Extraction of Muscle Synergies in Spinal Cord Injured Patients

Area of Science:

• Neuroscience
• Biomedical Engineering
• Rehabilitation Science

Background:

• Muscle synergies represent motor activity through the linear combination of muscle activation patterns.
• Extracting muscle synergies from electromyography (EMG) data is crucial for understanding motor control.
• Standard non-negative matrix factorization (NNMF) methods do not account for physiological signal delays.

Purpose of the Study:

• To develop and validate a novel algorithm for extracting muscle synergies that incorporates physiological signal delays.
• To improve the analysis of muscle synergies in individuals with complete spinal cord injury (SCI) undergoing epidural spinal stimulation.
• To enhance the physiological relevance and descriptive utility of muscle synergy features.

Main Methods:

• A new algorithm, regularized ShiftNMF, was developed to extract muscle synergies while accounting for signal propagation delays.
• The algorithm was applied to EMG data from a complete SCI patient performing simple standing tasks under epidural spinal stimulation.
• Performance was evaluated based on EMG activity reconstruction accuracy and the physiological consistency of extracted synergies.

Main Results:

• Regularized ShiftNMF significantly improved the reconstruction of EMG activity compared to standard NNMF.
• The muscle synergies extracted using the new algorithm demonstrated greater physiological consistency.
• The derived features provided a more effective description of the patient's motor behavior.

Conclusions:

• The regularized ShiftNMF algorithm offers a more accurate and physiologically relevant method for analyzing muscle synergies, particularly in SCI populations with delayed neural signaling.
• This advancement can lead to better insights into motor control deficits and inform rehabilitation strategies.
• Accounting for physiological delays is essential for robust muscle synergy analysis in clinical neurorehabilitation settings.