Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Non-Intrusive Neural Quality Assessment Model for Surface Electromyography Signals

Cho-Yuan Lee, Kuan-Chen Wang, Kai-Chun Liu, Yu-Te Wang, Xugang Lu, Ping-Cheng Yeh, Yu Tsao

TL;DR

ECG contamination in surface electromyography (sEMG) from muscles near the heart impairs downstream applications and lacks objective, reference-free quality metrics. The authors introduce QASE-net, a non-intrusive SNR predictor based on a CNN–BLSTM architecture with an attention mechanism, trained end-to-end on raw sEMG to estimate signal quality without clean references. Validated on real-world paired sEMG/ECG data from the NINAPro-DB2 and MIT-BIH NSRD databases, QASE-net achieves very high linear correlation (LCC > 0.99) and low prediction error (MSE < 0.40), outperforming a handcrafted-feature baseline and other NN baselines. This approach offers a robust, data-driven means to quantify sEMG quality and can guide ECG removal and denoising in practical, non-intrusive scenarios, improving reliability of sEMG-based applications.

Abstract

In practical scenarios involving the measurement of surface electromyography (sEMG) in muscles, particularly those areas near the heart, one of the primary sources of contamination is the presence of electrocardiogram (ECG) signals. To assess the quality of real-world sEMG data more effectively, this study proposes QASE-net, a new non-intrusive model that predicts the SNR of sEMG signals. QASE-net combines CNN-BLSTM with attention mechanisms and follows an end-to-end training strategy. Our experimental framework utilizes real-world sEMG and ECG data from two open-access databases, the Non-Invasive Adaptive Prosthetics Database and the MIT-BIH Normal Sinus Rhythm Database, respectively. The experimental results demonstrate the superiority of QASE-net over the previous assessment model, exhibiting significantly reduced prediction errors and notably higher linear correlations with the ground truth. These findings show the potential of QASE-net to substantially enhance the reliability and precision of sEMG quality assessment in practical applications.

A Non-Intrusive Neural Quality Assessment Model for Surface Electromyography Signals

TL;DR

ECG contamination in surface electromyography (sEMG) from muscles near the heart impairs downstream applications and lacks objective, reference-free quality metrics. The authors introduce QASE-net, a non-intrusive SNR predictor based on a CNN–BLSTM architecture with an attention mechanism, trained end-to-end on raw sEMG to estimate signal quality without clean references. Validated on real-world paired sEMG/ECG data from the NINAPro-DB2 and MIT-BIH NSRD databases, QASE-net achieves very high linear correlation (LCC > 0.99) and low prediction error (MSE < 0.40), outperforming a handcrafted-feature baseline and other NN baselines. This approach offers a robust, data-driven means to quantify sEMG quality and can guide ECG removal and denoising in practical, non-intrusive scenarios, improving reliability of sEMG-based applications.

Abstract

In practical scenarios involving the measurement of surface electromyography (sEMG) in muscles, particularly those areas near the heart, one of the primary sources of contamination is the presence of electrocardiogram (ECG) signals. To assess the quality of real-world sEMG data more effectively, this study proposes QASE-net, a new non-intrusive model that predicts the SNR of sEMG signals. QASE-net combines CNN-BLSTM with attention mechanisms and follows an end-to-end training strategy. Our experimental framework utilizes real-world sEMG and ECG data from two open-access databases, the Non-Invasive Adaptive Prosthetics Database and the MIT-BIH Normal Sinus Rhythm Database, respectively. The experimental results demonstrate the superiority of QASE-net over the previous assessment model, exhibiting significantly reduced prediction errors and notably higher linear correlations with the ground truth. These findings show the potential of QASE-net to substantially enhance the reliability and precision of sEMG quality assessment in practical applications.
Paper Structure (12 sections, 5 equations, 4 figures, 1 table)

This paper contains 12 sections, 5 equations, 4 figures, 1 table.

Table of Contents

  1. Introduction
  2. Related Work
  3. SNR estimation of sEMG with ECG interference
  4. NN-based signal quality assessment
  5. Materials and methods
  6. Public database of sEMG and ECG
  7. Data preparation
  8. Network architecture of the proposed method
  9. Evaluation Metrics
  10. Implementation details
  11. Results and discussion
  12. Conclusion

Figures (4)

  • Figure 1: Architecture of the proposed method.
  • Figure 2: The MSE values of SNR predictions of QASE-net and comparative models under different SNR inputs.
  • Figure 3: Scatter plots of the SNR predictions by (a) WL+MLP oo2020signal and (b) the proposed QASE-net.
  • Figure 4: Box plots of SNR predictions by (a) WL+MLP oo2020signal and (b) the proposed QASE-net.