Dexterous hand gestures recognition based on low-density sEMG signals for upper-limb forearm amputees

Abstract

Introduction: Intuitive prosthesis control is one of the most important challenges in order to reduce the user effort in learning how to use an artificial hand. This work presents the development of a novel method for pattern recognition of sEMG signals able to discriminate, in a very accurate way, dexterous hand and fingers movements using a reduced number of electrodes, which implies more confidence and usability for amputees. Methods: The system was evaluated for ten forearm amputees and the results were compared with the performance of able-bodied subjects. Multiple sEMG features based on fractal analysis (detrended fluctuation analysis and Higuchi's fractal dimension) combined with traditional magnitude-based features were analyzed. Genetic algorithms and sequential forward selection were used to select the best set of features. Support vector machine (SVM), K-nearest neighbors (KNN) and linear discriminant analysis (LDA) were analyzed to classify individual finger flexion, hand gestures and different grasps using four electrodes, performing contractions in a natural way to accomplish these tasks. Statistical significance was computed for all the methods using different set of features, for both groups of subjects (able-bodied and amputees). Results: The results showed average accuracy up to 99.2% for able-bodied subjects and 98.94% for amputees using SVM, followed very closely by KNN. However, KNN also produces a good performance, as it has a lower computational complexity, which implies an advantage for real-time applications. Conclusion:The results show that the method proposed is promising for accurately controlling dexterous prosthetic hands, providing more functionality and better acceptance for amputees.

Control of a robotic knee exoskeleton for assistance and rehabilitation based on motion intention from sEMG

“…In this stage, are used the following features: slope sign changes (SSC) and zero crossing (ZC) at time domain; and total power (TTP) and autoregressive (AR) coefficients of order 4 at frequency domain. These are common features used in literature for applications aimed controlling robotic assistive devices with sEMG signals (Lee et al, 2015;Mayor et al, 2017). Each feature is normalized individually based on average and standard deviation values.…”

Empirical Myoelectric Feature Extraction and Pattern Recognition in Hemiplegic Distal Movement Decoding

“…Over the past several decades, a multitude of novel features for myoelectrical classification were verified by SL algorithms [24,[28][29][30][31][32][33][34]. However, for simple decoding tasks (e.g., forearm locomotion) in stroke-related applications or more sophisticated tasks of gesture recognition (usually a few simple hand gestures), the original time domain Hudgins' set (TD-4) is mainly used [14,17,18,[35][36][37].…”

“…Collectively, these studies (mainly conducted on chronic stroke) are connected by actual performance directly tied to the type and quantity of movement intentions [35][36][37]. To conclude, present paretic muscle feature assortments for decoding strategies resonate with research performed on healthy individuals or amputees and cannot provide complete domain expertise [33,[39][40][41][42]. With this in mind, the most rational way to secure precise stroke-oriented feature selection is the validation of all possible combinations between usable feature vectors.…”

Design of a Hand Orthosis for People with Deficiency of the Medial, Radial, and Ulnar Nerves

“…The active orthosis can be controlled by myoelectric signals (MES, also termed electromyography, sEMG) through a hierarchical structure of high-and low-level controls [3]. In the high-level control, surface MES are recorded as non-invasive information of the muscle contraction related to the user's intention.…”