A Novel Algorithm for EMG Signal Processing and Muscle Timing Measurement

Abstract

This paper presents a new method for the automated processing of surface electromyography (SEMG) signals, particularly suited for the detection of muscle activation timing. The method has an intermediate level of complexity between simpler (but less performing) and more complex (but in general slower) methods, and is successfully used in the development of biomedical devices for rehabilitation carried out by our group.\ud The method proposed here is based on a statistical approach for threshold computation that is implemented without the need of maximum voluntary contraction or relaxed state, usually required to overcome the difficulty in obtaining the threshold value. The method is compared to 10 popular automated standard methods using different types of simulated signals that approximate the behavior of real SEMG signals. Both the number of activations detected and the onset time measured are analyzed. The algorithm is then applied to real SEMG signals, acquired from healthy subjects. The results are finally compared with literature values.\ud The results show that the proposed algorithm is the best performing method when both the number of activations and the activation timing are considered. In real applications, the algorithm gives results compatible with well-agreed literature data

An Embedded, Eight Channel, Noise Canceling, Wireless, Wearable sEMG Data Acquisition System With Adaptive Muscle Contraction Detection

“…Therefore, the most popular choice is to use RMS or some variant as the primary metric for determining whether there is a contraction. The general approach is to look at a limited bandwidth of the spectrum to determine the existence of contractions (generally with a lowpass or a band-pass filter) [57]. However, since PLN frequency overlaps with the bandwidth sEMG is most dominant, it can significantly degrade the performance of these metrics.…”

“…Another technique is to set the threshold to the RMS of the last contraction and to compare this with the RMS of the incoming frames while incrementally decreasing it with a decay parameter. In [57], the authors calculate the 5% and 95% percentile RMS values and use a convex combination 1 Spectrum Estimation of them as threshold. In [59], the authors propose a peak detection algorithm, where the number of sEMG measurements above a selected threshold are determined to make a decision.…”

Inter-Subject Domain Adaptation for CNN-Based Wrist Kinematics Estimation Using sEMG

“…The surface electromyography (sEMG) reflects the electrical activity of muscle fibres during contraction, and it has been widely used for intelligent prostheses or exoskeleton robotics control [1,2]. To decode human intentions from sEMG more intuitively, artificial intelligence (AI) can be leveraged in either the classification-based hand gesture recognition [3,4] or regression-based kinematic estimation [5,6].…”

On the OCRA Measurement: Automatic Computation of the Dynamic Technical Action Frequency Factor

“…Although it was demonstrated that threshold-based algorithms are less robust in activity recognition rather than machine-learning algorithms [34], it should be underlined that machine-learning algorithms cannot prescind from the training phase and they should be performed for each worker and each industrial workspace, making it difficult to apply these procedures in a real industrial scenario. For this reason, we decided to implement a threshold-based algorithm, also taking into account that this methodology is often used in human motion identification [35][36][37].…”