. Laboratory measurements included bicarbonate, sodium, potassium, chloride, calcium, glucose, lactate, blood urea nitrogen (BUN), creatinine, packed cell volume and total protein. The results were analysed using two-way analysis of variance (ANOVA).

 

The researchers found only minor differences between the two groups.

 

Horses that received IV fluids had lower total protein concentrations shortly after treatment, reflecting the expected short-term dilution of the blood following fluid administration. Blood urea nitrogen concentrations were also lower in the treated horses at the 32 km checkpoint, suggesting a modest effect on hydration status or kidney-related blood parameters during exercise.

 

However, these changes did not translate into clinically meaningful benefits.

 

Heart rate, one of the most important indicators of cardiovascular effort and recovery in endurance horses, was similar in both groups. Horses receiving IV fluids had a mean heart rate of 36 ± 5 beats per minute compared with 39 ± 3 beats per minute in the untreated horses, a difference that was not statistically significant.

 

Overall, the study suggests that while pre-competition IV fluids can produce measurable changes in some blood parameters, these changes are small and may have little practical significance. The reduction in total protein reflects temporary haemodilution rather than improved fitness for exercise, while the lower BUN concentration observed during the ride was not accompanied by improved cardiovascular responses or other indicators of enhanced performance.

 

The findings therefore challenge the assumption that routinely administering intravenous fluids before competition benefits healthy endurance horses. In this study, pre-ride fluid therapy did not reduce cardiovascular strain, improve heart-rate recovery, or produce widespread improvements in laboratory markers of hydration.

 

For veterinarians, riders and event organisers, the study provides valuable evidence when considering the routine use of pre-competition IV fluids. Fluid therapy requires veterinary supervision, venous catheterisation, specialised equipment and additional time. If it offers little measurable benefit in healthy horses, its routine use as a preventive hydration strategy may not be justified.

 

These findings do not mean intravenous fluids are never appropriate. Horses showing clinical signs of dehydration, heat stress, illness or poor recovery may still require fluid therapy as part of their veterinary treatment. However, this study suggests that administering IV fluids simply because a horse is about to tackle a demanding endurance ride may provide far less benefit than many riders and veterinarians have assumed.

 

For more details, see:

 

Weintraub, L., Fielding, C. L., Carli, I., & Greene-Dittz, G.

The effect of intravenous fluids administered prior to competition: A randomized clinical trial. 

Journal of Equine Veterinary Science,(2026) 106121.

https://doi.org/10.1016/j.jevs.2026.106121

New virus found in Swiss horses

© Viacheslav Nemyrivskyi Dreamstime.com

  

Swiss researchers have identified a novel virus in the brains of four horses that were
euthanased after developing neurological disease, prompting investigations into whether the virus may represent an emerging cause of equine central nervous system (CNS) disease.

Since late July 2026, equine referral hospitals at the Universities of Bern and Zurich have seen an increase in horses presenting with neurological signs including ataxia, muscle fasciculations, hypersensitivity and lethargy. Extensive diagnostic investigations failed to identify recognised infectious causes of neurological disease in many of these cases.

Researchers at the Swiss Institute of Virology and Immunology (IVI), working with the Vetsuisse Faculty at the University of Bern, subsequently detected genetic material from a Shamonda-like virus, an Orthobunyavirus in the Simbu serogroup, in brain tissue collected from four affected horses following post-mortem examination. Three cases originated from Zurich and one from Bern. This is the first reported detection of a Simbu serogroup Orthobunyavirus in horses in Europe.

The virus has been confirmed in post-mortem brain tissue from four horses with neurological disease. It has not yet been found in blood or cerebrospinal fluid from affected live horses, so ante-mortem diagnosis is currently not possible. Antibody testing can show previous exposure, but cannot confirm that the virus caused the neurological disease. There is currently no specific treatment, and affected horses receive supportive care.

Like other members of the Simbu serogroup, the virus is believed to be transmitted by Culicoides biting midges. Related viruses, including Schmallenberg virus, are well recognised in European livestock.

