Contribution to collective works (Parts of books)
Recurrent Neural Networks for Temporal Data Processing: Toward an Integrative Dynamic Recurrent Neural Network for Sensorimotor Coordination Dynamics
Cheron, Guy; Duvinage, Matthieu; Castermans, Thierry et al.
2011In Recurrent Neural Networks
 

Files


Full Text
Proof Chapter DRNN2010-1.pdf
Author postprint (470.22 kB)
Request a copy

All documents in ORBi UMONS are protected by a user license.

Send to



Details



Abstract :
[en] The utilization of dynamic recurrent neural networks (DRNN) for the interpretation of biological signals coming from human brain and body has acquired a significant growth in the field of brain-machine interface. DRNN approaches may offer an ideal tool for the identification of input-output relationships in numerous types of neural-based signals, such as intracellular synaptic potentials, local field potentials, EEG and EMG which integrate multiple sources of activity distributed across large neuronal assemblies. In the field of motor control, the output signals of the DRNN mapping concern movement-related kinematics signals such as position, velocity or acceleration of the different body segments involved. The most direct input signals consist in the electromyographic signals (EMG) recorded over the different superficial muscles implicated in movement generation. It is generally recognized that the non-invasive recording of the EMG envelope signal represents a reasonable reflection of the firing rate of the motoneuronal pools including both central and afferent influences (Cheron and Godaux, 1986). In addition, the combination of the multiple EMG signals may reveal the basic motor coordination dynamics of the gesture (Scholz and Kelso, 1990; Kelso, 1995). A major interest of the EMGs to kinematics mapping by a DRNN is that it may represent a new indirect way for a better understanding of motor organization elaborated by the central nervous system. After the learning phase and whatever the type of movement, the identification performed by the DRNN offers a dynamic memory which is able to recognize the preferential direction of the physiological action of the studied muscles (Cheron et al. 1996, 2003, 2006, 2007). In this chapter, we present the DRNN structure and the training procedure applied in case of noisy biological signals. Different DRNN applications are here reviewed in order to illustrate their usefulness in the field of motor control as diagnostic tools and potential prosthetic controllers.
Research center :
BIOSYS - Biosys
Disciplines :
Laboratory medicine & medical technology
Author, co-author :
Cheron, Guy ;  Université de Mons > Faculté de Psychologie et des Sciences de l'Education > Electrophysiologie
Duvinage, Matthieu ;  Université de Mons > Faculté Polytechnique > Information, Signal et Intelligence artificielle
Castermans, Thierry ;  Université de Mons > Faculté Polytechnique > Information, Signal et Intelligence artificielle
Leurs, F.
Cebolla, A.M.
Bengoetxea, A.
De Saedeleer, C.
Petieau, Mathieu
Hoellinger, Thomas
Seetharaman, Karthik
Draye, Jean-Philippe
Dan, B.
Language :
English
Title :
Recurrent Neural Networks for Temporal Data Processing: Toward an Integrative Dynamic Recurrent Neural Network for Sensorimotor Coordination Dynamics
Publication date :
02 March 2011
Main work title :
Recurrent Neural Networks
Pages :
67-80
Research unit :
P310 - Electrophysiologie
F105 - Information, Signal et Intelligence artificielle
Research institute :
R500 - Institut des Sciences et du Management des Risques
Available on ORBi UMONS :
since 19 January 2011

Statistics


Number of views
4 (0 by UMONS)
Number of downloads
0 (0 by UMONS)

Bibliography


Similar publications



Contact ORBi UMONS