Accelerometry-Based Step Count Validation for Horse Movement Analysis During Stall Confinement

Area of Science:

• Equine science
• Biomechanics
• Veterinary technology

Background:

• Quantitative tracking of equine movement during stall confinement can detect subtle mobility changes indicative of injury.
• Early detection of mobility changes is vital for directing rehabilitation protocols in horses.
• Inertial measurement units (IMUs) offer a readily available, objective method for recording equine step counts.

Purpose of the Study:

• To compare IMU-based step counts with video analysis for three distinct equine movements in a stall.
• To evaluate three different sensor positions for IMUs to identify the optimal location for assessing horse movement.
• To determine the ideal sensor placement for accurate step counting during stall confinement.

Main Methods:

• An IMU was attached to six horses at the withers, right forelimb, and hindlimb.
• Data were collected during free movement, circles, and figure-eights in 5-minute intervals.
• Accuracy was assessed by comparing IMU step counts to a video-based criterion measure using mean difference, absolute error, and intraclass correlation coefficients (ICCs).

Main Results:

• The right-forelimb vertical-axis IMU position yielded the best results (ICC = 1.0, % error = 6.8, mean difference = 1.3).
• The right-hindlimb position also showed high accuracy (ICC = 0.999, % error = 15.2, mean difference = 1.8).
• Excellent agreement was observed between IMU vertical axis data and video analysis at both forelimb and hindlimb locations across all movement types.

Conclusions:

• IMUs demonstrate high accuracy in quantifying equine steps during stall confinement when placed on the limbs.
• The forelimb and hindlimb sensor positions are suitable for monitoring horse mobility and detecting subtle changes.
• This technology has potential for early detection of complications and guiding rehabilitation strategies in horses.