The Poses for Equine Research Dataset (PFERD).

Studies of quadruped animal motion help us to identify diseases, understand behavior and unravel the mechanics behind gaits in animals. The horse is likely the best-studied animal in this aspect, but data capture is challenging and time-consuming. Computer vision techniques improve animal motion extraction, but the development relies on reference datasets, which are scarce, not open-access and often provide data from only a few anatomical landmarks. Addressing this data gap, we introduce PFERD, a video and 3D marker motion dataset from horses using a full-body set-up of densely placed over 100 skin-attached markers and synchronized videos from ten camera angles. Five horses of diverse conformations provide data for various motions from basic poses (eg. walking, trotting) to advanced motions (eg. rearing, kicking). We further express the 3D motions with current techniques and a 3D parameterized model, the hSMAL model, establishing a baseline for 3D horse markerless motion capture. PFERD enables advanced biomechanical studies and provides a resource of ground truth data for the methodological development of markerless motion capture.

3D pose estimation to detect posture transition in free-stall housed dairy cows.

Free stall comfort is reflected in various indicators, including the ability for dairy cattle to display unhindered posture transition movements in the cubicles. To ensure farm animal welfare, it is instrumental for the farm management to be able to continuously monitor occurrences of abnormal motions. Advances in computer vision have enabled accurate kinematic measurements in several fields such as human, equine and bovine biomechanics. An important step upstream to measuring displacement during posture transitions is to determine that the behavior is accurately detected. In this study, we propose a framework for detecting lying to standing posture transitions from 3D pose estimation data. A multi-view computer vision system recorded posture transitions between Dec. 2021 and Apr. 2022 in a Swedish stall housing 183 individual cows. The output data consisted of the 3D coordinates of specific anatomical landmarks. Sensitivity of posture transition detection was 88.2% while precision reached 99.5%. Analyzing those transition movements, breakpoints detected the timestamp of onset of the rising motion, which was compared with that annotated by observers. Agreement between observers, measured by intra-class correlation, was 0.85 between 3 human observers and 0.81 when adding the automated detection. The intra-observer mean absolute difference in annotated timestamps ranged from 0.4s to 0.7s. The mean absolute difference between each observer and the automated detection ranged from 1.0s to 1.3s. There was a significant difference in annotated timestamp between all observer pairs but not between the observers and the automated detection, leading to the conclusion that the automated detection does not introduce a distinct bias. We conclude that the model is able to accurately detect the phenomenon of interest and that it is equatable to an observer.

Changes in the equine facial repertoire during different orthopedic pain intensities.

A number of facial expressions are associated with pain in horses, however, the entire display of facial activities during orthopedic pain have yet to be described. The aim of the present study was to exhaustively map changes in facial activities in eight resting horses during a progression from sound to mild and moderate degree of orthopedic pain, induced by lipopolysaccharides (LPS) administered in the tarsocrural joint. Lameness progression and regression was measured by objective gait analysis during movement, and facial activities were described by EquiFACS in video sequences (n = 348, total length 892.5 min) of the horses obtained when resting in their box stalls. Predictive modeling identified 16 action units and action descriptors, related to ears, eyes, and lower face. Lower lip depressor (AU16), lips part (AU25), half blink (AU47), single ear forward (SEAD101) and single ear rotator (SEAD104) were selected as co-occurring significantly more in horses with pain than in horses without pain. The major change in co-occurring facial activities occurred in the transition from no pain to mild pain. In conclusion, resting horses with induced orthopedic pain showed a dynamic upper and lower facial repertoire and the relationship between level of pain intensity and facial activity appears complex.

Adaptation strategies of the Icelandic horse with induced forelimb lameness at walk, trot and tölt.

