Effect of gamified perceptual learning on visual detection and discrimination skills in equine gait assessment.
Authors: S. Starke, Gregory C Miles, S. Channon, S. May
Journal: The Veterinary record
Summary
# Editorial Summary Visual assessment remains the cornerstone of equine lameness diagnosis, yet considerable inter-observer variability and diagnostic errors persist in veterinary practice. Starke and colleagues evaluated whether a computerised perceptual learning game ('LamenessTrainer') could accelerate skill development in detecting and discriminating between sound and lame horses, using 36 undergraduate veterinary students who completed four modules presenting animated horses with vertical movement asymmetries ranging from 0–70%. The gamified learning approach proved effective: following structured practice, over 80% of participants reliably detected forelimb asymmetries ≥20% and hindlimb asymmetries ≥40–50% depending on presentation complexity. When assessed on randomised animations, students achieved 82% accuracy identifying sound forelimbs and 65% for lame forelimbs, though hindlimb assessment proved more challenging (39% sound, 56% lame). Remarkably, this proficiency developed within two hours of deliberate practice, reaching accuracy levels comparable to expert assessors evaluating identical animations. The findings suggest significant potential for structured digital training to supplement traditional clinical teaching and standardise assessment competency across the profession. However, persistent misclassification of sound horses and mild lameness cases—particularly in hindlimb evaluation—indicates that training protocols should incorporate greater emphasis on subtle movement abnormalities and transition zones between normal and pathological gait, informing both undergraduate curricula and continuing professional development programmes.
Read the full abstract on PubMed
Practical Takeaways
- •Gamified training tools like LamenessTrainer can rapidly improve visual lameness detection skills in veterinary students and may provide a scalable method for improving diagnostic accuracy in clinical practice
- •Practitioners should recognize that hindlimb lameness assessment is inherently more difficult than forelimb evaluation and requires additional training focus, especially for detecting mild cases
- •Structured perceptual learning with deliberate practice is an efficient investment of training time—less than 2 hours can yield clinically meaningful improvements in gait assessment accuracy
Key Findings
- •After gamified perceptual learning, >80% of students reliably detected forelimb lameness at ≥20% vertical asymmetry and hindlimb lameness at ≥40-50% asymmetry depending on presentation complexity
- •During random presentation, diagnostic accuracy was 82% for sound horses and 65% for lame horses in forelimb assessment, but dropped to 39% sound and 56% lame for hindlimb assessment
- •Less than two hours of systematic perceptual learning through deliberate practice achieved diagnostic accuracy comparable to expert assessors on the same animations
- •Hindlimb lameness detection remains significantly more challenging than forelimb assessment, with higher rates of misclassification of sound and mildly lame horses