Minisymposium "Movement Augmentation with Supernumerary Limbs"
Detailed Program
Thursday, May 6, 2021
Chair: Carsten Mehring, University of Freiburg
Movement Augmentation with Natural and Artificial Supernumerary Effectors
Mehring, Carsten, University of Freiburg
Keywords: Motor learning, neural control, and neuromuscular systems, Brain-computer/machine Interface, Human Performance - Sensory-motor
Abstract: In this first presentation for the suggested mini-symposium “Movement augmentation with supernumerary limbs” we (i) introduce and discuss different types of movement augmentation and (ii) present findings from a study with human subjects with six-fingered hands that show augmented manipulation abilities [1] and from a study on concurrent control of an ECoG based brain-computer interface (BCI) and natural movements [2]
ACKNOWLEDGEMENTS We thank M Akselrod, L Bashford, M Mace, H Choi, M Blüher, A-S Buschhoff, T Pistohl, R Salmon, A Cheah, O Blanke, A Serino, E Burdet, J Wu, D Sarma, K Collins, RPN Rao, JG Ojeman, J Eden, D Farina, M Bräcklein for their contributions to the work presented here.
REFERENCES [1] Mehring, C., Akselrod, M., Bashford, L. et al. Augmented manipulation ability in humans with six-fingered hands. Nat Commun 10, 2401 (2019). https://doi.org/10.1038/s41467-019-10306-w [2] Bashford L et al. Concurrent control of a brain-computer interface and natural overt movements, J. Neural Eng., 15(4) 066021
Trimanipulation with Body Interfaces
Huang, Yanpei, Imperial College London
Eden, Jonathan, Imperial College London
Burdet, Etienne, Imperial Collge of Science, Technology and Medicine
Keywords: Neural Interfaces - Sensors and body Interfaces, Human performance, Brain-Computer/Machine Interface - Robotics applications
Abstract: Body interfaces that move with one limb. e.g. a foot interface, can be used to control a robotic arm. However, the use of such interfaces for controlling a supernumerary robotic limb (SL) in coordination with the natural limbs (NL) has not been well studied. In our presentation, we will first review recent studies on trimanual control with body interfaces using one foot and the arms. We will then present a recent study that was conducted with 14 subjects to evaluate human performance of trimanipulation with haptic feedback. The results of trimanipulation with the hands and foot of a single subject exhibited similar or even superior performance than similar operation carried out by two cooperating subjects.
Voluntary Decoupling of Low-Frequency and Beta Band Power in Motoneuron Behavior
Bracklein, Mario, Imperial College London
Ibáñez, Jaime, Institute of Neurology, University College London
Barsakcioglu, Deren Yusuf, Imperial College London
Farina, Dario, Imperial College London
Keywords: Neural Interfaces - Recording, Neural signal processing, Motor learning, neural control, and neuromuscular systems
Abstract: Supernumerary artificial limbs controlled concurrently with natural limbs would allow extending human motor capacities beyond anatomical boundaries. However, to ensure flexible and parallel navigation of supernumerary and natural limbs, control signals independent from those activating natural degrees-of-freedom are needed. We introduce a novel approach of extracting true augmented control signals by using a non-invasive neural interface to facilitate spectral separation of the neural code of spinal motor neurons. We demonstrate that spectral components outside the bandwidth of musculoskeletal control, i.e. inside the beta range (13-30Hz), can be partially uncoupled from low-frequency components and thus from voluntary muscle contractions.
An Intuitive Control for a Wearable Supernumerary Robotic Limb
Jarrassé, Nathanael, Sorbonne Université, ISIR UMR 7222 CNRS
Poignant, Alexis, Sorbonne Université, ISIR UMR 7222 CNRS
Legrand, Mathilde, Sorbonne Université, ISIR UMR 7222 CNRS
Khoramshahi, Mahdi, Sorbonne Université, ISIR UMR 7222 CNRS
Morel, Guillaume, Université Pierre Et Marie Curie - Paris 6
Keywords: Neurorehabilitation - Robotics, Neurorehabilitation - Wearable systems, Human performance
Abstract: Supernumerary robotic limbs (SRL) are wearable robotic devices that promise to augment the wearers' motor capabilities and coordinate with their natural limbs. However, current SRLs are controlled with little robust and hardly flexible automatic behaviours inspired from conventional robotics. We introduce here a concept of intuitive voluntary control of an SRL, exploiting an approach we recently developed for prosthetics limbs relying on the minimization of the body compensation of the user. We present preliminary results obtained on the control of a virtual SRL to perform advanced manipulation task involving both hands and the SRL. These preliminary results illustrate the possibility offered by this control strategy and by controllable SRL in general to enhance operator’s manipulation abilities.
11:20-11:40 | Paper ThA3.5
Cartesian Space Feedback for Real Time Tracking of a Supernumerary Robotic Limb: A Pilot Study
Pinardi, Mattia, Campus Bio-Medico, University of Rome
Raiano, Luigi, Unit of Biomedical Robotics and Biomicrosystems, Department of E
Noccaro, Alessia, Università Campus Bio-Medico Di Roma
Formica, Domenico, Campus Bio-Medico University
Di Pino, Giovanni, Campus Biomedico University
Keywords: Human Performance - Sensory-motor, Human performance, Neuromuscular Systems - Wearable systems
Abstract: We present a system to provide the user with real time proprioceptive feedback regarding the state of a supernumerary robotic limb (SRL). The system converted the robot kinematics into a vibration amplitude-frequency value, using a custom electronic board. Four eccentric-motors placed on the leg delivered the vibrotactile pattern to subjects. We measured the accuracy in real-time tracking of the robot end-effector position and the delay from the robot movement onset. We tested four subjects in a preliminary study, and we found an average Position Error and a Delay of 0.084 ± 0.01 m and 1.169 ± 0.408 s respectively, which validated the feasibility of the presented setup. Increasing the learning phase duration should further improve subject performance. Additionally, the present platform could easily be employed to test the efficacy of dynamic feedbacks (such as joint angles and torques) for real time tracking of SRL.