Search

For centuries humans have been fascinated by the natural beauty of horses in motion and their different gaits. Gait classification (GC) is commonly performed through visual assessment and reliable, automated methods for real-time objective GC in horses are warranted. In this study, we used a full body network of wireless, high sampling-rate sensors combined with machine learning to fully automatically classify gait. Using data from 120 horses of four different domestic breeds, equipped with seven motion sensors, we included 7576 strides from eight different gaits. GC was trained using several machine-learning approaches, both from feature-extracted data and from raw sensor data. Our best GC model achieved 97% accuracy. Our technique facilitated accurate, GC that enables in-depth biomechanical studies and allows for highly accurate phenotyping of gait for genetic research and breeding. Our approach lends itself for potential use in other quadrupedal species without the need for developing gait/animal specific algorithms. [ABSTRACT FROM AUTHOR]

An amendment to this paper has been published and can be accessed via a link at the top of the paper. [ABSTRACT FROM AUTHOR]

Summary: Background: Quantitative gait analysis is rapidly gaining ground in equine practice, and pros and cons are regularly discussed within the scientific literature. However, no data exist on the appreciation of the technique by equine clinicians, their motivation to use it or not, and their perception of its value in daily practice. Objectives: To make a first inventory of opinions, expectances and experiences of equine veterinarians concerning the use of quantitative gait analysis in their daily general practice. Study design: Survey. Methods: A questionnaire was sent out to a group of equine orthopaedic clinicians working in an equine clinic or practice. Respondents were classified as users (having clinical experience with quantitative gait analysis) or not (nonusers). Data were analysed using descriptive statistics. Results: Within the sample population, users were more positive about the usefulness of quantitative gait analysis than nonusers. Veterinarians who purchased a system were motivated by better objectivity, transparency, documentation and client service. Main reasons not to purchase a system were costs and complexity of data interpretation. A minority of both users and nonusers deemed quantitative gait analysis also suitable for equine professionals other than veterinarians. Main limitations: Users (n = 40) outnumbered nonusers (n = 32), sample size was limited (n = 72) and insufficient to allow for generalisation of results. Conclusions: Users of quantitative gait analysis were more positive about the technology than nonusers. More data are needed to allow for generalisation of the results. Regularly repeating this survey may help in monitoring, and eventually guiding the process of integration of gait analysis technology within equine clinical practice by providing valuable information for individual clinics, educational institutions and the industry producing this technology. [ABSTRACT FROM AUTHOR]

Work on curved tracks, e.g. on circles, is commonplace within all forms of horse training. Horse movements in circles are naturally asymmetric, including the load distribution between inner and outer limbs. Within equestrian dressage the horse is expected to bend the back laterally to follow the circle, but this has never been studied scientifically. In the current study 12 horses were measured (optical motion capture, 100 Hz) trotting on left and right circles and on the straight without rider (soft surface). Data from markers placed along the spine indicated increased lateral bending to the inside (e.g. left bending on the left circle) of the thoracolumbar back (difference left circle vs. straight − 3.75°; right circle + 3.61°) and the neck (left − 5.23°; right + 4.80° vs. straight). Lateral bending ROM increased on the circle (+ 0.87° and + 0.62°). Individual variation in straight-circle differences was evident, but each horse was generally consistent over multiple trials. Differences in back movements between circle and straight were generally small and may or may not be visible, but accompanying changes in muscle activity and limb movements may add to the visual impression. [ABSTRACT FROM AUTHOR]

Background: Lameness assessment in horses is still predominantly performed using subjective methods. Visual assessment is known to have moderate to good intra-rater agreement but relatively poor inter-rater agreement. Little is known about inter- and intra-rater agreement on the evaluation of back motion, for which no objective measurement technique in a clinical setting is available thus far. Objectives: To describe inter- and intra-rater agreement of visual evaluation of equine back mobility. Study design: Rater reliability study using a fully crossed design in which all horses are rated by all observers. This data is compared with objective gait analysis. Methods: Seventy equine professionals (veterinarians and physiotherapists) and veterinary students evaluated videos of 12 healthy horses at walk and trot on a hard, straight line. Nine parameters related to back mobility were scored: general mobility, thoracic, lumbar, lumbosacral flexion and extension and left and right thoracolumbar latero-flexion. All parameters were compared with simultaneously measured quantitative motion parameters. After 1 month, six randomly chosen horses were re-evaluated by 57 observers. Results: For each parameter inter- and intra-rater agreements were calculated using intra-class correlation coefficients. For all parameters, inter-rater agreement was very poor (<0.2). The mean intra-rater agreement of all observers and for all parameters was poor (~0.4) but varied between 0.0 and 0.96 for individual observers. There was no correlation between the visual subjective scoring and objective gait analysis measurements. Main limitations: Horses were scored from videos and by lack of any existing (semi-) quantitative system, a custom-made system had to be used. Conclusions: The poor inter- and intra-rater agreements of visual scoring of mobility of the equine back and the disagreement between subjective and objective gait analysis data, demonstrate the need for the development and introduction of objective, quantitative and repeatable techniques to assess equine back motion. [ABSTRACT FROM AUTHOR]

