An approach to surface EMG decomposition based on higher-order cumulants

Area of Science:

• Biomedical Engineering
• Signal Processing
• Neuroscience

Background:

• Surface electromyograms (SEMGs) are complex signals reflecting motor unit activity.
• Accurate decomposition of SEMGs into individual motor-unit action potentials (MUAPs) is crucial for clinical and research applications.
• Existing decomposition methods face challenges with noise and signal complexity.

Purpose of the Study:

• To develop and validate a novel approach for SEMG decomposition using higher-order cumulants.
• To accurately estimate motor-unit action potentials (MUAPs) from noisy SEMG signals.
• To assess the performance of the proposed method on synthetic SEMG data.

Main Methods:

• A two-step decomposition procedure was implemented.
• The first step involved a multivariate w-slice method for initial MUAP extraction.
• The second step refined MUAP estimates using modified Newton-Gauss iteration for optimal cumulant fitting.

Main Results:

• The proposed method demonstrated effective SEMG decomposition on synthetic data.
• Low first-norm differences (5.4% in noise-free, 6.0% at 10dB SNR, 6.5% at 0dB SNR) were achieved between original and decomposed MUAPs.
• The method proved robust to varying signal-to-noise ratios (SNR).

Conclusions:

• The higher-order cumulant-based approach provides a powerful tool for SEMG decomposition.
• The method accurately estimates MUAPs, offering potential for improved analysis of neuromuscular activity.
• This technique shows promise for advancing quantitative electromyography in research and clinical settings.