A novel approach for SEMG signal classification with adaptive local binary patterns

Abstract

Feature extraction plays a major role in the pattern recognition process, and this paper presents a novel feature extraction approach, adaptive local binary pattern (aLBP). aLBP is built on the local binary pattern (LBP), which is an image processing method, and one-dimensional local binary pattern (1D-LBP). In LBP, each pixel is compared with its neighbors. Similarly, in 1D-LBP, each data in the raw is judged against its neighbors. 1D-LBP extracts feature based on local changes in the signal. Therefore, it has high a potential to be employed in medical purposes. Since, each action or abnormality, which is recorded in SEMG signals, has its own pattern, and via the 1D-LBP these (hidden) patterns may be detected. But, the positions of the neighbors in 1D-LBP are constant depending on the position of the data in the raw. Also, both LBP and 1D-LBP are very sensitive to noise. Therefore, its capacity in detecting hidden patterns is limited. To overcome these drawbacks, aLBP was proposed. In aLBP, the positions of the neighbors and their values can be assigned adaptively via the down-sampling and the smoothing coefficients. Therefore, the potential to detect (hidden) patterns, which may express an illness or an action, is really increased. To validate the proposed feature extraction approach, two different datasets were employed. Achieved accuracies by the proposed approach were higher than obtained results by employed popular feature extraction approaches and the reported results in the literature. Obtained accuracy results were brought out that the proposed method can be employed to investigate SEMG signals. In summary, this work attempts to develop an adaptive feature extraction scheme that can be utilized for extracting features from local changes in different categories of time-varying signals.

Current Trends and Confounding Factors in Myoelectric Control: Limb Position and Contraction Intensity

“…Among common contaminants, power-line interference is often removed using notch filters at key harmonic frequencies; motion artefacts are removed using a high-pass Butterworth filter with a typical cutoff frequency of 10 or 20 Hz [26]; ECG interference is sometimes removed using high-pass filters with cutoff of 100 Hz [27,28]; and electrical noise may be minimized by using high-performance components and/or a low-pass filter with cutoff-frequency of 450 or 500 Hz (the active energy of the EMG signal during contraction is considered negligible above 500 Hz). Other advanced signal processing techniques employed for EMG denoising include wavelet transform [29][30][31], wavelet packet transform [32,33], empirical mode decomposition [34][35][36], one-dimensional (1D) local binary pattern [37][38][39], and adaptive Wiener filtering [40].…”

MyoNet: A Transfer-Learning-Based LRCN for Lower Limb Movement Recognition and Knee Joint Angle Prediction for Remote Monitoring of Rehabilitation Progress From sEMG

“…For comparison with the state-of-the-art methodologies, a comparative analysis of the proposed MyoNet model for movement classification with existing study by [4] and [14][15][16] is detailed in Table 9 that has used the same dataset included in our study. However, comparison for knee joint angle prediction with the existing study cannot be done because they have used their own data-set for evaluation.…”

“…(i) decoding user intention and (ii) prediction of joint angle information. The first task is 'decoding user intention', which is usually accomplished by applying pattern recognition techniques on surface electromyography signals (sEMG) captured from user's muscles for inferring the intent of performing desired movements [14]- [16]. However, since the lower limb muscles are present deep beneath the skin with significant overlap among them, therefore the classification of sEMG signals from such muscles is more challenging when compared to the upper limb muscles [14].…”

Implementation and experimental validation of surface electromyogram and force model of Tibialis Anterior muscle for examining muscular factors

“…[5][6][7] The relationship of muscle properties and sEMG features has only been heuristically investigated. 1,8,9 Research studies [10][11][12][13][14] have reported various computational sEMG models to understand the changes in sEMG due to neuromuscular parameters. While the earlier models provide conceptual and generic explanation of the signal, some of the approximations that have limited their suitability for investigating age-or diseaseassociated changes are as follows:…”