Classification of racehorse limb radiographs using deep convolutional neural networks

Purpose

To assess the capability of deep convolutional neural networks to classify anatomical location and projection from a series of 48 standard views of racehorse limbs.

Materials and methods

Radiographs (N = 9504) of horse limbs from image sets made for veterinary inspections by 10 independent veterinary clinics were used to train, validate and test (116, 40 and 42 radiographs, respectively) six deep learning architectures available as part of the open source machine learning framework PyTorch. The deep learning architectures with the best top-1 accuracy had the batch size further investigated.

Results

Top-1 accuracy of six deep learning architectures ranged from 0.737 to 0.841. Top-1 accuracy of the best deep learning architecture (ResNet-34) ranged from 0.809 to 0.878, depending on batch size. ResNet-34 (batch size = 8) achieved the highest top-1 accuracy (0.878) and the majority (91.8%) of misclassification was due to laterality error. Class activation maps indicated that joint morphology, not side markers or other non-anatomical image regions, drove the model decision.

Conclusions

Deep convolutional neural networks can classify equine pre-import radiographs into the 48 standard views including moderate discrimination of laterality, independent of side marker presence.

What’s in a warm-up? A preliminary investigation of how European dressage riders and show jumpers warm-up their horses for training and at competition

Equestrian sports such as dressage and show jumping cause physical and physiological stress on the horses’ musculoskeletal structures, which can lead to decreased performance and injury. Warming-up prior to intense exercise can increase utilisation of the aerobic pathway, increase performance and decrease injury risk. Whilst duration of equestrian warm-up regimes has been reported, details of which gaits and skills related tasks, such as jumping and lateral movements, riders elect to use have not been evaluated to date. The purpose of this study was to understand dressage and show jumping riders’ decision-making when warming up at home and prior to a competition. Surveys (dressage: 39 questions; show jumping: 41 questions) were distributed online via social media. Mann Whitney U tests identified significant differences in warming up practice between dressage and show jumping riders. Most riders reported that a warm-up was beneficial for getting the horse ready for work, increasing responsiveness to aids, enhancing suppleness and relaxation, and decreasing injury risk. Both dressage and show jumping riders typically warm-up between 10-20 min. While dressage riders use the walk as their main warm-up gait, show jumpers preferred the trot. Both dressage riders and show jumpers incorporate technical skills in their warm-up such as lateral work, and quick transitions (when riders change gait for only few strides before changing again). Show jumpers include 4-10 jumping efforts, using different fence types. During a competition most dressage and show jumping riders agreed that factors such as perceived stress level of both the horse and rider, crowdedness of the arena, arena footing and size, as well as time allocated by the venue, were important factors that could impact the duration and content of their warm-up routines. Both groups of riders considered horses were sufficiently ‘warmed up’ when they were responsive to the aids and felt supple and relaxed.