Shared human–robot proportional control of a dexterous myoelectric prosthesis

Abstract

Myoelectric prostheses allow users to recover lost functionality by controlling a robotic device with their remaining muscle activity. Such commercial devices can give users a high level of autonomy, but still do not approach the dexterity of the intact human hand. We present here a method to control a robotic hand, shared between user intention and robotic automation. The algorithm allows user-controlled movements when high dexterity is desired, but also assisted grasping when robustness is paramount. This combination of features is currently lacking in commercial prostheses and can greatly improve prosthesis usability. First, we design and test a myoelectric proportional controller that can predict multiple joint angles simultaneously and with high accuracy. We then implement online control with both able-bodied and amputee subjects. Finally, we present a shared control scheme in which robotic automation aids in object grasping by maximizing contact area between hand and object, greatly increasing grasp success and object hold times in both a virtual and a physical environment. Our results present a viable method of prosthesis control implemented in real time, for reliable articulation of multiple simultaneous degrees of freedom. In the United States alone, about 1.6 million people live with an amputation, 541,000 of which affect the upper limbs 1. This condition diminishes quality of life, mobility and independence, while also imparting a social stigma 2. Upper limb prostheses controlled using surface electromyographic (sEMG) signals attempt to restore hand and arm functionality by using the amputee's remaining muscle activity to control movements of a prosthetic device. However, the capabilities of current commercial prostheses are still grossly inferior compared to the dexterity of the human hand. Commercial devices usually use a two-recording-channel system to control a single degree of freedom (DoF), i.e. one sEMG channel for flexion and one for extension 3. While intuitive, the system provides little dexterity. Patients abandon myoelectric prostheses at high rates, in part because they feel that the level of control is insufficient to merit the price and complexity of these devices 4-6 In recent years, various research groups have made significant advances in myoelectric prosthesis control in laboratory and prototype environments. Many groups have demonstrated great success in grasp classification, which is a common approach for prosthesis control, but limits the user to a library of trained hand postures 7-10. However a few groups have now attempted to decode single finger movements 11-13. Despite high decoding accuracy, these studies showed results mainly from able-bodied subjects performing offline tests. With cited decoding performances of upwards of 90-95% for each method, we see a clear dichotomy between laboratory experiments and clinical viability, a point that is addressed by Jiang et al 14. The idea of "shared control", that is, automation of some portion of the motor command, is already a topic...

Soft Embodiment for Engineering Artificial Limbs

“…the redundancies that exist in both the central and peripheral nervous systems to guide motor control of a limb. If we could recruit a subset of these resources to train the user to control an external device using simplified operating principles (relative to the biological body), we could better benefit from already existing technological advancements [8].…”

“…Soft embodiment will also allow us to exploit sensorimotor resources without necessarily impacting bodily awareness and our sense of body ownership. This provides wider opportunities for implementation, such as with regard to currently emerging technologies for autonomous prosthesis control [8].…”

Discrete action control for prosthetic digits

“…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]. This indicates that, to date, reconstruction of position/velocity trajectories from surface EMG signals in real-time remains a challenging problem.…”

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

“…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 [11][12][13][14][15] , finger positions [16][17][18][19][20][21][22] and velocities 23,24 , as well as fingertip forces [25][26][27][28] . Only a few studies have, however, thus far demonstrated the feasibility of real-time prosthetic finger control in amputee users 17,21,22 .…”