Use of a Collar-Mounted Triaxial Accelerometer to Predict Speed and Gait in Dogs

Area of Science:

• Veterinary Science
• Biomechanics
• Animal Locomotion

Background:

• Accelerometry is used to assess osteoarthritis treatment efficacy in dogs, but its interpretation is challenging.
• Simplifying accelerometry data into gait or speed categories could enhance interpretability.

Purpose of the Study:

• To evaluate if collar-mounted accelerometers can estimate canine speed and categorize gait on a treadmill.
• To determine the feasibility of using accelerometry for objective gait analysis in dogs.

Main Methods:

• Eight Huntaway dogs were fitted with triaxial accelerometers and recorded via high-speed video on a treadmill.
• Data from accelerometers (delta-G) were synchronized with video and treadmill speed/gait records.
• Mixed linear and logistic regression models, including delta-G and skeletal size, predicted speed and gait.

Main Results:

• Gait was accurately categorized into walking versus faster gaits (trot/canter) with high reliability (pseudo-R2 = 0.87).
• Separate categorization of faster gaits (trot, canter) was not achieved.
• Speed estimation above 3 m/s proved inaccurate, with potential device or inherent limitations.

Conclusions:

• Collar-mounted accelerometry reliably categorizes canine gaits into two broad types: walking and faster.
• Precise speed estimation or differentiation of faster gaits requires individualized modeling and validation.
• This method can enhance the utility of accelerometry for detecting osteoarthritis treatment effects by focusing on activity periods most affected.