BACKGROUND AND OBJECTIVE: Lameness assessment in the gaited Icelandic horse is complex. We aimed to describe their kinematic and temporal adaptation strategies in response to forelimb lameness at walk, trot and tölt. STUDY DESIGN: In vivo experiment. METHODS: Ten clinically non-lame Icelandic horses were measured before and after reversible forelimb lameness induction. Upper body and limb kinematics were measured using 11 inertial measurement units mounted on the poll, withers, pelvis (tubera sacrale) and all four limbs and hoofs (Equimoves®, 500 Hz). Horses were measured on a straight line at walk and trot in-hand and at walk, trot and tölt while ridden. Linear mixed models were used to compare baseline and lame conditions (random factor = 'horse'), and results are presented as the difference in estimated marginal means or percentage of change. RESULTS: Lameness induction significantly (p < 0.05) increased head vertical movement asymmetry at walk (HDmin/HDmaxHAND: 18.8/5.7 mm, HDmin/HDmaxRIDDEN: 9.8/0.3 mm) and trot (HDmin/HDmaxHAND: 18.1/7.8 mm, HDmin/HDmaxRIDDEN: 24.0/9.3 mm). At the tölt, however, HDmin did not change significantly (1.1 mm), but HDmax increased by 11.2 mm (p < 0.05). Furthermore, pelvis vertical movement asymmetry (PDmax) increased by 4.9 mm, sound side dissociation decreased (-8.3%), and sound diagonal dissociation increased (6.5%). Other temporal stride variables were also affected, such as increased stance duration of both forelimbs at walk, tölt and in-hand trot. MAIN LIMITATIONS: Only one degree of lameness (mild) was induced with an acute lameness model. CONCLUSIONS: Classical forelimb lameness metrics, such as vertical head and withers movement asymmetry, were less valuable at tölt compared to walk and trot, except for HDmax. Therefore, it is advised to primarily use the walk and trot to detect and quantify forelimb lameness in the Icelandic horse.

Evaluating Overall Performance in High-Level Dressage Horse-Rider Combinations by Comparing Measurements from Inertial Sensors with General Impression Scores Awarded by Judges.

In the sport of dressage, one or more judges score the combined performance of a horse and rider with an emphasis on the technical correctness of the movements performed. At the end of the test, a single score is awarded for the 'general impression', which considers the overall performance of the horse and rider as a team. This study explored original measures that contributed to the general impression score in a group of 20 horse-rider combinations. Horses and riders were equipped with inertial measurement units (200 Hz) to represent the angular motion of a horse's back and the motions of a rider's pelvis and trunk. Each combination performed a standard dressage test that was recorded to video. Sections of the video were identified for straight-line movements. The videos were analyzed by two or three judges. Four components were scored separately: gaits of the horse, rider posture, effectiveness of aids, and harmony with the horse. The main contributor to the score for gaits was stride frequency (R = -0.252, p = 0.015), with a slower frequency being preferred. Higher rider component scores were associated with more symmetrical transverse-plane trunk motion, indicating that this original measure is the most useful predictor of rider performance.

Perceived sidedness and correlation to vertical movement asymmetries in young warmblood horses.

The prevalence of vertical asymmetries is high in "owner-sound" warmblood riding horses, however the origin of these asymmetries is unknown. This study investigated correlations between vertical asymmetries and motor laterality. Young warmblood riding horses (N = 65), perceived as free from lameness were evaluated on three visits, each comprising objective gait analysis (inertial measurement units system) and a rider questionnaire on perceived sidedness of the horse. A subgroup (N = 40) of horses were also subjected to a forelimb protraction preference test intended as an assessment of motor laterality. We hypothesized associations between vertical asymmetry and motor laterality as well as rider-perceived sidedness. Vertical asymmetry was quantified as trial means of the stride-by-stride difference between the vertical displacement minima and maxima of the head (HDmin, HDmax) and pelvis (PDmin, PDmax). Laterality indexes, based on counts of which limb was protracted, and binomial tests were used to draw conclusions from the preference tests. In the three visits, 60-70% of horses exhibited vertical asymmetries exceeding clinically used thresholds for ≥1 parameter, and 22% of horses exhibited a side preference in the preference test as judged by binomial tests. Linear mixed models identified a weak but statistically significant correlation between perceived hindlimb weakness and higher PDmin values attributable to either of the hindlimbs (p = 0.023). No other statistically significant correlations to vertical asymmetry were seen for any of the questionnaire answers tested. Tests of correlation between the absolute values of laterality index and asymmetry parameters (HDmin, HDmax, PDmin, PDmax) identified a weak correlation (p = 0.049) with PDmax, but when accounting for the direction of asymmetry and motor laterality, no correlations were seen for either of the asymmetry parameters. No convincing evidence of associations between vertical asymmetries and motor laterality were seen and further studies investigating motor laterality and the origin of vertical asymmetries are needed.

