Linear Discriminant Analysis for Investigating Differences in Upper Body Movement Symmetry in Horses before/after Diagnostic Analgesia in Relation to Expert Judgement.
Authors: Pfau Thilo, Bolt David M, Fiske-Jackson Andrew, Gerdes Carolin, Hoenecke Karl, Lynch Lucy, Perrier Melanie, Smith Roger K W
Journal: Animals : an open access journal from MDPI
Summary
# Editorial Summary Diagnostic analgesia is a cornerstone of equine lameness work-up, yet the subjective visual assessment of whether a horse responds negatively, partially, or completely to a regional nerve block can be clouded by examiner expectation bias. Researchers at multiple institutions employed inertial sensors to track upper body movement symmetry in 53 horses before and after diagnostic analgesia—recording data both on the straight and on the lunge over different ground surfaces—then applied linear discriminant analysis to determine whether quantitative movement asymmetry could reliably distinguish between these three response categories. Classification accuracy was modest (38–57% for forelimb blocks and 36–56% for hindlimb blocks depending on conditions), with notably better separation when the blocked limb was positioned on the inside of a circle on hard ground (forelimb) or soft ground (hindlimb). Head and pelvic upward displacement, along with withers minimum differences, emerged as the most discriminatory movement variables, indicating that diagnostic analgesia effects propagate throughout the upper body rather than being localised. The considerable overlap between response categories in the quantitative data suggests that movement changes after diagnostic analgesia are nuanced and multifaceted, highlighting both the value and the limitations of sensor-based assessment as a complement to—rather than replacement for—skilled clinical observation; practitioners should pay particular attention to whole-body kinematics on the lunge when evaluating diagnostic analgesia responses.
Read the full abstract on PubMed
Practical Takeaways
- •When using quantitative gait analysis to assess lameness responses to diagnostic analgesia, pay particular attention to upper body symmetry (head, pelvis, withers) rather than just distal limb movement.
- •Exercise conditions matter: test forelimb blocks on hard ground with the blocked leg on the inside of circles, and hindlimb blocks on soft ground, for clearest movement pattern changes.
- •Expect moderate accuracy (38-57%) when using movement data alone to classify negative, partial, or complete responses to diagnostic analgesia; this tool should complement rather than replace clinical judgment.
Key Findings
- •Leave-one-out classification accuracy for diagnostic analgesia response categories ranged from 38.3-57.4% for forelimb and 36.1-56.1% for hindlimb, indicating substantial overlap between subjective categories.
- •Best discrimination between diagnostic analgesia responses occurred with the blocked limb on the inside of the circle on hard ground for forelimb and soft ground for hindlimb conditions.
- •Head and pelvic upward movement and withers minimum differences were the most important movement asymmetry features in discriminating diagnostic analgesia responses across all conditions.
- •Quantitative gait analysis data shows considerable overlap with subjective veterinary judgement of diagnostic analgesia efficacy, suggesting data-driven approaches could help reduce expectation bias in lameness diagnosis.