Classification performance of sEMG and kinematic parameters for distinguishing between non-lame and induced lameness conditions in horses.
Authors: St George Lindsay B, Spoormakers Tijn J P, Hobbs Sarah Jane, Clayton Hilary M, Roy Serge H, Richards Jim, Serra Bragança Filipe M
Journal: Frontiers in veterinary science
Summary
# Editorial Summary Whilst surface electromyography (sEMG) has demonstrated value in equine research settings, its clinical utility for lameness detection remains unproven—a gap this 2024 study addressed by comparing sEMG measurements against established kinematic asymmetry parameters in eight clinically sound horses subjected to induced fore- and hindlimb lameness (2–3/5 AAEP grade). Researchers collected bilateral sEMG and three-dimensional kinematic data during in-hand trot at baseline and following lameness induction via modified horseshoe, processing sEMG signals through high-pass filtering and full-wave rectification to generate both absolute values (sEMGabs) and left-right asymmetry metrics (sEMGasym), then evaluated discriminatory accuracy using receiver-operating-characteristic analysis. Hindlimb lameness detection proved substantially more reliable than forelimb detection across both sEMG approaches (AUC 0.97 versus 0.77 maximum), with sEMGabs outperforming asymmetry-based sEMG metrics; kinematic parameters (particularly poll MinDiff and pelvic asymmetry measures) remained superior overall, achieving AUC >0.95 for their respective limb classifications. The findings suggest that integrating sEMG neuromuscular data with conventional kinematic analysis could enhance clinical lameness assessment by revealing functional deficits at the muscular level, though current sEMG methodology requires refinement before standalone clinical application—particularly for detecting forelimb dysfunction where muscle recruitment patterns may differ substantially from hindlimb compensatory mechanisms.
Read the full abstract on PubMed
Practical Takeaways
- •sEMG shows promise for detecting hindlimb lameness in practice but cannot yet reliably detect forelimb lameness—kinematic assessment remains superior for forelimb evaluation
- •Single-parameter sEMG analysis is insufficient; absolute muscle activity values are more useful than asymmetry comparisons for lameness discrimination
- •Combining kinematic and sEMG data in a multivariate approach may eventually provide better diagnostic precision and monitoring during treatment than either method alone
Key Findings
- •sEMG absolute value (sEMGabs) parameters detected induced hindlimb lameness with AUC ≥0.97 but performed poorly for forelimb lameness (AUC ≤0.77)
- •sEMGabs outperformed sEMG asymmetry (sEMGasym) parameters for both forelimb and hindlimb lameness detection
- •Kinematic asymmetry parameters (MinDiff Poll, Hip Hike) maintained excellent discrimination for induced lameness (AUC >0.95) consistent with previous research
- •Combined multivariate approaches integrating kinematics and sEMG may improve comprehensive lameness diagnosis by measuring neuromuscular functional causes