Prevalence of vertical movement asymmetries at trot in Standardbred and Swedish Warmblood foals.

Many horses, just before and during their athletic career, show vertical movement asymmetries, to the same degree as clinically lame horses. It is unknown whether these asymmetries are caused by pain or have alternative explanations, such as inherent biological variation. In the latter case, movement asymmetries would be expected to be present at a very young age. This study aimed to investigate the prevalence of movement asymmetries in foals. Motion analysis, using an inertial measurement unit-based system (Equinosis), was performed on 54 foals (31 Swedish Warmbloods, 23 Standardbreds) during straight-line trot. The foals were between 4-13 weeks old and considered sound by their owners. Differences between the vertical minimum and maximum values recorded for the head (HDmin, HDmax) and pelvis (PDmin, PDmax) between left and right stance were calculated for each stride and an average was computed for each trial. Thresholds for asymmetry were defined as absolute trial mean >6 mm for HDmin and HDmax, and >3 mm for PDmin and PDmax. These thresholds were exceeded for one or several parameters by 83% of Standardbred foals and 45% of Swedish Warmblood foals, demonstrating surprisingly high prevalence of asymmetries in young foals, although the risk of repetitive strain injuries and cumulative risk of trauma injuries was expected to be low in this age group. Standardbred foals showed similar prevalence of asymmetries to that reported previously for yearling Standardbred trotters, so relatively higher prevalence of movement asymmetries may be expected among trotters as a breed. In general, vertical head and pelvic movement asymmetries can be anticipated among foals considered sound by their owners. A better understanding of the aetiology of asymmetries is needed for correct interpretation of objective symmetry measurements in different populations of horses.

Is Markerless More or Less? Comparing a Smartphone Computer Vision Method for Equine Lameness Assessment to Multi-Camera Motion Capture.

Computer vision is a subcategory of artificial intelligence focused on extraction of information from images and video. It provides a compelling new means for objective orthopaedic gait assessment in horses using accessible hardware, such as a smartphone, for markerless motion analysis. This study aimed to explore the lameness assessment capacity of a smartphone single camera (SC) markerless computer vision application by comparing measurements of the vertical motion of the head and pelvis to an optical motion capture multi-camera (MC) system using skin attached reflective markers. Twenty-five horses were recorded with a smartphone (60 Hz) and a 13 camera MC-system (200 Hz) while trotting two times back and forth on a 30 m runway. The smartphone video was processed using artificial neural networks detecting the horse's direction, action and motion of body segments. After filtering, the vertical displacement curves from the head and pelvis were synchronised between systems using cross-correlation. This rendered 655 and 404 matching stride segmented curves for the head and pelvis respectively. From the stride segmented vertical displacement signals, differences between the two minima (MinDiff) and the two maxima (MaxDiff) respectively per stride were compared between the systems. Trial mean difference between systems was 2.2 mm (range 0.0-8.7 mm) for head and 2.2 mm (range 0.0-6.5 mm) for pelvis. Within-trial standard deviations ranged between 3.1-28.1 mm for MC and between 3.6-26.2 mm for SC. The ease of use and good agreement with MC indicate that the SC application is a promising tool for detecting clinically relevant levels of asymmetry in horses, enabling frequent and convenient gait monitoring over time.

Timing of Vertical Head, Withers and Pelvis Movements Relative to the Footfalls in Different Equine Gaits and Breeds.

