The Effect of Filtering on Signal Features of Equine sEMG Collected During Overground Locomotion in Basic Gaits.

This research paper evaluates how varying methods of signal processing impact the reliability and interpretability of data when studying horse locomotion using surface electromyography (sEMG). With a focus on the balance between signal loss and noise reduction, the authors uncover which filtering methods present the most trustworthy results.

Study Aim and Methodology

The primary goal of this research was to shed light on how different filtering methods impact the basic signal features when studying horse locomotion via sEMG.

• This study used raw sEMG signals collected from the quadriceps muscle of six horses performing various basic locomotive activities – walking, trotting, and cantering.
• The signals were then filtered through eight different methods with varying cut-off frequencies. These methods include both high-pass and low-pass filtering, plus bandpass filtering covering a range of frequencies.
• Post signal filtration, various signal features like amplitude, root mean square (RMS), integrated electromyography (iEMG), median frequency (MF), and signal-to-noise ratio (SNR) were calculated along with signal loss metrics and power spectral density (PSD) for each signal variation.

Key Findings

The research discovered significant filtering-induced changes in signal features based on the filtering method used.

• A high-pass filtering set at 40 Hz, and a bandpass filtering set at 40-450 Hz contributed to significant changes in signal features but facilitated complete attenuation of low-frequency noise.
• Interestingly, these two methods did not result in differing degrees of signal loss.
• Other filtering methods led to only partial reduction of low-frequency noise contamination and resulted in lower signal loss. However, they also led to less consistent signal features changes across different gaits.

Implications and Recommendations

The results of the study underline the importance of carefully considering the method of signal filtration when comparing sEMG studies of equine locomotion and drawing reliable inferences.

• It emphasizes the significance of adopting standardized filtering methods, particularly recommending a high-pass cut-off frequency set at 40 Hz.
• This research is beneficial for equine sEMG studies used for training, rehabilitation programs, and diagnosing musculoskeletal diseases in horses.
• It can help in creating a more reliable cross-referencing system among sEMG studies involving different filtering approaches.