Development of a deep neural network for automated electromyographic pattern classification

Abstract

Determining the signal quality of surface electromyography (sEMG) recordings is time consuming and requires the judgement of trained observers. An automated procedure to evaluate sEMG quality would streamline data processing and reduce time demands. This paper compares the performance of two supervised and three unsupervised artificial neural networks (ANNs) in the evaluation of sEMG quality. Manually classified sEMG recordings from various lower-limb muscles during motor tasks were used to train (n=28,000), test performance (n=12,000) and evaluate accuracy (n=47,000) of the five ANNs in classifying signals into four categories. Unsupervised ANNs demonstrated a 30-40% increase in classification accuracy (>98%) compared with supervised ANNs. AlexNet demonstrated the highest accuracy (99.55%) with negligible false classifications. The results indicate that sEMG quality evaluation can be automated via an ANN without compromising human-like classification accuracy. This classifier will be publicly available and will be a valuable tool for researchers and clinicians using electromyography.

Deep hip muscle activation during squatting in femoroacetabular impingement syndrome

“…EMG signals for all squatting trials were visually inspected for artefacts. Trials without acceptable signal-to-noise ratios were excluded (28). EMG data were rectified and smoothed using a 6 Hz dual-pass, zero-lag fourth order, low-pass Butterworth filter.…”

Machine learning methods to support personalized neuromusculoskeletal modelling

“…Machine learning methods can facilitate neuromusculoskeletal modelling across five key domains (Table 1). These domains are: i) feature extraction (Diamond et al 2017;Zhang et al 2014), (ii) synthesizing data (Bahl et al 2019;Clouthier et al 2019;Davico et al 2019b;Nolte et al 2016a;Suwarganda et al 2019;Zhang and Besier 2017), (iii) model generation (Bahl et al 2019;Clouthier et al 2019;Johnson et al 2019a;Johnson et al 2018;Nolte et al 2016b;Zhang and Besier 2017), (iv) execution (Eskinazi and Fregly 2015;Eskinazi and Fregly 2018;Ziaeipoor et al 2019b), and v) data digitization, processing (Ambellan et al 2019;Heimann and Meinzer 2009;Liu et al 2018) and classification (Akhundov et al 2019). Importantly, machine learning can be applied to measured data (e.g., medical imaging, EMG, ground reaction forces) as well as results of created (e.g., rigid multi-body joint model, tendon mesh) and/or executed (e.g., muscle tendon lengths and moment arms, FEA stresses and strains) models.…”

Ankle Joint Torque Prediction Based on Surface Electromyographic and Angular Velocity Signals

“…RMSE is a measure of the difference between the measured and predicted values. The calculation is as follows: (6) In the above formulas, Tm(i) is the measured and Te(i) is the predicted joint torque value for sample i, n corresponds to the total number of samples tested, Tmmax and Tmmin represent the maximum and minimum of Tm.…”

“…The Surface electromyography (sEMG) signal is the sum of the action potentials generated by the active motor units and detected over the skin, the signal contains a wealth of information about muscle functions, and it is one of the basic processing techniques to detect muscle activity and provide the motion intentions of the user. Because of the significant advantages such as noninvasive, real-time, and multi-point measurement, sEMG has been widely used in medical [4,5], engineering studies [6,7], control of prosthesis [8], biomechanics and movement analysis [9][10][11], genomics and exoskeleton robot control. In recent years, the recording and analysis of sEMG signals provide important information to the field of limb joint force prediction [12,13], used as an feedback prediction controller of FES [3] and provide continuous real-time control signal for human musculoskeletal system.…”