Knowledge of vertical motion patterns of the axial body segments is a prerequisite for the development of algorithms used in automated detection of lameness. To date, the focus has been on the trot. This study investigates the temporal synchronization between vertical motion of the axial body segments with limb kinematic events in walk and trot across three popular types of sport horses (19 Warmbloods, 23 Iberians, 26 Icelandics) that are known to have different stride kinematics, and it presents novel data describing vertical motion of the axial body segments in tölting and pacing Icelandic horses. Inertial measurement unit sensors recorded limb kinematics, vertical motion of the axial body at all symmetrical gaits that the horse could perform (walk, trot, tölt, pace). Limb kinematics, vertical range of motion and lowest/highest positions of the head, withers and pelvis were calculated. For all gaits except walk and pace, lowest/highest positions of the pelvis and withers were found to be closely related temporally to midstance and start of suspension of the hind/fore quarter, respectively. There were differences in pelvic/withers range of motion between all breeds where the Icelandic horses showed the smallest motion, which may explain why lameness evaluation in this breed is challenging.

Changes in Head, Withers, and Pelvis Movement Asymmetry in Lame Horses as a Function of Diagnostic Anesthesia Outcome, Surface and Direction.

Evaluation of diagnostic anesthesia during equine lameness examination requires comparison of complex movement patterns and can be influenced by expectation bias. There is limited research about how changes in movement asymmetries after successful analgesia are affected by different exercise conditions. Movement asymmetry of head, withers and pelvis was quantified in N = 31 horses undergoing forelimb or hindlimb diagnostic anesthesia. Evaluation on a straight line and a circle was performed with subjective diagnostic anesthesia outcome and quantitative changes recorded. Mixed linear models (P < .05) analyzed the differences in movement asymmetry before/after diagnostic anesthesia - random factor: horse, fixed factors: surface (soft, hard), direction (straight, inside, outside, inside-outside average), diagnostic anesthesia outcome (negative, partially positive, positive) and two-way interactions. Forelimb diagnostic anesthesia influenced primary movement asymmetry (all head and withers parameters) and compensatory movement asymmetry (two pelvic parameters) either individually (P≤.009) or in interaction with surface (P≤.03). Hindlimb diagnostic anesthesia influenced primary movement asymmetry (all pelvic parameters) and compensatory movement asymmetry (two head and two withers parameters) either individually (P≤.04) or in interaction with surface (P≤.01;) or direction (P≤.006). Direction was also significant individually for two pelvic parameters (P≤.04). Changes in primary movement asymmetries after partially positive or positive outcomes indicated improvement in the blocked limb. Compensatory changes were mostly in agreement with the 'law of sides'. The changes were more pronounced on the hard surface for hindlimb lameness and on the soft surface for forelimb lameness. Withers asymmetry showed distinct patterns for forelimb and hindlimb lameness potentially aiding clinical decision-making.

Evaluation of feedback methods for improved detection of hindlimb lameness in horses among riding instructors and trainers.

Lameness, a wellknown issue in sport horses, impedes performance and impairs welfare. Early detection of lameness is essential for horses to receive needed treatment, but detection of hindlimb lameness is challenging. Riding instructors and trainers observe horses in motion in their daily work and could contribute to more efficient lameness detection. In this cross-sectional and prospective study, we evaluated the ability of riding instructors and trainers to assess hindlimb lameness. We also evaluated different feedback methods for improved lameness detection. For the cross-sectional part, n = 64 riding instructors and trainers of varying level and n = 23 high-level trainers were shown 13 videos of trotting horses, lameness degree: 0-3.5 (test 1) and tasked with classifying the horses as sound, left hindlimb lame, or right hindlimb lame. For the prospective part, the riding instructors and trainers of varying levels were randomly allocated to three different groups (a, b, c) and given 14 days of feedback-based, computer-aided training in identifying hindlimb lameness, where they assessed 13 videos (of which three were repeated from test 1) of horses trotting in a straight line. Participants in groups a-c received different feedback after each video (group a: correct answer and re-viewing of video at full and 65% speed; group b: correct answer, re-viewing of video at full and 65% speed, narrator providing explanations; group c: correct answer and re-viewing of video at full speed). After computer-aided training, the participants were again subjected to the video test (test 2). Participants also provided background information regarding level of training etcetera. Effects of participants' background on results were analyzed using analysis of variance, and effects of the different feedback methods were analyzed using generalized estimation equations. On test 1, 44% (group a), 48% (b), 46% (c), and 47% (high-level trainers) of horses were correctly classified. Group a participants significantly improved their test score, both with (p < 0.0001) and without (p = 0.0086) inclusion of repeated videos. For group c, significant improvement was only seen with inclusion of repeated videos (p = 0.041). For group b, no significant improvement was seen (p = 0.51). Although test 2 scores were low, computer-aided training may be useful for improving hindlimb lameness detection.

