Validation of distal limb mounted inertial measurement unit sensors for stride detection in Warmblood horses at walk and trot.
Authors: Bragança F M, Bosch S, Voskamp J P, Marin-Perianu M, Van der Zwaag B J, Vernooij J C M, van Weeren P R, Back W
Journal: Equine veterinary journal
Summary
# Editorial Summary Researchers at Utrecht University validated whether inertial measurement units (IMUs)—small accelerometers and gyroscopes mounted on the metacarpal and metatarsal bones—could accurately detect hoof contact and lift-off timing during walk and trot by comparing their readings against force plate measurements and motion capture data in seven Warmblood horses. Four custom algorithms were tested, with the best-performing versions achieving accuracies within 28.5 ms at walk and −27.6 ms at trot for forelimbs, translating to errors of only 3.7% and −8.4% respectively relative to actual stance duration. Whilst these results demonstrate that IMU sensors can reliably capture basic temporal stride parameters, the precision margins—particularly the 33.5 ms standard deviation for hindlimb trot data—may be insufficient to detect subtle lameness-related gait changes that clinicians need to identify in practice. The authors acknowledge significant limitations, notably the lack of testing across different surfaces and the absence of validation in lame horses, yet conclude the technology shows sufficient promise to justify further development as an objective locomotor assessment tool for performance and gait analysis.
Read the full abstract on PubMed
Practical Takeaways
- •IMU sensors mounted on distal limbs offer a portable, objective method for detecting stride timing in sound horses, potentially useful for pre-purchase evaluation and baseline gait documentation
- •Current precision limitations (±8–34 ms) suggest this technology is suitable for gross gait assessment but may miss early or subtle lameness indicators that require more sensitive detection
- •Before adopting IMU systems for lameness diagnosis in clinical practice, further validation on varied surfaces and lame populations is needed to confirm clinical utility
Key Findings
- •IMU sensors achieved accuracy within 28.5 ms and precision of 31.6 ms for forelimb stance detection at walk, with 3.7% error relative to force plate data
- •At trot, forelimb accuracy was -27.6 ms with 8.8 ms precision (-8.4% error), while hindlimbs showed 6.3 ms accuracy and 33.5 ms precision (9.1% error)
- •IMU technology can reliably determine temporal kinematic stride variables at walk and trot but may lack sufficient precision to detect subtle lameness-related gait changes
- •Algorithm performance was not validated across different surface types, limiting current generalizability