Proof of Concept of an Online EMG-Based Decoding of Hand Postures and Individual Digit Forces for Prosthetic Hand Control

Abstract

IntroductionOptions currently available to individuals with upper limb loss range from prosthetic hands that can perform many movements, but require more cognitive effort to control, to simpler terminal devices with limited functional abilities. We attempted to address this issue by designing a myoelectric control system to modulate prosthetic hand posture and digit force distribution.MethodsWe recorded surface electromyographic (EMG) signals from five forearm muscles in eight able-bodied subjects while they modulated hand posture and the flexion force distribution of individual fingers. We used a support vector machine (SVM) and a random forest regression (RFR) to map EMG signal features to hand posture and individual digit forces, respectively. After training, subjects performed grasping tasks and hand gestures while a computer program computed and displayed online feedback of all digit forces, in which digits were flexed, and the magnitude of contact forces. We also used a commercially available prosthetic hand, the i-Limb (Touch Bionics), to provide a practical demonstration of the proposed approach’s ability to control hand posture and finger forces.ResultsSubjects could control hand pose and force distribution across the fingers during online testing. Decoding success rates ranged from 60% (index finger pointing) to 83–99% for 2-digit grasp and resting state, respectively. Subjects could also modulate finger force distribution.DiscussionThis work provides a proof of concept for the application of SVM and RFR for online control of hand posture and finger force distribution, respectively. Our approach has potential applications for enabling in-hand manipulation with a prosthetic hand.

Effect of user adaptation on prosthetic finger control with an intuitive myoelectric decoder

“…This opportunity has been explored in the context of controlling both virtual 23 as well as prosthetic 24 hands. Besides finger position and velocity decoding, individual fingertip forces have also been reconstructed offline 25,26 and in real-time [27][28][29] using surface EMG signals.…”

Discrete action control for prosthetic digits

“…In the context of myoelectric control, multi-label (i.e., simultaneous) classification has been previously used only to decode wrist and/or whole hand functions [28][29][30][31][32]. For prosthesis digit control, the focus has been on using multi-output regression to decode joint angles (i.e., positions) [9][10][11][12]14], velocities [15,16] or fingertip forces [18][19][20]. Action control can be seen as an extreme, discretised case of velocity control (see figure 2), whereby the velocity can be either zero or take a constant value, which is only parametrised by its sign.…”

“…The term proportional control is often used to describe the feasibility of controlling one or more prosthesis output(s) in a continuous space [3]. To that end, many research groups, including us, have used multi-output regressionbased algorithms to map features extracted from surface or intramuscular EMG channels onto wrist [4][5][6][7][8] and/or finger position [9][10][11][12][13][14], velocity [15,16], and force trajectories [17][18][19][20]. For prosthetic digit control, several studies have shown premise in decoding position/velocity offline [9, 11-13, 15, 16], however, only a smaller number have achieved real-time digit control in amputees [10,14,21].…”

Myoelectric digit action decoding with multi-label, multi-class classification: an offline analysis

“…In myoelectric control, multi-label classification has been previously used to decode simultaneous wrist and hand motions [4][5][6][7][8][9] . For control of prosthetic digits, however, previous efforts have focused on using multi-output regression to reconstruct position [15][16][17][18]20 , velocity 22,23 or fingertip force trajectories [25][26][27] . One study has previously adopted a similar approach to ours 36 , but with three main differences: firstly, the labels corresponded to digit positions instead of actions; secondly, labels were binary (i.e., digits could be fully open or closed), whereas with our approach actions can take three values (i.e., open, close, or stall); finally, due to using binary outputs, a controller using the approach proposed previously would be limited to extreme digit positions.…”

“…From a technical perspective, the ultimate goal of the myoelectric control field is to approximate this dexterity via simultaneous and independent control of multiple degrees of freedom (DOFs) in a continuous space. To that end, several groups, including us, have used regression-based methods to reconstruct wrist kinematic trajectories [10][11][12][13][14] , finger positions [15][16][17][18][19][20][21] and velocities 22,23 , as well as fingertip forces [24][25][26][27] . Only a few studies have, however, thus far demonstrated the feasibility of real-time prosthetic finger control in amputee users 16,20,21 .…”