Bridging the gap between motor imagery and motor execution with a brain–robot interface

• Published in: NeuroImage (Orlando, Fla.) Vol. 108; pp. 319 - 327
• Main Authors: Bauer, Robert, Fels, Meike, Vukelić, Mathias, Ziemann, Ulf, Gharabaghi, Alireza
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.03.2015; Elsevier Limited
• Subjects: Adult; Brain - physiology; Brain research; Brain-computer interface; Brain-machine interface; Brain–robot interface; Effective connectivity; Electroencephalography; Emulation; Feedback; Female; Functional Restoration; Humans; Imagery; Imagination - physiology; Indicators; Male; Middle Aged; Motor ability; Motor Activity - physiology; Motor imagery; Motors; Networks; Neural Pathways - physiology; Neurorehabilitation; Oscillations; Patients; Perceptions; Phase slope index; Principal Component Analysis; Principal components analysis; Rehabilitation; Resting state; Robotics; Robots; Signal Processing, Computer-Assisted; Slopes; Stroke; Training; User-Computer Interface; Young Adult; Brain-computer interface; Functional Restoration; Resting state; Brain–robot interface; Motor imagery; Neurorehabilitation; Phase slope index; Brain-machine interface; Effective connectivity
• ISSN: 1053-8119; 1095-9572; 1095-9572
• DOI: 10.1016/j.neuroimage.2014.12.026

According to electrophysiological studies motor imagery and motor execution are associated with perturbations of brain oscillations over spatially similar cortical areas. By contrast, neuroimaging and lesion studies suggest that at least partially distinct cortical networks are involved in motor imagery and execution. We sought to further disentangle this relationship by studying the role of brain-robot interfaces in the context of motor imagery and motor execution networks. Twenty right-handed subjects performed several behavioral tasks as indicators for imagery and execution of movements of the left hand, i.e. kinesthetic imagery, visual imagery, visuomotor integration and tonic contraction. In addition, subjects performed motor imagery supported by haptic/proprioceptive feedback from a brain–robot-interface. Principal component analysis was applied to assess the relationship of these indicators. The respective cortical resting state networks in the α-range were investigated by electroencephalography using the phase slope index. We detected two distinct abilities and cortical networks underlying motor control: a motor imagery network connecting the left parietal and motor areas with the right prefrontal cortex and a motor execution network characterized by transmission from the left to right motor areas. We found that a brain–robot-interface might offer a way to bridge the gap between these networks, opening thereby a backdoor to the motor execution system. This knowledge might promote patient screening and may lead to novel treatment strategies, e.g. for the rehabilitation of hemiparesis after stroke. Panel A shows the component coefficients for the visual and the kinesthetic imagery scale (VIS & KIS), for performances in the brain-robotic-interface (BRI), in the tonic contraction task (EMG) and in the task based on visuomotor integration (VMI). The gap between motor imagery (VIS & KIS) and motor execution (EMG, VMI) is bridged by the BRI. Panel B shows the connections with the strongest prediction for motor imagery (blue arrows) and motor execution (red arrows). [Display omitted] •The abilities of motor imagery and motor execution are uncorrelated.•They can be predicted with high specificity from resting state connectivity.•Motor imagery is linked to a parieto-centro-frontal network.•Motor execution is linked to a bilateral motor network.•A brain–robot interface can form a bridge between these abilities and networks.