Quantitative lameness assessment in the horse based on upper body movement symmetry: The effect of different filtering techniques on the quantification of motion symmetry

Serra Bragança, F M and Roepstorff, C and Rhodin, M and Pfau, T and Van Weeren, P R and Roepstorff, L (2020) Quantitative lameness assessment in the horse based on upper body movement symmetry: The effect of different filtering techniques on the quantification of motion symmetry. Biomolecular NMR Assignments, 57.

Full text not available from this repository.

Abstract

Quantitative gait analysis in horses is rapidly gaining importance, both clinically and in research. The number of available systems is increasing, but the methods of signal analysis differ between systems and research groups. Our objectives are to describe and evaluate the effects of different methods of signal analysis for processing of data from equine kinematic gait analysis. To this end, we use theoretical signals based on previously published work, followed by the evaluation of the performance of each technique using real data from horses with induced lameness. Two infinite impulse response (IIR), high-pass filters (Butterworth and Chebyshev), a signal decomposition method and a moving average filtering technique were evaluated. First, we describe methods to fine-tune each filter to the optimal settings based on residual analysis. Second the performance of each filter is evaluated based on differences in calculated symmetry parameters from horses with induced lameness. We show that optimisation of filtering techniques is crucial when processing signals used for objective lameness quantification. Improper selection of the cut-off frequency for IIR filters can result in false negative results (average values above or below predefined reference values). The IIR Butterworth filter and the signal decomposition method achieved the best reduction of unwanted signal components. Knowledge of the available filtering techniques is a pre-requisite for adequate signal processing of gait data from quantitative analysis systems in horses.