Effect of Muscle Fatigue on Surface Electromyography-Based Hand Grasp Force Estimation

Area of Science:

• Biomedical Engineering
• Rehabilitation Technology
• Signal Processing

Background:

• Surface electromyography (sEMG) offers promising hand grasp force estimation in labs.
• Clinical applicability is limited by physiological changes, including muscle fatigue.
• Current qualitative muscle fatigue assessments yield minimal improvements in force estimation models.

Purpose of the Study:

• To propose an easy-to-implement quantitative method for evaluating muscle fatigue.
• To demonstrate the impact of quantitative muscle fatigue metrics on hand grasp force estimation.
• To improve the accuracy and reliability of sEMG-based force estimation in the presence of muscle fatigue.

Main Methods:

• Developed a quantitative muscle fatigue metric based on reduced maximal force capacity.
• Utilized a back-propagation neural network (BPNN) for sEMG-hand grasp force estimation.
• Compared three experimental cases: time-domain features only, frequency-domain features, and incorporating the quantitative muscle fatigue metric.

Main Results:

• The proposed quantitative metric effectively distinguishes muscle fatigue levels and task intensities.
• Incorporating the muscle fatigue metric as an additional input significantly enhanced the BPNN model's performance.
• Achieved a 6.3797% increase in the coefficient of determination (R²) for hand grasp force estimation.

Conclusions:

• Quantitative muscle fatigue evaluation is crucial for accurate sEMG-based hand grasp force estimation.
• The proposed method offers a substantial improvement over conventional approaches.
• This approach enhances the clinical applicability of sEMG for hand grasp force estimation.