Inter-evaluator and Intra-evaluator Reliability of a Software Program Used to Extract Kinematic Variables Obtained by an Extremity-Mounted Inertial Measurement Unit System in Sound Horses at the Trot Under Soft and Hard Ground Conditions and Treadmill Exercise.
Authors: Schwarz Julia, Vidondo Beatriz, Maninchedda Ugo E, Sprick Miriam, Schöpfer Melina C, Cruz Antonio M
Journal: Frontiers in veterinary science
Summary
# Editorial Summary Researchers assessed whether a software program reliably extracts kinematic data from extremity-mounted inertial measurement units (IMUs) in horses, a question that matters considerably if such systems are to become trustworthy clinical tools in equine practice. Thirty sound French Montagne stallions were trotted on three different surfaces—treadmill, soft ground and hard ground—with six IMUs positioned on the distal metacarpal, metatarsal and tibial regions of each limb; six evaluators (three experienced, three inexperienced) independently analysed the captured data three times each, with measurements assessed using intra-class correlation coefficients and variance partitioning coefficients. All tested gait variables demonstrated high reliability both within individual evaluators and between different evaluators, with negligible differences across the three surface types, and critically, evaluator experience had no bearing on result quality. The software's robust inter-evaluator and intra-evaluator reliability—independent of ground conditions and operator expertise—suggests that with proper standardised application protocols, IMU-based gait analysis could transition from research settings into practical clinical use for lameness assessment, rehabilitation monitoring and performance evaluation, provided farriers, vets and physiotherapists receive appropriate training in correct data collection technique.
Read the full abstract on PubMed
Practical Takeaways
- •This IMU-based gait analysis software is reliable for consistent extraction of kinematic data regardless of ground surface, making it suitable for field use and clinical monitoring
- •Training level doesn't significantly affect measurement quality—both experienced and inexperienced operators can produce reliable data when following correct procedures
- •The tool shows promise for tracking individual horse gait changes over time and detecting subtle lameness or performance issues, provided standardized application protocols are followed
Key Findings
- •Software for extracting kinematic variables from extremity-mounted IMU systems demonstrated high inter-evaluator and intra-evaluator reliability across all tested gait variables (ICC values high)
- •No significant difference in reliability between soft ground, hard ground, and treadmill surfaces
- •Evaluator experience level (experienced vs. inexperienced) had no influence on software reliability results
- •Horse-to-horse variation accounted for the majority of variance in gait parameters, confirming software sensitivity to genuine individual differences rather than evaluator error