Performance of four equine pain scales and their association to movement asymmetry in horses with induced orthopedic pain.

Objective: This study investigated the relationship between orthopedic pain experienced at rest, and degree of movement asymmetry during trot in horses with induced reversible acute arthritis. Orthopedic pain was assessed with the Horse Grimace Scale (HGS), the Equine Utrecht University Scale of Facial Assessment of Pain (EQUUS-FAP), the Equine Pain Scale (EPS), and the Composite Orthopedic Pain Scale (CPS). Reliability and diagnostic accuracy were evaluated with intraclass correlation coefficients (ICC) and area under the curve (AUC). Study design and animals: Eight healthy horses were included in this experimental study, with each horse acting as its own control. Methods: Orthopedic pain was induced by intra-articular lipopolysaccharide (LPS) administration. Serial pain assessments were performed before induction and during pain progression and regression, where three observers independently and simultaneously assessed pain at rest with the four scales. Movement asymmetry was measured once before induction and a minimum of four times after induction, using objective gait analysis. Results: On average 6.6 (standard deviation 1.2) objective gait analyses and 12.1 (2.4) pain assessments were performed per horse. The ICC for each scale was 0.75 (CPS), 0.65 (EPS), 0.52 (HGS), and 0.43 (EQUUS-FAP). Total pain scores of all scales were significantly associated with an increase in movement asymmetry (R 2 values ranging from -0.0649 to 0.493); with CPS pain scores being most closely associated with movement asymmetry. AUC varied between scales and observers, and CPS was the only scale where all observers had a good diagnostic accuracy (AUC > 0.72). Conclusions and clinical relevance: This study identified significant associations between pain experienced at rest and degree of movement asymmetry for all scales. Pain scores obtained using CPS were most closely associated with movement asymmetry. CPS was also the most accurate and reliable pain scale. All scales had varying linear and non-linear relations between total pain scores and movement asymmetry, illustrating challenges with orthopedic pain assessment during rest in subtly lame horses since movement asymmetry needs to be rather high before total pain score increase.

Modelling fore- and hindlimb peak vertical force differences in trotting horses using upper body kinematic asymmetry variables.

Differences in peak vertical ground reaction forces (dFzpeak) between contralateral forelimbs and hindlimbs are considered the gold standard for quantifying weight-bearing lameness. However, measuring kinematics for the same purpose is more common and practical. Vertical movement asymmetries (VMA) of the horse's upper body have previously been correlated to fore- and hindlimb lameness. But the combined response of head, withers and pelvis VMA to fore- and hindlimb dFzpeak has not yet been thoroughly investigated. Deriving the kinetic responses from kinematics would help the interpretation and understanding of quantified weight-bearing lameness. In this retrospective study, 103 horses with a wide range of fore- and hindlimb dFzpeak had been trotted on a force-measuring treadmill synchronized with an optical motion capture system. VMA of the head, withers and pelvis as well as dFzpeak were extracted. Multiple linear mixed models and linear regressions of kinematic variables were used to model the dFzpeak. It was hypothesised that all included VMA would have a significant influence on the dFzpeak outcome variables. The results showed a complex relationship between VMA and dFzpeak where both amplitude and timing of the VMA were of importance. On average, the contribution percentage of VMA to fore/hind dFzpeak were 66/34% for head, 76/24% for withers and 33/67% for pelvis. The linear regressions for the fore/hindlimb models achieved mean measurement root mean squared errors of 0.83%/0.82% dFzpeak. These results might help determine the clinical relevance of upper body VMA and distinguish between primary fore, hind, ipsilateral and diagonal weight-bearing lameness.