In horses, hip hike asymmetry, i.e. left–right difference in hip upwards movement during hind limb protraction in trot, is a crucial lameness sign. Vertical hip movements are complex, influenced by both pelvic roll and pelvic vertical motion. Veterinarians find it challenging to identify low-grade lameness, and knowledge of normal variation is a prerequisite for discerning abnormalities. This study, which included 100 clinically sound Warmblood horses, aimed to describe normal variation in pelvic roll stride patterns. Data were collected during straight-line trot in hand using optical motion capture. Stride-segmented pelvic roll data, normalised with respect to time (0–100% of the stride) and amplitude (± 0.5 of horse average stride range of motion), were modelled as a linear combination of sine and cosine curves. A sine curve with one period per stride and a cosine curve with three periods per stride explained the largest proportions of roll motion: model estimate 0.335 (p < 0.01) and 0.138 (p < 0.01), respectively. Using finite mixture models, the horses could be separated into three groups sharing common pelvic roll characteristics. In conclusion, pelvic roll motion in trot follows a similar basic pattern in most horses, yet there is significant individual variation in the relative prominence of the most characteristic features. [ABSTRACT FROM AUTHOR]

In horses, hip hike asymmetry, i.e. left–right difference in hip upwards movement during hind limb protraction in trot, is a crucial lameness sign. Vertical hip movements are complex, influenced by both pelvic roll and pelvic vertical motion. Veterinarians find it challenging to identify low-grade lameness, and knowledge of normal variation is a prerequisite for discerning abnormalities. This study, which included 100 clinically sound Warmblood horses, aimed to describe normal variation in pelvic roll stride patterns. Data were collected during straight-line trot in hand using optical motion capture. Stride-segmented pelvic roll data, normalised with respect to time (0–100% of the stride) and amplitude (± 0.5 of horse average stride range of motion), were modelled as a linear combination of sine and cosine curves. A sine curve with one period per stride and a cosine curve with three periods per stride explained the largest proportions of roll motion: model estimate 0.335 (p < 0.01) and 0.138 (p < 0.01), respectively. Using finite mixture models, the horses could be separated into three groups sharing common pelvic roll characteristics. In conclusion, pelvic roll motion in trot follows a similar basic pattern in most horses, yet there is significant individual variation in the relative prominence of the most characteristic features. [ABSTRACT FROM AUTHOR]

The reliability of surface electromyography (sEMG) has not been adequately demonstrated in the equine literature and is an essential consideration as a methodology for application in clinical gait analysis. This observational study investigated within-session, intra-subject (stride-to-stride) and inter-subject reliability, and between-session reliability of normalised sEMG activity profiles, from triceps brachii (triceps), latissimus dorsi (latissimus), longissimus dorsi (longissimus), biceps femoris (biceps), superficial gluteal (gluteal) and semitendinosus muscles in n = 8 clinically non-lame horses during in-hand trot. sEMG sensors were bilaterally located on muscles to collect data during two test sessions (session 1 and 2) with a minimum 24-hour interval. Raw sEMG signals from ten trot strides per horse and session were DC-offset removed, high-pass filtered (40 Hz), full-wave rectified, and low-pass filtered (25 Hz). Signals were normalised to peak amplitude and percent stride before calculating intra- and inter-subject ensemble average sEMG profiles across strides for each muscle and session. sEMG profiles were assessed using waveform similarity statistics: the coefficient of variation (CV) to assess intra- and inter-subject reliability and the adjusted coefficient of multiple correlation (CMC) to evaluate between-session reliability. Across muscles, CV data revealed that intra-horse sEMG profiles within- and between-sessions were comparatively more reliable than inter-horse profiles. Bilateral gluteal, semitendinosus, triceps and longissimus (at T14 and L1) and right biceps showed excellent between-session reliability with group-averaged CMCs > 0.90 (range 0.90–0.97). Bilateral latissimus and left biceps showed good between-session reliability with group-averaged CMCs > 0.75 (range 0.78–0.88). sEMG profiles can reliably describe fundamental muscle activity patterns for selected equine muscles within a test session for individual horses (intra-subject). However, these profiles are more variable across horses (inter-subject) and between sessions (between-session reliability), suggesting that it is reasonable to use sEMG to objectively monitor the intra-individual activity of these muscles across multiple gait evaluation sessions at in-hand trot. [ABSTRACT FROM AUTHOR]

Clinical assessment of spinal motion in horses is part of many routine clinical exams but remains highly subjective. A prerequisite for the quantification of spinal motion is the assessment of the expected normal range of motion and variability of back kinematics. The aim of this study was to objectively quantify spinal kinematics and between -measurement, -surface and -day variation in owner-sound horses. In an observational study, twelve owner-sound horses were trotted 12 times on four different paths (hard/soft straight line, soft lunge left and right). Measurements were divided over three days, with five repetitions on day one and two, and two repetitions on day three (recheck) which occurred 28–55 days later. Optical motion capture was used to collect kinematic data. Elements of the outcome were: 1) Ranges of Motion (ROM) with confidence intervals per path and surface, 2) a variability model to calculate between-measurement variation and test the effect of time, surface and path, 3) intraclass correlation coefficients (ICC) to determine repeatability. ROM was lowest on the hard straight line. Cervical lateral bending was doubled on the left compared to the right lunge. Mean variation for the flexion-extension and lateral bending of the whole back were 0.8 and 1 degrees. Pelvic motion showed a variation of 1.0 (pitch), 0.7 (yaw) and 1.3 (roll) degrees. For these five parameters, a tendency for more variation on the hard surface and reduced variation with increased repetitions was observed. More variation was seen on the recheck (p<0.001). ICC values for pelvic rotations were between 0.76 and 0.93, for the whole back flexion-extension and lateral bending between 0.51 and 0.91. Between-horse variation was substantially higher than within-horse variation. In conclusion, ROM and variation in spinal biomechanics are horse-specific and small, necessitating individual analysis and making subjective and objective clinical assessment of spinal kinematics challenging. [ABSTRACT FROM AUTHOR]