Assessment of mild hindlimb lameness during over ground locomotion using linear discriminant analysis of inertial sensor data.

The study focuses on the use of an inertial sensor system to detect and measure the degree of hindlimb lameness in horses. The system measures the movement of the tuber coxae in horses, providing data that allows for a more objective evaluation of lameness.

Objective and Background

• The objective of this study was to explore the potential of an inertial sensor system in detecting and quantifying hindlimb lameness in horses. Hindlimb lameness is a common problem in horses and can be tough to diagnose. An objective method to measure lameness is vital to accurately evaluate the effectiveness of treatment and to contribute to evidence-based medicine.
• Specifically, the research aimed to assess the movement of the tuber coxae, a part of the horse’s body that can provide useful data for distinguishing between lame and non-lame horses. This was measured during the horse’s over-ground movement, or locomotion.

Methodology

• Two inertial sensors were fixed on 21 horses, over their left and right tuber coxae. The sample included both non-lame and mildly lame horses.
• The horses were made to trot on a hard surface, during which data from a total of 1021 strides were collected from the sensors. The sensors captured six degrees of freedom of movement, primarily focusing on the dorsoventral and craniocaudal movements.
• The collected data, featuring 34 distinguishing aspects, were then analyzed to classify the horses into various groups, specifically lame vs. non-lame and right lame, left lame, and non-lame.

Results

• The sensor system successfully gathered continuous kinematic data from the left and right tuber coxae during the horse’s over-ground locomotion.
• The data were used to automatically classify lameness with an 89% sensitivity and 75% specificity in the first scenario (lame vs. non-lame).
• In the second instance (right-lame, left-lame, non-lame), the horses could be accurately allocated into the correct groups in a simple 3-class reclassification test.

Conclusions

• This study demonstrated that a mobile system that reliably detects and quantifies hindlimb lameness in horses during unrestricted movement could be an essential tool in a clinical setting.
• Such a system enables evidence-based examination of lameness before initiating treatments such as nerve blocks or surgical intervention, leading to an enhanced understanding of the problem and more effective intervention strategies.