Comparison of speed-dependent time, force and spatial parameters between Franches-Montagnes and European Warmblood horses walking and trotting on a treadmill.

Speed alterations affect many gait analysis parameters. How horses adapt to speed is relevant in many equestrian disciplines and may differ between breeds. This study described changes in gait parameters in 38 Warmblood (WB) and 24 Franches-Montagnes (FM) horses subjected to an incremental speed test at walk (1.35-2.05 m/s) and trot (3.25-5.5 m/s). Time, force and spatial parameters of each limb were measured with an instrumented treadmill and analysed with regression analysis using speed as the independent variable. With higher speeds, stride rate, length, over-tracking distance and vertical ground reaction forces increased while the impulses decreased. The parameters followed the same linear or polynomial regression curves independent of breed, while the slope (linear) or incurvation (polynomial) often differed significantly between breeds. Some differences between the breeds were associated with height and speed (e.g. stride length at walk), and would disappear when scaling the data. The main differences between the breeds seem to stem from the movement of the hind limbs, with the FM obtaining long over-tracking distances despite the shorter height at withers. Some parameters relevant to gait quality could be improved in the FM to resemble WB movement by strict selection using objective measurements systems.

New genomic insights into the conformation of Lipizzan horses.

Conformation traits are important selection criteria in equine breeding, as they describe the exterior aspects of the horse (height, joint angles, shape). However, the genetic architecture of conformation is not well understood, as data of these traits mainly consist of subjective evaluation scores. Here, we performed genome-wide association studies on two-dimensional shape data of Lipizzan horses. Based on this data, we identified significant quantitative trait loci (QTL) associated with cresty neck on equine chromosome (ECA)16 within the MAGI1 gene, and with type, hereby differentiating heavy from light horses on ECA5 within the POU2F1 gene. Both genes were previously described to affect growth, muscling and fatty deposits in sheep, cattle and pigs. Furthermore, we pin-pointed another suggestive QTL on ECA21, near the PTGER4 gene, associated with human ankylosing spondylitis, for shape differences in the back and pelvis (roach back vs sway back). Further differences in the shape of the back and abdomen were suggestively associated with the RYR1 gene, involved in core muscle weakness in humans. Therefore, we demonstrated that horse shape space data enhance the genomic investigations of horse conformation.

Prediction of continuous and discrete kinetic parameters in horses from inertial measurement units data using recurrent artificial neural networks.

Vertical ground reaction force (GRFz) measurements are the best tool for assessing horses' weight-bearing lameness. However, collection of these data is often impractical for clinical use. This study evaluates GRFz predicted using data from body-mounted IMUs and long short-term memory recurrent neural networks (LSTM-RNN). Twenty-four clinically sound horses, equipped with IMUs on the upper-body (UB) and each limb, walked and trotted on a GRFz measuring treadmill (TiF). Both systems were time-synchronised. Data from randomly selected 16, 4, and 4 horses formed training, validation, and test datasets, respectively. LSTM-RNN with different input sets (All, Limbs, UB, Sacrum, or Withers) were trained to predict GRFz curves or peak-GRFz. Our models could predict GRFz shapes at both gaits with RMSE below 0.40 N.kg-1. The best peak-GRFz values were obtained when extracted from the predicted curves by the all dataset. For both GRFz curves and peak-GRFz values, predictions made with the All or UB datasets were systematically better than with the Limbs dataset, showing the importance of including upper-body kinematic information for kinetic parameters predictions. More data should be gathered to confirm the usability of LSTM-RNN for GRFz predictions, as they highly depend on factors like speed, gait, and the presence of weight-bearing lameness.

Improving gait classification in horses by using inertial measurement unit (IMU) generated data and machine learning.

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.