At present, there is no evidence that the Shamonda-like virus causes neurological disease in horses. Although its detection in affected brain tissue makes it a plausible candidate, researchers have not established a causal relationship. Further studies are needed to determine whether the virus is responsible for the recent increase in neurological cases in Switzerland or whether its presence is incidental.

Veterinarians investigating horses with neurological signs should therefore continue to consider established differential diagnoses, including equine herpesvirus (EHV), West Nile virus (WNV) and tick-borne encephalitis, while research into the Shamonda-like virus continues.

The UK's Equine Infectious Disease Surveillance (EIDS), based at the University of Cambridge, reports that there is currently no evidence of Shamonda-like virus-associated disease in horses in the UK. However, because infected Culicoides midges may be transported long distances by wind, and horses continue to move between mainland Europe and the UK, the situation is being monitored closely. 

 

For more details, see:

https://www.ivi.admin.ch/en/neues-orthobunyavirus-der-simbu-serogruppe-bei-rindern-entdeckte

Cold Plasma Therapy shows promise for tackling resistant equine wound infections

   Cuts, lacerations and puncture wounds are an unfortunate fact of life for horses. While many wounds heal quickly after prompt veterinary treatment, others, particularly those involving the lower limbs, can become prolonged and frustrating cases. Constant movement, contamination and infection can all interfere with healing, while the development of exuberant granulation tissue, or "proud flesh", may further delay recovery. 

Standard treatment relies on thorough wound debridement to remove dead and infected tissue, together with copious irrigation to reduce bacterial contamination. Current veterinary recommendations suggest antibiotics should not automatically be the first line of treatment for every infected wound, particularly given the growing global concern over antimicrobial resistance. However, infections caused by multidrug-resistant bacteria such as methicillin-resistant Staphylococcus aureus (MRSA) present an increasingly difficult therapeutic challenge.

Researchers in Germany have now investigated an innovative technology that could offer a valuable alternative. Led by Sandra Kurras of Tierklinik Luesche GmbH, the team evaluated the use of Cold Atmospheric Plasma-Aerosol (CAP-A)* as a non-antibiotic treatment for infected equine wounds containing antibiotic-resistant bacteria.

Although the name sounds highly technical, the principle is relatively straightforward. The CAP-A system uses a high-voltage electrical field to energise ordinary air, producing highly reactive oxygen molecules known as Reactive Oxygen Species (ROS). These include compounds such as hydroxyl radicals together with very small amounts of ozone.

An ultrasonic nebuliser then combines these reactive molecules with a fine water mist, creating electrostatically active droplets that are directed onto the wound from a distance of around 5–10 cm. A single three-minute treatment can cover an area of up to 20 × 20 cm.

The reactive oxygen species damage bacterial cell membranes, proteins and DNA, making it difficult for bacteria to survive or develop resistance. At the same time, the fine mist keeps the wound moist, helping regulate wound pH and preventing the tissues from drying out – both factors known to support normal healing.

The retrospective case series included four horses ranging in age from just five weeks to 11 years. Each horse had a deep infected wound or sinus infection containing multidrug-resistant bacteria, including methicillin-resistant Staphylococcus aureus (MRSA).

The horses received CAP-A treatment once daily for between 20 and 55 days. Each session consisted of two three-minute applications, while identical wound dressings were used throughout the study. Bacterial cultures and antibiotic sensitivity testing were performed before and after treatment, and wound healing was monitored photographically.

The results were encouraging. All four horses showed substantial reductions in bacterial contamination, with MRSA becoming undetectable after just 10 days of treatment. Clinical healing progressed in every case despite the absence of antibiotic therapy, and no adverse effects or signs of discomfort were observed during treatment sessions. The researchers also reported improvements in wounds infected with multidrug-resistant Gram-negative bacteria.

Although the study involved only four horses and lacked a control group, the findings suggest that CAP-A could become a valuable addition to equine wound care, particularly when antimicrobial resistance limits conventional treatment options. By reducing bacterial burden without relying on antibiotics, the technology also aligns with the wider One Health initiative, which seeks to reduce unnecessary antimicrobial use across both human and veterinary medicine.

