Analysis and Comparison of Features and Algorithms to Classify Shoulder Movements From sEMG Signals

Abstract

Shoulder movements are not considered for electromyography-based pattern classification control, due to the difficulty to manufacture three-degrees-of-freedom shoulder prostheses. This paper aims at exploring the feasibility of classifying up to nine shoulder movements by processing surface electromyography signals from eight trunk muscles. Experimenting with different pattern recognition methods, two classifiers were developed, considering six different combinations of window sizes and increments, and three feature sets for each channel. Applying linear discriminant analysis the best performance was obtained on a window length of 500 ms associated to temporal increments of 62 ms. This setting yielded a 100% accuracy for recognizing four movements, and progressively degraded to 92% for nine movements. Using neural networks, higher accuracy was obtained in particular in the 9-class problem. Finally, the signals from the eight channels were analyzed in order to check the possibility to reduce the number of acquisition channels.

Applying LDA-based pattern recognition to predict isometric shoulder and elbow torque generation in individuals with chronic stroke with moderate to severe motor impairment

“…This likely impacted the accuracy of the classifier and could potentially be improved through EMG acquisition from rotator cuff muscles and other involved muscle groups. For example, high classification accuracies (> 92%) have been achieved for shoulder movements within a healthy population using eight channels of EMG over muscles of the back and torso and slightly longer window lengths [20, 21]. Despite the aforementioned limitation, the EMG data does a fair job of classifying and, as seen in the combined data set, offers an improvement especially to the most confused classes, increasing the group average for each class closer to the control scheme goal of > 90%.…”

“…Historically, these devices were controlled using simple amplitude-based thresholds, but recently the use of linear discriminant analysis (LDA) based pattern recognition has proven to be both accurate and computationally efficient and enables intuitive control of a greater number of degrees of freedom [17–19]. Although LDA-based pattern recognition is often focused on controlling distal joints, pattern recognition of shoulder motions of healthy controls has been explored for the purposes of application to the population with amputation and has achieved classification accuracies above 90% [20, 21]. 90% is significant, as it has been implicated as a transitional value between high functionality and extremely variable levels of functionality of a myoelectric prosthesis based on the user, classifier, and their interaction [22].…”

Classification of Hand Movements based on Discrete Wavelet Transform and Enhanced Feature Extraction

“…LDA is the most robustness learning algorithm in EMG studies. Additionally, LDA is high speed training and computationally less expensive [29]. In LDA, the data is assumed to be normally distributed with equal covariance matrices.…”

Shoulder girdle recognition using electrophysiological and low frequency anatomical contraction signals for prosthesis control

“…In the literature, key works by Gonzalez et al [7], Rivela et al [8], Soma et al [9], and Horiuchi et al [10] showed that shoulder motions could be identified for a group of nonamputees, whereas Kaur et al [11] managed to differentiate shoulder motions for a group of amputees across four shoulder motions, and using the wavelet transform, obtained an accuracy of 98%. More recently, Sharba et al [4], were able to differentiate between various shoulder motions for a group of non-amputees and amputees, where they showed that high classification accuracies (CAs) were obtainable with a slight degradation in the accuracy for the case of the amputee participants.…”