Brain-actuated gait trainer with visual and proprioceptive feedback

OBJECTIVE Brain-machine interfaces (BMIs) have been proposed in closed-loop applications for neuromodulation and neurorehabilitation. This study describes the impact of different feedback modalities on the performance of an EEG-based BMI that decodes motor imagery (MI) of leg flexion and extension. APPROACH We executed experiments in a lower-limb gait trainer (the legoPress) where nine able-bodied subjects participated in three consecutive sessions based on a crossover design. A random forest classifier was trained from the offline session and tested online with visual and proprioceptive feedback, respectively. Post-hoc classification was conducted to assess the impact of feedback modalities and learning effect (an improvement over time) on the simulated trial-based performance. Finally, we performed feature analysis to investigate the discriminant power and brain pattern modulations across the subjects. MAIN RESULTS (i) For real-time classification, the average accuracy was [Formula: see text]% and [Formula: see text]% for the two online sessions. The results were significantly higher than chance level, demonstrating the feasibility to distinguish between MI of leg extension and flexion. (ii) For post-hoc classification, the performance with proprioceptive feedback ([Formula: see text]%) was significantly better than with visual feedback ([Formula: see text]%), while there was no significant learning effect. (iii) We reported individual discriminate features and brain patterns associated to each feedback modality, which exhibited differences between the two modalities although no general conclusion can be drawn. SIGNIFICANCE The study reported a closed-loop brain-controlled gait trainer, as a proof of concept for neurorehabilitation devices. We reported the feasibility of decoding lower-limb movement in an intuitive and natural way. As far as we know, this is the first online study discussing the role of feedback modalities in lower-limb MI decoding. Our results suggest that proprioceptive feedback has an advantage over visual feedback, which could be used to improve robot-assisted strategies for motor training and functional recovery.

[1]  Jose L Pons,et al.  Powered exoskeletons for bipedal locomotion after spinal cord injury , 2016, Journal of neural engineering.

[2]  Robert Leeb,et al.  Evidence Accumulation in asynchronous BCI , 2011 .

[3]  F. Reynard,et al.  Lower Limbs Robotic Rehabilitation Case Study with Clinical Trials , 2014 .

[4]  G. Pfurtscheller,et al.  EEG-based discrimination between imagination of right and left hand movement. , 1997, Electroencephalography and clinical neurophysiology.

[5]  Bram Vanderborght,et al.  An exoskeleton for gait rehabilitation: Prototype design and control principle , 2008, 2008 IEEE International Conference on Robotics and Automation.

[6]  José del R. Millán,et al.  Transferring brain-computer interfaces beyond the laboratory: Successful application control for motor-disabled users , 2013, Artif. Intell. Medicine.

[7]  Ricardo Chavarriaga,et al.  Teaching brain-machine interfaces as an alternative paradigm to neuroprosthetics control , 2015, Scientific Reports.

[8]  Ricardo Chavarriaga,et al.  Adaptive Assistance for Brain-Computer Interfaces by Online Prediction of Command Reliability , 2016, IEEE Computational Intelligence Magazine.

[9]  J. Millán,et al.  Detection of self-paced reaching movement intention from EEG signals , 2012, Front. Neuroeng..

[10]  José del R. Millán,et al.  Freeing the visual channel by exploiting vibrotactile BCI feedback , 2013, 2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC).

[11]  Jun Morimoto,et al.  Decoding the ERD/ERS: influence of afferent input induced by a leg assistive robot , 2014, Front. Syst. Neurosci..

[12]  Jose L. Contreras-Vidal,et al.  Design and Optimization of an EEG-Based Brain Machine Interface (BMI) to an Upper-Limb Exoskeleton for Stroke Survivors , 2016, Front. Neurosci..

[13]  José Luis Pons Rovira,et al.  A Closed-Loop Brain–Computer Interface Triggering an Active Ankle–Foot Orthosis for Inducing Cortical Neural Plasticity , 2014, IEEE Transactions on Biomedical Engineering.

[14]  José del R. Millán,et al.  Brain-Controlled Wheelchairs: A Robotic Architecture , 2013, IEEE Robotics & Automation Magazine.

[15]  J. del R. Millán.  Brain-Machine Interfaces: The Perception-Action Closed Loop: A Two-Learner System , 2015, IEEE Systems, Man, and Cybernetics Magazine.

[16]  Gernot R. Müller-Putz,et al.  Random Forests for Feature Selection in Non-invasive Brain-Computer Interfacing , 2013, CHI-KDD.

[17]  J. Millán,et al.  Single trial prediction of self-paced reaching directions from EEG signals , 2014, Front. Neurosci..

[18]  V. Caggiano,et al.  Proprioceptive Feedback and Brain Computer Interface (BCI) Based Neuroprostheses , 2012, PloS one.

[19]  Hannes Bleuler,et al.  The LegoPress: A Rehabilitation, Performance Assessment and Training Device Mechanical Design and Control , 2014, EuroHaptics.