Larger prospective clinical trials will now be needed to confirm the effectiveness of CAP-A, establish treatment protocols and determine where it best fits alongside established wound management techniques. Nevertheless, these initial results indicate that cold atmospheric plasma-aerosol could represent an important new tool in the battle against antibiotic-resistant wound infections in horses.

For more details, see:

Kurras, Sandra, Theresa Maria Conze, Sinje Obermann, Robert Fuchs, Tim Tischendorf, Derek C. Knottenbelt, and Marc Koene. 2026.

Cold Atmospheric Plasma-Aerosol Treatment of Equine Antibiotic-Resistant Wound Infection: A Preliminary Case Series

 Animals (2026) 16, no. 15: 2380.

https://doi.org/10.3390/ani16152380

* PLASMO®VET CAP-A (WK-MedTec GmbH, Bückeburg, Germany)

New research explores different types of stereotypic behaviour

© Callipso88 Dreamstime.com

 

Stereotypic behaviours such as crib-biting, weaving and box-walking are among the most recognisable signs of compromised welfare in horses. These repetitive behaviours appear to serve no obvious purpose and are most seen in horses kept under restrictive management conditions, where opportunities for natural behaviours such as continuous grazing, free movement and social interaction are limited.

Although stereotypies are widely regarded as indicators of reduced welfare, scientists are increasingly recognising that not all stereotypic behaviours arise for the same reasons. Understanding these differences could help develop more effective management strategies for affected horses.

Stereotypies are generally divided into two broad categories. Oral stereotypies involve repetitive movements of the mouth, with crib-biting being the best-known example. Locomotor stereotypies involve repeated movement of part or all of the body, such as weaving or box-walking.

As part of her PhD research at the Centre for Comparative and Evolutionary Psychology, University of Portsmouth, Kate Lewis and colleagues investigated whether horses displaying different forms of stereotypic behaviour also differed in their responses to a potential source of social enrichment: a stable mirror.

Mirrors have previously been shown to provide some horses with the illusion of another horse being present, potentially helping to reduce stress associated with social isolation. The researchers wanted to determine whether horses exhibiting locomotor stereotypies might be more socially motivated than those showing oral stereotypies.

The study involved 16 horses: six horses that regularly box-walked, five that crib-bit, and five control horses with no stereotypic behaviour. Each horse underwent a 24-hour baseline observation period, followed by 24 hours with a large stable mirror measuring one by 1.5 metres installed in the stable. Behaviour was then monitored for a further 24 hours after the mirror was removed.

The results revealed clear differences between the groups. Horses that box-walked showed a noticeable reduction in the behaviour while the mirror was present, and this reduction persisted even after the mirror had been removed. In contrast, crib-biting behaviour was unaffected by the presence of the mirror.

The horses also differed in how much attention they paid to the mirror. Both the box-walking horses and the control horses spent more time orientated towards it than the crib-biters, suggesting that the locomotor stereotypy group may have been more interested in the apparent social companion.

However, the findings also showed that the picture is more complex than simple social attraction. Box-walking horses did not display more ear or head movements associated with affiliative social behaviour towards the mirror, nor did they show greater signs of frustration after the mirror was removed. This suggests that although mirrors appeared to reduce box-walking, the underlying reasons remain unclear.

The researchers conclude that the study provides preliminary evidence that horses displaying different stereotypies respond differently to environmental enrichment. In particular, locomotor stereotypies such as box-walking may have stronger links with social factors than oral stereotypies like crib-biting.

Because the study involved relatively small numbers of horses, the authors emphasise that larger, well-controlled studies are needed before firm conclusions can be drawn. Nevertheless, the findings add to growing evidence that stereotypic behaviours should not be viewed as a single condition, but rather as behaviours with potentially different underlying causes that may require different approaches to management and welfare improvement.

For more details, see:

Lewis, K., McBride, S. D., Parker, M. O., Faithfull, R., & Proops, L.

Differences between horses (Equus caballus) displaying oral and locomotor stereotypies in their responses to a stable mirror. 