Continuous versus discrete data analysis for gait evaluation of horses with induced bilateral hindlimb lameness.

BACKGROUND: Gait kinematics measured during equine gait analysis are typically evaluated by analysing (asymmetry-based) discrete variables (eg, peak values) obtained from continuous kinematic signals (eg, timeseries of datapoints). However, when used for the assessment of complex cases of lameness, such as bilateral lameness, discrete variable analysis might overlook relevant functional adaptations. OBJECTIVES: The overall aim of this paper is to compare continuous and discrete data analysis techniques to evaluate kinematic gait adaptations to lameness. STUDY DESIGN: Method comparison. METHODS: Sixteen healthy Shetland ponies, enrolled in a research programme in which osteochondral defects were created on the medial trochlear ridges of both femurs, were used in this study. Kinematic data were collected at trot on a treadmill before and at 3 and 6 months after surgical intervention. Statistical parametric mapping and linear mixed models were used to compare kinematic variables between and within timepoints. RESULTS: Both continuous and discrete data analyses identified changes in pelvis and forelimb kinematics. Discrete data analyses showed significant changes in hindlimb and back kinematics, where such differences were not found to be significant by continuous data analysis. In contrast, continuous data analysis provided additional information on the timing and duration of the differences found. MAIN LIMITATIONS: A limited number of ponies were included. CONCLUSIONS: The use of continuous data provides additional information regarding gait adaptations to bilateral lameness that is complementary to the analysis of discrete variables. The main advantage lies in the additional information regarding time dependence and duration of adaptations, which offers the opportunity to identify functional adaptations during all phases of the stride cycle, not just the events related to peak values.

The Challenges of Equestrian Arena Surfaces: The Unprecedented Use of a Raised Platform at the 2012 Olympic Games.

The design of equestrian arenas can be challenged by time constraints and specific restrictions at a venue but are nonetheless a critical element to the success and sustainability of equestrian sport. The equestrian arenas for the 2012 Olympic Games were an example of a temporary arena constructed on a raised platform and supported by struts, a design unprecedented for equestrian activities. This study assessed the developmental stages of the Olympic surfaces from 2011 to the actual event in 2012 and aimed to confirm that accelerations and forces experienced by horses were comparable to those on solid ground. Assessment took place at (1) the Olympic test event;(2) a developmental mock-up arena; and (3) the Olympic venue in 2012. A Clegg impact hammer measured peak vertical deceleration and an Orono Biomechanical Surface Tester quantified peak load and peak loading rate. General Linear Models using the arena's structural features as explanatory variables highlighted surface heterogeneity. Peak vertical deceleration (P < .0001) and peak load (P < .0001) were significantly higher and peak loading rate was significantly lower (P < .0001) following iterative testing and modifications to the arena. Data were comparable with surfaces on solid ground by the final testing at the 2012 Olympic Games. Findings highlighted the importance of testing surfaces throughout their development and demonstrated the impact that surface composition, time elapsed since installation, water management, and type of construction have on surface functional properties, with relevance to future temporary arena initiatives.

Effect of Speed and Surface Type on Individual Rein and Combined Left-Right Circle Movement Asymmetry in Horses on the Lunge.

