A hybrid method for real-time stimulation artefact removal during functional electrical stimulation with time-variant...

Area of Science:

• Biomedical Engineering
• Neuroscience
• Rehabilitation Technology

Background:

• Functional electrical stimulation (FES) is crucial for restoring motor function but is hindered by stimulation artefacts (SAs).
• Extracting clean surface electromyography (sEMG) signals during FES is challenging due to signal interference.
• Accurate sEMG analysis is vital for effective FES control and rehabilitation.

Purpose of the Study:

• To develop and validate a hybrid hardware-software system for real-time SA removal and sEMG extraction during FES.
• To improve the accuracy and reliability of sEMG signals in the presence of time-variant FES parameters.
• To enhance the utility of FES in clinical applications and research.

Main Methods:

• A novel sEMG detection front-end (DFE) with fast recovery and blanking was designed to prevent saturation.
• A database-based Gram-Schmidt (DBGS) algorithm was employed for SA removal, utilizing high-similarity templates.
• The system was tested on able-bodied volunteers and individuals with stroke to assess performance.

Main Results:

• The 6th-order DBGS algorithm achieved significant SA attenuation (12.77 dB) and high sEMG correlation (0.84).
• Performance metrics surpassed existing methods like empirical mode decomposition and notch filters.
• Strong sEMG-torque correlations were observed in both able-bodied (0.78) and stroke (0.48) participants.

Conclusions:

• The proposed hybrid method effectively removes SAs and extracts volitional sEMG in real time during dynamic FES.
• This technology holds promise for advancing FES-based neurorehabilitation and control systems.
• The DBGS algorithm offers a robust solution for artifact-contaminated biosignal processing.