Behavioural processes, (2026). 241, 105415.

https://doi.org/10.1016/j.beproc.2026.105415

What riding school horses' faces reveal about their experience

  

© Igordabari  Dreamstime.com

   

Horse owners are often taught to watch for obvious signs of discomfort, such as pinned ears, tail swishing, bucking or resistance under saddle. But what if a horse's face could reveal more subtle clues about how it is coping long before those more obvious behaviours appear?

New research from the University of Guelph suggests that tiny changes in facial expression may provide valuable insights into how lesson horses experience their work and interactions with riders.

 As part of her PhD research in Equine Exercise Physiology, Aveil Oberhammer used the Equine Facial Action Coding System (EquiFACS) to study the facial expressions of 35 riding school horses before, during and after both beginner and advanced hunter/jumper lessons. Unlike the Horse Grimace Scale, which was developed specifically to identify pain, EquiFACS is a comprehensive system that records individual facial muscle movements without assigning them an emotional meaning. Researchers can then investigate what those movements may indicate in different situations, from learning and communication to stress and welfare.

The study examined short video clips taken at five stages of a typical lesson: before tacking up, after saddling, after bridling, immediately after riding and following untacking. A certified EquiFACS observer analysed each clip, recording facial action units including eye movements, ear position, nostril shape and oral behaviours such as chewing and lip movements.

Some facial expressions are surprisingly difficult to interpret. Chewing, for example, has traditionally been viewed as a sign of relaxation, yet growing evidence suggests it may instead function as a self-soothing behaviour that helps horses regulate their emotional state. Similarly, nostril dilation may reflect physical exertion, heightened attention or anticipation depending on the circumstances.

"You need to be very observant," Oberhammer explained. "Doing it live is much harder than analysing video. Sometimes you're trying to determine whether a tiny lip movement is upwards or backwards because that changes the facial action score."

One of the study's key questions challenged a common assumption within riding schools: that beginner riders are easier for horses because the physical demands are generally lower. The findings suggest the reality is more complex.

Overall, rider ability alone had surprisingly little effect on the horses' facial expressions. However, certain groups of horses responded differently. Less experienced lesson horses, particularly those with fewer than three years in a lesson programme, were more likely to display facial movements such as nostril lifting, nostril dilation and tongue movements when working with beginner riders. Mares also showed more facial activity than geldings and exhibited more eye blinks, nostril movements and lip parting with novice riders.

These findings may reflect the additional mental effort required to interpret inconsistent rider cues. Horses constantly attempt to understand what their riders are asking, and unclear timing or conflicting signals may increase their cognitive workload.

"We know that horses learn from trial and error," said Oberhammer. "If they don't associate a certain cue with a positive outcome, they might just do nothing."

The research also highlighted how facial expressions changed throughout the lesson itself. Horses were most expressive before preparation began, showing a variety of movements around the eyes, nostrils and mouth. Overall facial activity declined after saddling and became even more subdued once bridled. Following the ride, oral behaviours such as chewing became more common again, suggesting horses may communicate differently once tack and riding pressures are removed.

The authors conclude that the greatest changes in facial expression were linked not to rider ability but to different stages of the lesson, potentially reflecting anticipation before work and recovery or relief afterwards.

As equitation science continues to reshape understanding of horse behaviour and welfare, subtle facial expressions are emerging as another valuable tool for looking "behind the bridle". While they should never be interpreted in isolation, careful observation of these quiet signals could help riders, coaches and instructors better understand how horses experience their work, refine communication and support more horse-centred training and management practices.

 

For more details, see:

Aveil Oberhammer, Caleigh Copelin, Lais Rabel, Katrina Merkies

Behind the bridle: Understanding facial expressions in lesson horses before and after riding.

Applied Animal Behaviour Science,(2026) Vol 300,106982,

https://doi.org/10.1016/j.applanim.2026.106982

 

Interested in learning more about subtle behavioural cues, learning theory, and how horses communicate? Discover how the latest research in equine behaviour can help you better understand your horse and build stronger partnerships through informed, welfare-focused training.
Register for Equine Guelph’s 12-week accredited, online course Equine Behaviour, offered in partnership with School of Continuing Studies through the University of Guelph.  Go to:

https://courses.scs.uoguelph.ca/search/publicCourseSearchDetails.do?method=load&courseId=17912