Fifty ways to do your lameness: the relationship between various pelvis-based visual lameness pointers during hindlimb lameness

Reasons for performing study:

The relationship between pelvic hindlimb lameness pointers remains unclear.

Objectives:

To examine sacrum and tubera coxae (TC) based movement asymmetry across different lameness degrees, patterns and changes in pelvic rotation.

Study design:

Parameter regression and rigid body model.

Methods:

Vertical displacement of left and right tubera coxae (LTC, RTC) and sacrum was quantified experimentally in 107 horses with varying lameness degrees using inertial sensors. To investigate the theoretical relationship between pointers, a geometrical model of pelvis movement was created in Matlab. The model predicts TC movement based on experimentally derived pelvic rotation and computer-generated asymmetrical sacrum movement. Further, the effect of changes in pelvic rotation on TC-based lameness pointers was quantified. For both, experimental and model data, asymmetry measures corresponding to visual assessment strategies of veterinarians were calculated and contrasted.

Results:

Experimental data showed that upward movement amplitudes for sacrum and TC were strongly correlated (r≥0.96). Regression showed that differences were greater between amplitudes of LTC/RTC than sacrum (slope 1.18; r2 = 0.94; P<0.001); however, for ratios the reverse held true (slope 0.86; r2 = 0.92; P<0.001). The model predicted experimentally determined tubera coxae asymmetry well. While the model data matched the experimental scaling relationship between amplitude ratios, it predicted a slightly shallower slope for amplitude differences. Perturbing the model with a different rotational range of movement, an offset in pelvic rotation and asymmetrical pelvic rotation affected observable TC-based asymmetry depending on the chosen assessment strategy.

Conclusions:

Detecting lameness based on sacrum or comparative TC movement can give similar results. Using comparative TC movement for lameness detection would be susceptible to changes in pelvic rotation which a lame horse may adopt. Hence, assessment of multiple pelvis-based lameness pointers can help establishing nonobvious hindlimb lameness.