Comparison of an inertial sensor based system of lameness quantification to subjective lameness evaluation

Abstract

Recently developed wireless inertial sensor systems provide a means to objectively detect lameness as a horse is trotted over natural footing. The purpose of this study was to compare one such inertial sensor based system to subjective evaluation performed by experienced equine practitioners of horses with lameness induced by sole pressure. Fifteen horses were fitted with special shoes, which allowed for lameness induction via sole pressure. Before each trial the horses were fitted with three wireless inertial sensors. Horses were subjected to multiple trials: 1) immediately before inserting the screw, 2) immediately after inserting the screw to just touch the sole, and 3) after gradually tightening the screw in half turn increments. The number of half turns of the screw required for consistent identification of lameness in the correct limb by the inertial sensors and through the consensus of three equine veterinarians was compared using the Wilcoxon's sum rank test. The inertial sensor system selected the correct limb sooner (i.e., after few screw turns) than the veterinarians (P<0.0001). The 3 veterinarians selected the correct limb before the inertial sensors in five trials (8.3%). The opposite occurred in 43 trials (71.6%). The 3 veterinarians and the computer simultaneously selected the correct limb in 12 trials (20%). The inertial sensor based system was more sensitive relative to the consensus of 3 equine veterinarians to identify the affected limb in horses subjected to sole pressure induced lameness.