Differences in movement asymmetry between surfaces and with increasing speed increase the complexity of incorporating gait analysis measurements from lunging into clinical decision making. This observational study sets out to quantify by means of quantitative gait analysis the influence of surface and speed on individual-rein movement asymmetry measurements and their averages across reins (average-rein measurements). Head, withers, and pelvic movement asymmetry was quantified in 27 horses, identified previously as presenting with considerable movement asymmetries on the straight, during trot in hand and on the lunge on two surfaces at two speeds. Mixed linear models (p < 0.05) with horse as the random factor and surface and speed category (and direction) as fixed factors analyzed the effects on 11 individual-rein and average-rein asymmetry measures. Limits of agreement quantified differences between individual-rein and average-rein measurements. A higher number of individual-rein asymmetry variables-particularly when the limb that contributed to movement asymmetry on the straight was on the inside of the circle-were affected by speed (nine variables, all p ≤ 0.047) and surface (three variables, all p ≤ 0.037) compared with average-rein asymmetry variables (two for speed, all p ≤ 0.003; two for surface, all p ≤ 0.046). Six variables were significantly different between straight-line and average-rein assessments (all p ≤ 0.031), and asymmetry values were smaller for average-rein assessments. Limits of agreement bias varied between +0.4 and +4.0 mm with standard deviations between 3.2 and 12.9 mm. Fewer average-rein variables were affected by speed highlighting the benefit of comparing left and right rein measurements. Only one asymmetry variable showed a surface difference for individual-rein and average-rein data, emphasizing the benefit of assessing surface differences on each rein individually. Variability in straight-line vs. average-rein measurements across horses and exercise conditions highlight the potential for average-rein measurements during the diagnostic process; further studies after diagnostic analgesia are needed.

Towards Machine Recognition of Facial Expressions of Pain in Horses.

Automated recognition of human facial expressions of pain and emotions is to a certain degree a solved problem, using approaches based on computer vision and machine learning. However, the application of such methods to horses has proven difficult. Major barriers are the lack of sufficiently large, annotated databases for horses and difficulties in obtaining correct classifications of pain because horses are non-verbal. This review describes our work to overcome these barriers, using two different approaches. One involves the use of a manual, but relatively objective, classification system for facial activity (Facial Action Coding System), where data are analyzed for pain expressions after coding using machine learning principles. We have devised tools that can aid manual labeling by identifying the faces and facial keypoints of horses. This approach provides promising results in the automated recognition of facial action units from images. The second approach, recurrent neural network end-to-end learning, requires less extraction of features and representations from the video but instead depends on large volumes of video data with ground truth. Our preliminary results suggest clearly that dynamics are important for pain recognition and show that combinations of recurrent neural networks can classify experimental pain in a small number of horses better than human raters.

Non-banked curved tracks influence movement symmetry in two-year-old Standardbred trotters.

BACKGROUND: Little is known regarding how trotting through curves affects locomotion symmetry in Standardbred trotters. OBJECTIVES: To investigate differences in objectively measured Standardbred trotter vertical motion symmetry between straight and non-banked, curved sections of oval trotting tracks during exercise warm-up, using a wireless inertial measurement unit (IMU) system. STUDY DESIGN: Cross-sectional, observational study. METHODS: Sixteen horses were included. Mixed models were used to assess associations between symmetry, track segment (straight vs curve) and stride duration. RESULTS: Significant results for forelimb parameters were dependent on interactions between track segments and stride duration. At mean stride duration (0.611 second), during the curved track segment horses showed a lower maximum vertical position of the head after push-off of the outside forelimb (estimate -2.3 mm, P < 0.0001, 95% CI -1.7 to -2.9) and higher minimum vertical position of the head during stance of the outside forelimb (estimate -1.8 mm, P < 0.0001, 95% CI -1.2 to -2.5) compared to straight track, mimicking outside forelimb impact and push-off asymmetry during track curves. For hindlimb parameters, during the curve there was a decreased downward motion of the pelvis during outer hindlimb stance (estimate-0.7 mm, P < 0.0001, 95% CI -0.4 to -1.0), mimicking outside hindlimb impact asymmetry. MAIN LIMITATIONS: Horses were evaluated going in one direction only on the track (clockwise). CONCLUSIONS: Systematic differences between straight and curved track segments were found but did not fully correspond to previously described findings for horses lunged in circles. Effect sizes were overall small. Data in our study were collected from horses trotting on 1000 m tracks with curve radii of 80-85 m. On non-banked tracks of this size, collecting IMU symmetry data at jogging speeds without distinguishing between straight and curved parts is unlikely to adversely affect clinical decision-making.

