A Fatigue Involved Modification Framework for Force Estimation in Fatiguing Contraction.

Area of Science:

• Biomechanics
• Biomedical Engineering
• Rehabilitation Science

Background:

• Muscle fatigue negatively impacts the accuracy of force estimation models.
• Existing models often fail to account for fatigue-induced changes in muscle behavior.
• Accurate force estimation is crucial for biomechanical analysis, rehabilitation, and prosthetic control.

Purpose of the Study:

• To develop a fatigue-modified force estimation framework to mitigate the effects of muscle fatigue.
• To establish a relationship between fatigue indices and force estimation model parameters.
• To enhance the performance and robustness of muscle force estimation models.

Main Methods:

• Collected muscle force and surface electromyography (sEMG) data using high-density electrodes and a dynamometer.
• Applied multi-step signal pre-processing and non-negative matrix factorization (NMF) for signal optimization and fatigue index extraction.
• Trained a degree 4 polynomial fitting model, correlating its parameters with extracted fatigue indices for real-time modification.

Main Results:

• The proposed fatigue-modified force estimation model demonstrated significant improvements across various testing scenarios, including different sexes and ages.
• Achieved notable reductions in percentage root mean square error (RMSE): 36.5% for young males, 20.7% for young females, and 20.4% for elderly males.
• The model's performance enhancement was consistent, indicating its effectiveness in real-world applications.

Conclusions:

• The developed framework effectively alleviates the negative impacts of muscle fatigue on force estimation accuracy.
• This approach offers a valuable tool for advancing biomechanical studies, optimizing rehabilitation treatments, and improving prosthetic device functionality.
• The integration of fatigue indices provides a robust method for dynamic force prediction in diverse populations.