Fault-Tolerant Sensor Detection of sEMG signals: Quality Analysis Using a Two-Class Support Vector Machine

Area of Science:

• Biomedical Engineering
• Signal Processing
• Rehabilitation Technology

Background:

• Reliable surface electromyography (sEMG) signal acquisition is crucial for advanced prosthetic control.
• Contamination in sEMG signals can significantly impair the performance and reliability of myoelectric prostheses.
• Automated methods for identifying sEMG signal artifacts are needed for real-world applications.

Purpose of the Study:

• To evaluate a method for automatic identification of common contaminants in sEMG signals.
• To assess the effectiveness of a support vector machine (SVM) approach for sEMG signal validation.
• To enhance the reliability of intelligent recognition systems for sEMG-controlled prostheses.

Main Methods:

• Development and evaluation of a two-class support vector machine (SVM) classifier.
• Training the SVM with both clean and artificially contaminated sEMG datasets.
• Consideration of common contaminants: electrocardiogram interference, motion artifact, power line interference, amplifier saturation, and electrode displacement.

Main Results:

• The SVM method effectively distinguished between clean and contaminated sEMG signals.
• Detection accuracy was maintained even with an increased number of degraded channels.
• Signal contamination detection (SFTD) performance varied based on noise type, subject status (amputee/non-amputee), and channel analyzed.

Conclusions:

• The proposed SVM-based method offers a viable solution for detecting contaminants in sEMG signals.
• Automated signal validation prior to movement recognition enhances the reliability of intelligent prosthetic control.
• This approach contributes to the development of more robust and dependable sEMG-based assistive technologies.