Digital filter design for peak detection of surface EMG

Area of Science:

• Biomedical Engineering
• Signal Processing
• Electromyography

Background:

• Motor Unit Action Potential (MUAP) peak detection is crucial for surface EMG decomposition.
• Traditional Low-Pass Differential (LPD) filters are effective for needle EMG but less so for complex surface EMG signals.
• Surface EMG signals are inherently more mixed, posing challenges for accurate MUAP identification.

Purpose of the Study:

• To evaluate the suitability of simple LPD filters for surface MUAP peak detection.
• To propose and analyze Weighted Low-Pass Differential (WLPD) filters as an alternative for surface EMG.
• To assess the performance of WLPD filters under various conditions and parameters.

Main Methods:

• Investigated the performance of LPD filters for surface MUAP detection.
• Developed and analyzed Weighted Low-Pass Differential (WLPD) filters.
• Evaluated filter performance using simulated and recorded surface EMG data.
• Assessed the impact of different window selections and varying MUAP characteristics.

Main Results:

• Simple LPD filters were found to be inadequate for surface MUAP detection.
• WLPD filters demonstrated superior performance compared to simple LPD filters.
• The sinusoidal WLPD filter exhibited enhanced signal-to-noise ratio (SNR) improvement.
• WLPD filters proved more robust against variations in MUAPs and recording conditions.

Conclusions:

• Weighted Low-Pass Differential (WLPD) filters are a more suitable approach for surface EMG decomposition than simple LPD filters.
• WLPD filters enhance the accuracy of Motor Unit Action Potential peak detection in surface EMG.
• The proposed WLPD filters offer improved SNR and robustness, advancing EMG decomposition techniques.