Elements Influencing sEMG-Based Gesture Decoding: Muscle Fatigue, Forearm Angle and Acquisition Time.

Area of Science:

• Biomedical Engineering
• Neuroscience
• Rehabilitation Technology

Background:

• Surface Electromyography (sEMG) signals are crucial for decoding human movement intentions.
• sEMG-based control is widely used in robotics, prosthetics, and rehabilitation.
• Factors affecting sEMG decoding accuracy require thorough investigation for improved system performance.

Purpose of the Study:

• To investigate the impact of muscle fatigue, forearm angle, and acquisition time on sEMG-based gesture decoding accuracy.
• To evaluate the performance of Linear Discriminant Analysis (LDA) and Probabilistic Neural Network (PNN) models under varying conditions.
• To identify the most influential factors affecting the reliability of sEMG gesture decoding.

Main Methods:

• Selected 11 static gestures and sampled sEMG signals from four forearm muscles: SFD, FCU, ECRL, and FE.
• Extracted signal eigenvalues including Root Mean Square (RMS), Waveform Length (WL), Zero Crossing (ZC), and Slope Sign Change (SSC).
• Employed Linear Discriminant Analysis (LDA) and Probabilistic Neural Network (PNN) for gesture classification.

Main Results:

• Muscle fatigue, forearm angle, and acquisition time reduced decoding accuracy by an average of 7%, 10%, and 13%, respectively.
• Acquisition time demonstrated the most significant impact, causing a maximum accuracy reduction of nearly 20%.
• Both LDA and PNN models showed decreased performance with increased muscle fatigue, altered forearm angles, and longer acquisition times.

Conclusions:

• Muscle fatigue, forearm angle, and acquisition time are critical factors influencing sEMG gesture decoding accuracy.
• Acquisition time is the most detrimental factor, necessitating careful consideration in real-world applications.
• Optimizing data acquisition protocols and developing robust algorithms are essential for enhancing the reliability of sEMG-controlled devices.