[20]  Jose L. Contreras-Vidal,et al.  Multiple Kernel Based Region Importance Learning for Neural Classification of Gait States from EEG Signals , 2017, Front. Neurosci..

[21]  Arnaud Delorme,et al.  EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis , 2004, Journal of Neuroscience Methods.

[22]  Thanassis Rikakis,et al.  Adaptive Mixed Reality Rehabilitation Improves Quality of Reaching Movements More Than Traditional Reaching Therapy Following Stroke , 2013, Neurorehabilitation and neural repair.

[23]  Mark Schram Christensen,et al.  Modulation of proprioceptive feedback during functional electrical stimulation: an fMRI study , 2013, The European journal of neuroscience.

[24]  Jarrod A. Lewis-Peacock,et al.  Closed-loop brain training: the science of neurofeedback , 2017, Nature Reviews Neuroscience.

[25]  Muhammad Abd-El-Barr,et al.  Long-term Training With a Brain-Machine Interface-Based Gait Protocol Induces Partial Neurological Recovery in Paraplegic Patients. , 2016, Neurosurgery.

[26]  Yun Dong,et al.  Reliable assessment of lower limb motor representations with fMRI: Use of a novel MR compatible device for real-time monitoring of ankle, knee and hip torques , 2008, NeuroImage.

[27]  J. J. Gil,et al.  Lower-Limb Robotic Rehabilitation: Literature Review and Challenges , 2011, J. Robotics.

[28]  T. Hortobágyi,et al.  Teager–Kaiser energy operator signal conditioning improves EMG onset detection , 2010, European Journal of Applied Physiology.

[29]  C. Braun,et al.  Hand Movement Direction Decoded from MEG and EEG , 2008, The Journal of Neuroscience.

[30]  Yili Liu,et al.  EEG-Based Brain-Controlled Mobile Robots: A Survey , 2013, IEEE Transactions on Human-Machine Systems.

[31]  Abderrahmane Kheddar,et al.  Audio-visual feedback improves the BCI performance in the navigational control of a humanoid robot , 2014, Front. Neurorobot..

[32]  J. del R. Millán,et al.  Characterizing the EEG Correlates of Exploratory Behavior , 2008, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[33]  D. Farina,et al.  Detection of movement intention from single-trial movement-related cortical potentials , 2011, Journal of neural engineering.

[34]  Dong Liu,et al.  A brain-controlled exoskeleton with cascaded event-related desynchronization classifiers , 2017, Robotics Auton. Syst..

[35]  M. Castelo‐Branco,et al.  Brain Activity during Lower-Limb Movement with Manual Facilitation: An fMRI Study , 2015, Neurology research international.

[36]  Martin A. Lindquist,et al.  Dynamic connectivity detection: an algorithm for determining functional connectivity change points in fMRI data , 2015, Front. Neurosci..

[37]  J. Millán,et al.  Personalized Neuroprosthetics , 2013, Science Translational Medicine.

[38]  Ricardo Chavarriaga,et al.  Action prediction based on anticipatory brain potentials during simulated driving , 2015, Journal of neural engineering.

[39]  Luis Montesano,et al.  Control of an Ambulatory Exoskeleton with a Brain–Machine Interface for Spinal Cord Injury Gait Rehabilitation , 2016, Front. Neurosci..

[40]  Junichi Ushiba,et al.  EEG-based classification of imaginary left and right foot movements using beta rebound , 2013, Clinical Neurophysiology.

[41]  Jose L. Contreras-Vidal,et al.  Sitting and standing intention can be decoded from scalp EEG recorded prior to movement execution , 2014, Front. Neurosci..

[42]  Dario Farina,et al.  Factors of Influence on the Performance of a Short-Latency Non-Invasive Brain Switch: Evidence in Healthy Individuals and Implication for Motor Function Rehabilitation , 2016, Front. Neurosci..

[43]  G. Pfurtscheller,et al.  Enhancement of left-right sensorimotor EEG differences during feedback-regulated motor imagery. , 1999, Journal of clinical neurophysiology : official publication of the American Electroencephalographic Society.

[44]  F. L. D. Silva,et al.  Event-related EEG/MEG synchronization and desynchronization: basic principles , 1999, Clinical Neurophysiology.

[45]  Ashley N. Johnson,et al.  Dual-task motor performance with a tongue-operated assistive technology compared with hand operations , 2012, Journal of NeuroEngineering and Rehabilitation.

[46]  R. Riener,et al.  Towards more effective robotic gait training for stroke rehabilitation: a review , 2012, Journal of NeuroEngineering and Rehabilitation.

[47]  N. Birbaumer,et al.  ERD-Based Online Brain–Machine Interfaces (BMI) in the Context of Neurorehabilitation: Optimizing BMI Learning and Performance , 2011, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[48]  Oded Nov,et al.  Increasing Patient Engagement in Rehabilitation Exercises Using Computer-Based Citizen Science , 2015, PloS one.