CC BY
60
Volga Neuroscience School 2016 Astroglial control of rhythm genesis in the brain References
1. G. Dobreva, M Chahrour, M Dautzenberg L. ChiriveDa, B. Kanzler, I. Fariñas, G. Karsenty andR Grosschedl, Cell,2006, 125(5), 971-986.
Human-Machine Interface Built on sEMG Toolkit with Artificial Neural Network Feature Classifier
I.A. Kastalskiy1 *, V.A. Makarov1,2 and S.A. Lobov1
1 Lobachevsky State University of Nizhny Novgorod, Russia;
2 Universidad Complutense de Madrid, Spain. * Presenting e-mail: kastalskiy@neuro.nnov.ru
EMG (electromyographic) signal is a superposition of action potentials generated by motor units. Signal recording typically implement via the surface electrodes placed on the skin [1,2]. It is known that sEMG (surface EMG) toolkit is widely used to interpret bioelectric patterns for controlling a variety of devices [3-8]. The aim of this study was to investigate the possibility of using a wearable sEMG bracelet to control electronic devices, including interaction with the personal computer. In order to execute the set of commands needed to control the position of the cursor on the computer screen, we propose an artificial neural network (ANN) driven by the signals recorded on arm. The ANN passed through a supervised learning is able to measure the degree of arm muscles effort during movement and classify gestures. It has been shown that the human-computer interface allows controlling the cursor remotely by hand movements and simulating mouse clicks by clenched fist. The average classification accuracy of six gestures (right, left, up, down, left (single) click, right (double) click) varies around 97%.
We used the bracelet MYO (Thalmic Labs) with eight equispaced sEMG sensors acquiring raw myographic signals which were being sent through a bluetooth interface to a PC. The software allows for recognition of hand gestures and estimating muscle efforts that control the cursor on the screen in a way similar that one can achieve with ordinary computer mouse. We used root mean square (RMS) value calculating to evaluate the EMG signal obtained by each electrode. The RMS data, as a composite feature of the current hand gesture, are fed into an ANN. The network neurons apply weighted sum over inputs and use sigmoidal activation function to generate the output. The learning, i.e., adjustment of the neuron weights, is achieved by the back-propagation algorithm [9].
As a result, the cursor movement direction is defined by gestures, while its speed is controlled by the degree of muscle contraction. This significantly improved the user experience. Experimental data shows that all users were able to move the cursor and simulate left and right mouse clicks.
Acknowledgements
This work was supported by the Russian Science Foundation under project 15-12-10018. References
1. Bishop MD, Pathare N. Considerations for the use of surface electromyography. Phys Ther Korea. 2004;11(4):61-69.
2. Pullman SL, Goodin DS, Marquinez AI, Tabbal S, Rubin M. Clinical utility of surface EMG Report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Neurology. 2000;55(2):171-177.
3. Chowdhury A, Ramadas R, Karmakar S. Muscle Computer Interface: A Review. In: Chakrabarti A, Prakash RV, eds. ICoRD'13, Lecture Notes in Mechanical Engineering. Springer India 2013; 2013:411-421.
4. Hahne JM, Biessmann F, Jiang N, et al. Linear and nonlinear regression techniques for simultaneous and proportional myoelectric control. IEEE Trans Neural Syst Rehabil Eng. 2014;22(2):269-279.
5. Lorrain T, Jiang N, Farina D. Influence of the training set on the accuracy of surface EMG classification in dynamic contractions for the control of multifunction prostheses. J Neuroeng Rehabil. 2011;8:25.
6. Mironov VI, Lobov SA, Kastalskiy IA, Kazantsev VB. Myoelectric Control System of Lower Limb Exoskeleton for Re-training Motion Deficiencies. Lect Notes Comput Sci. 2015;9492:428-435.
7. Naik GR, Kumar DK, Palaniswami M. Multi run ICA and surface EMG based signal processing system for recognizing hand gestures. In: 8th IEEE International Conference on Computer and Information Technology, 2008. IEEE; 2008:700-705.
8. Peerdeman B, Boere D, Witteveen HJB, et al. Myoelectric forearm prostheses: State of the art from a user-centered perspective. J Rehabil Res Dev. 2011;4s(6):719-738.
9. Rumelhart DE, Hinton GE, Williams RJ. Learning Internal Representations by Error Propagation. In: Parallel Distributed Processing. San Diego: La Jolla Institute for Cognitive Science, California University; 1985:318-362.
OM&P
Opera Med Physiol 2016 Vol. 2 (S1) 89