Objectively measured movement asymmetry in yearling Standardbred trotters.

BACKGROUND: Lameness evaluation of Standardbred trotters can be challenging due to discrepancies in observed movement asymmetry between in-hand and track exercise, and between different trotting speeds. There are few studies on objective measurement of movement in Standardbreds, and little knowledge regarding biological variation and clinical significance of measured movement asymmetry in this breed. OBJECTIVES: To quantify the prevalence and magnitude of objectively measured movement asymmetry in young Standardbred trotters, and identify associations with trainer, sex, height, track type and in-hand measurement prior to or after track trials. STUDY DESIGN: Cross-sectional, observational study. METHODS: A total of 114 Standardbred yearlings were evaluated with a wireless inertial sensor system during trot in-hand and when driven on a track. After exclusions relating to lameness or technical difficulties, 103 horses were included in the study; 77 were evaluated in-hand and on the track, 24 only in-hand and 2 only on the track. RESULTS: Front and/or hindlimb parameters were above asymmetry thresholds previously established for other breeds during in-hand trials for 94 (93%) horses and during track trials for 74 (94%) horses. Most horses showed mild asymmetry. A minority of horses (20%) switched side of the asymmetry for one or more parameters between in-hand and track trials. Mixed model analyses revealed no significant effects of trial mode (in-hand or track trial, in-hand trial pre- or post-track trial, straight or oval track), trainer or horse height. Females had a significant but small reduction in asymmetry in one front limb parameter (HDmax ) compared with males (1.7 mm, 95% CI 0.18-3.28, P = .03). MAIN LIMITATIONS: High data variability, reflected in large trial standard deviations, relating mainly to a lack of horse compliance. CONCLUSIONS: A high proportion of Standardbred yearlings showed movement asymmetries. There was no group-level effect between in-hand and track trials, however, considerable individual variation was observed.

Asymmetries of horses walking and trotting on treadmill with and without rider.

BACKGROUND: Left-right movement symmetry is a highly desirable characteristic in sport horses. OBJECTIVES: This study compared movement symmetry in well-trained dressage horses in unridden and unrestrained position and ridden in a dressage frame, and investigated possible associations between gaits. STUDY DESIGN: Experimental study. METHODS: Seven sound, high-level dressage horses were measured at walk and sitting trot on a treadmill at several speeds under two conditions: with and without rider. Left-right differences in stance duration, stance protraction and retraction based on longitudinal hoof positions, ipsilateral limb tracking, minimum and maximum vertical positions of the dorsal spinous processes of the sixth thoracic (T6), third sacral vertebrae (S3) and wing of atlas, and vertical ground reaction forces were calculated and analysed in mixed models. RESULTS: At walk, five body variables indicated increased asymmetry in the ridden condition compared with unridden condition: forelimb stance duration (unridden/ridden left-right differences 9 vs 13 ms; P = .008), forelimb stance protraction (P = .004), stance retraction (P = .001) and first force peak (P = .003), and hindlimb stance retraction (P = .01). At trot, six body variables were more asymmetrical in the ridden condition: forelimb stance duration (2.5 vs 3.8 ms, P = .004); hindlimb stance protraction (P < .0001) and retraction (P = .01), T6 minimum (4 vs 6 mm, P = .001), T6 maximum (9 vs 11 mm, P = .01) and S3 maximum (6 vs 12 mm, P < .001). Five variables had significant associations between asymmetries at walk and trot, but only three demonstrated a positive slope. MAIN LIMITATIONS: A limited number of horses and riders were studied. Measurements were performed on a treadmill. CONCLUSIONS: High-level horses moved slightly more asymmetrically when ridden in a dressage frame than in the unridden condition.