Aston Business School, Aston University, Birmingham, UK.
Clinical Science and Services, The Royal Veterinary College, Hatfield, UK.
Vet Rec. 2021 May;188(10):e21. doi: 10.1002/vetr.21. Epub 2021 Mar 1.
Visual assessment of equine lameness is an everyday veterinary task suffering from poor diagnostic accuracy. The aim of this study was to quantify the impact of the perceptual learning game 'LamenessTrainer' on skill development.
Thirty-six undergraduate veterinary students engaged in four game modules teaching the assessment of fore- and hindlimb lameness. Computer animations of horses in this game displayed 0% (sound) to 70% (moderately lame) vertical movement asymmetry of head and pelvis. Performance, learning effects, diagnostic accuracy, detection thresholds and survey responses were analysed.
Following staircase learning, more than 80% of students reliably classified horses with ≥20% asymmetry for forelimb lameness, ≥40% asymmetry for simplified hindlimb lameness and ≥50% asymmetry for realistic hindlimb lameness. During random presentation, on average 82% of sound and 65% of lame horses were assessed correctly during forelimb lameness evaluation, dropping to 39% of sound and 56% of lame horses for hindlimb lameness.
In less than two hours, systematic perceptual learning through deliberate practice can develop visual assessment skills to an accuracy level comparable to expert assessors scoring the same animations. Skills should be developed further to improve misclassifications of sound and mildly lame horses, especially for hindlimb lameness evaluation.
马的跛行的视觉评估是兽医日常工作，但诊断准确性较差。本研究的目的是量化感知学习游戏“跛行训练器”对技能发展的影响。
36 名本科兽医学生参与了四个游戏模块，教授前肢和后肢跛行的评估。该游戏中的马的计算机动画显示头和骨盆的垂直运动不对称度为 0%（正常）至 70%（中度跛行）。分析了表现、学习效果、诊断准确性、检测阈值和调查响应。
经过阶梯式学习，超过 80%的学生能够可靠地将前肢跛行≥20%、简化后肢跛行≥40%、真实后肢跛行≥50%的马进行分类。在随机呈现中，在前肢跛行评估中，平均 82%的正常马和 65%的跛马被正确评估，而后肢跛行评估中，正常马和跛马的正确评估率分别下降至 39%和 56%。
通过有目的的练习进行系统的感知学习，在不到两个小时的时间内，可以将视觉评估技能提高到与评估相同动画的专家评估者相当的准确性水平。应进一步提高技能，以减少对正常和轻度跛行马的错误分类，特别是在后肢跛行评估中。
