Muscle Synergy-based Grasp Classification for Robotic Hand Prosthetics

Abstract

The main goal of this study is analyzing whether muscle synergies based on surface electromyography (EMG) measurements could be used for hand posture classification in the context of robotic prosthetic control. Target grasps were selected according to usefulness in daily activities. Additionally, due to the feasibility constraints of robotic prosthetics, only 14 gestures (13 feasible grasps and 1 resting state) were analyzed. EMG signals of intact-limb subjects were decomposed into base and activation components for muscle activity evaluation. The results demonstrate that features based on muscle synergies derived from non-negative matrix factorization (NMF) outperform the ones derived from principal component analysis (PCA). Moreover, we also examine the robustness of these methods in the absence of electrodes (muscle importance) and show that NMF is able to provide sufficiently accurate results.

Hierarchical Graphical Models for Context-Aware Hybrid Brain-Machine Interfaces

“…Without baseline correction, time-series windowing into 250 ms intervals was performed over the four second time interval starting after the first second until the end of the trials. Raw time-series features were obtained by calculating the root-mean-squared (RMS) values in each one of the six EMG recording sources in every time window [11].…”

“…In the context of neurophysiologically-driven hand prosthetics, hybrid brainmachine interfaces (hBMIs) based on fusion of electroencephalographic (EEG) and electromyographic (EMG) activities to decode upper limb movements gained significant interest [1], [2]. In that regard, to investigate neural correlates of human motor behavior, a variety of recent studies have shown promising results in both EEG-based [3][4][5][6][7][8] and EMGbased [9][10][11] settings for decoding of complex same hand gestures. In the light of recent promising work, we argue that probabilistic fusion of multimodal information sources in a unified framework would yield significant insights to develop robust hybrid BMIs for neural prosthetics.…”

Multimodal Fusion of EMG and Vision for Human Grasp Intent Inference in Prosthetic Hand Control

“…Despite the aforementioned challenges, there has been efforts in the field to meet these expectations, especially methods based on inferring the human's intent using the amputee's body signals e.g., Electroencephalography (EEG) and Electromyography (EMG) signals [4][5][6]. Despite the advances of robotic prosthetic hands that are based on bodily physiological signals, they generally lack robustness which reduces their effectiveness in amputees' daily life activities.…”

HANDS: a multimodal dataset for modeling toward human grasp intent inference in prosthetic hands

“…To enhance the quality of life of individuals with upper limb loss, further research to improve robotic prostheses is necessary. Along this direction, some electromyography (EMG)-based human intent inference solutions demonstrated promising results for patients with hand and wrist amputations [2,9,10]. The quality of EMG signals may dramatically vary across individuals depending on the specifics of their amputation.…”

“…The quality of EMG signals may dramatically vary across individuals depending on the specifics of their amputation. Some studies also investigated the possibility of complementing EMG with electroencephalography (EEG) signals to improve inference accuracy [2,9,15,19]; however, in many practical situations the added information from EEG does not have significant impact on performance. Both EMG and EEG models may need frequent calibration to account for signal nonstationarity due to various factors, such as electrode locations or skin conductance.…”