Electromyography signal segmentation method based on spectral subtraction backtracking

Abstract

Surface electromyography (EMG) is a bioelectrical signal that recognizes speech contents in a non-acoustic form. Activity detection is an important research direction in EMG research. However, in the low signal-to-noise ratio (SNR) environment, it is difficult for traditional methods to obtain accurate active signals. This paper proposes a new energy-based spectral subtraction backtracking (E-SSB) method to segment EMG active signal in the low SNR environment. Compared with traditional energy detection, the algorithm in this paper adds spectral subtraction (SS) to filter out the clutter, and raises a retrospective idea to improve the classification performance. The experiment results show the proposed activity detection method is more effective than other methods in the low SNR environment.

References

[1] LiuL, YueW H. Principles of neuro electromyography[M], 1983, Beijing, Science Press: 1-20(in Chinese)

[2] Gonzalez-LopezJ A, Gomez-AlanisA, DoñasJ M M, et al.. Silent speech interfaces for speech restoration: a review[J]. IEEE access, 2020, 8: 177995-178021

[3] MeltznerG S, HeatonJ T, DengY, et al.. Development of sEMG sensors and algorithms for silent speech recognition[J]. Journal of neural engineering, 2018, 15(4):046031

[4] ZhuM, ZhangH, WangX, et al.. Towards optimizing electrode configurations for silent speech recognition based on high-density surface electromyography[J]. Journal of neural engineering, 2021, 18(1): 016005

[5] HippenstielR D. Detection theory: applications and digital signal processing[M], 2017, Montrouge, CRC Press

[6] HwangI, ChangJ H. End-to-end speech endpoint detection utilizing acoustic and language modeling knowledge for online low-latency speech recognition[J]. IEEE access, 2020, 8: 161109-161123

[7] MeltznerG S, HeatonJ T, DengY, et al.. Silent speech recognition as an alternative communication device for persons with laryngectomy[J]. IEEE/ACM transactions on audio speech & language processing, 2017, 25(12):2386-2398

[8] ChengJ, ChenX, PengH. An onset detection method for action surface electromyography based on sample entropy[J]. Acta electonica sinica, 2016, 44(2):479

[9] BengacemiH, Abed-MeraimK, ButtelliO, et al.. A new detection method for EMG activity monitoring[J]. Medical & biological engineering & computing, 2020, 58(2): 319-334

[10] KangK, RheeK, ShinH C. A precise muscle activity onset/offset detection via EMG signal[C], 2021, New York, IEEE: 633-635

[11] DeA B P A, DeS M A M, NadalJ. Electromyographic activity of the lower limb in runners with anterior knee pain while running[J]. Research on biomedical engineering, 2021, 37(2):135-142

[12] BengacemiH, MesloubA, OuldaliA, et al.. Adaptive linear energy detector based on onset and offset electromyography activity detection[C], 2017, New York, IEEE: 409-413

[13] ZhangT, ZhangX B, ZhuX X. Speech endpoint detection with low SNR based on improved cepstrum distance method[J]. Audio engineering, 2017, 41(7):108-112

[14] SrisuwanN, PrukpattaranontP, LimsakulC. Comparison of classifiers for EMG based speech recognition[J]. Journal of physics: conference series, 2020, 1438(1):012032

[15] BollS. Suppression of acoustic noise in speech using spectral subtraction[J]. IEEE transactions on acoustics, speech, and signal processing, 1979, 27(2): 113-120

[16] NoderaH, OsakiY, YamazakiH, et al.. Classification of needle-EMG resting potentials by machine learning[J]. Muscle & nerve, 2019, 59(2):224-228

[17] WangJ, YangY, MaoJ, et al.. CNN-RNN: a unified framework for multi-label image classification[C], 2016, New York, IEEE: 2285-2294

[18] WangY, ZhangM, WuR M, et al.. Silent speech decoding using spectrogram features based on neuromuscular activities[J]. Brain sciences, 2020, 10(7):442

[19] GADDY D, KLEIN D. Digital voicing ofsilent speech[EB/OL]. (2020-10-06) [2022-04-10]. https://arxiv.org/pdf/2010.02960.pdf.

[20] AtzoriM, MüllerH. The Ninapro database: a resource for sEMG naturally controlled robotic hand prosthetics[C], 2015, New York, IEEE: 7151-7154