Compressed sensing study for the sEMG data of SCI survivors
Zongxian Feng1, Beining Cui2, Fan He1
1Department of Orthopedics, Ningbo Medical Center Lihuili Hospital, Ningbo, Zhejiang Province, China.
Computer Methods in Biomechanics and Biomedical Engineering
|September 8, 2025
Summary
Surface electromyography (sEMG) can evaluate walking function. Compressed sensing (CS) with regularized orthogonal matching pursuit (ROMP) and a specific matrix improved sEMG signal reconstruction for spinal cord injury (SCI) patients.
Area of Science:
- Biomedical Engineering
- Rehabilitation Technology
- Signal Processing
Background:
- Surface electromyography (sEMG) is crucial for assessing walking function.
- Compressed sensing (CS) reduces data acquisition needs by exploiting signal sparsity.
- Spinal cord injury (SCI) impacts motor control, necessitating effective evaluation methods.
Purpose of the Study:
- To introduce and evaluate a novel compressed sensing (CS) algorithm for surface electromyography (sEMG) signal reconstruction in spinal cord injury (SCI) patients.
- To compare the performance of the proposed regularized orthogonal matching pursuit (ROMP) algorithm against the conventional orthogonal matching pursuit (OMP) algorithm.
- To investigate the influence of different measurement matrices on the accuracy of sEMG signal reconstruction.
Main Methods:
- Development of a compressed sensing (CS) algorithm for sEMG signals using regularized orthogonal matching pursuit (ROMP).
- Reconstruction of multiple sEMG signals from spinal cord injury (SCI) subjects using the ROMP algorithm.
- Comparative analysis of ROMP with orthogonal matching pursuit (OMP) and evaluation of various measurement matrices, including the binary permuted block diagonal (BPBD) matrix.
Main Results:
- The regularized orthogonal matching pursuit (ROMP) algorithm demonstrated effective reconstruction of sEMG signals from SCI patients.
- The combination of ROMP with a binary permuted block diagonal (BPBD) matrix yielded superior reconstruction accuracy compared to the standard OMP algorithm.
- Different measurement matrices significantly impacted the reconstruction performance of the CS algorithms.
Conclusions:
- The proposed ROMP-based CS algorithm offers a promising approach for efficient sEMG data acquisition in walking function evaluation for SCI individuals.
- The BPBD matrix enhances the performance of CS algorithms, suggesting its utility in sEMG-based rehabilitation monitoring.
- This study highlights the potential of advanced signal processing techniques to improve the assessment and management of SCI.
