Brainwave control of a wearable robotic arm for rehabilitation and neurophysiological study in Cervical Spine Injury

01/04/14 to 31/03/15

CSI:Brainwave project’s full title is “Brainwave control of a wearable robotic arm for rehabilitation and neurophysiological study in Cervical Spine Injury” .

It constitutes a multidisciplinary neurophysiological project, developed by the Biomedical Electronics Robotics & Devices (BERD) research group, which is housed in the Medical Physics Laboratory of Aristotle University of Thessaloniki, Greece. The project is supported by two Neurosurgical Departments and has received funding by the 2013 “Mario Boni” Research Grant, awarded by the Cervical Spine Research Society – European Section (CSRS-ES).

The project meets critical acclaim wherever is presented. Recently the Hellenic Spine Society ( awarded the presentation of CSI:Brainwave project with the 2016 “Vicky Zachariou” award during the 10th Hellenic Spine Congress & Joint Meeting with French and Turkish Spine Society, that took place on 26-29 October at the Makedonia Palace Hotel in Thessaloniki ( For more info, please read the relevant piece of news (

This project started on April 2014 and, while originally was planned to run for three years (expected completion early 2017), it has been extended and currently recruits patients (as of December 2016). The project is registered with under NCT02443558 (

The CSI:Brainwave project involves:

  • A clinical study for rehabilitation of patients with Cervical Spinal Cord Injury (CSCI), using a Brain-Computer Interface (BCI) controlled wearable robotic arm.
  • A comparative neurophysiological analysis of cortical activation, connectivity and plasticity in patients with CSCI undergoing motor imagery (MI) practice.

Primary objectives:

  • To develop, test and optimize a mountable robotic arm controlled with wireless BCI.
  • To develop and validate self-paced neuro-rehabilitation protocols for patients with CSCI.
  • To identify and study the neurophysiological functionality and alteration of cortical activity in acute and chronic CSCI.

The CSI:Brainwave project aims at allowing patients suffering from tetraplegia due to CSCI to perform brainwave modulation, practicing Visual & Kinesthetic Motor Imagery (VMI  & KMI) and offering neurofeedback with the form of control of a 8-degree-of-freedom bimanual robot.

Eventually, we aim to design the robotic arms as to mount on the patient’s plegic arm and allow the patient to directly control it using a BCI module. The project aims at demonstrating the added value of neurofeedback for rehabilitation and/or motor restoration of CSCI patients.

The project makes use of the “MERCURY” robot, which consists of a bimanual, fully digital, 6-degrees-of-freedom robotic arms apparatus, which was developed under CSI:Brainwave to reach version 2.0. The robot was initially constructed and developed by members of the BERD group. Please see the MEDIA section at the end of this page.

Under the auspices of the project, the research group gained permission to translate and validate the Spinal Cord Independence Measure in the greek language to be used for patient evaluation. More relevant evaluation tools are also involved and the research team has asked for official permissions for their translation, as well.

Background & Methodology

In terms of brain networks we aim to: 1) identify how reorganization can be correlated according to neurological condition, 2) identify adaptive and maladaptive changes in connectivity and 3) produce imaging “biomarkers” of neurological condition & improvement. We also aim to 4) measure the impact of brain-computer interface (BCI) training in brain networks following SCI. Moreover the project also aims at 5) developing and validating a non-invasive network-based BCI for rehabilitation and motor restoration of SCI patients making use of the derived knowledge and BCI-controlled anthropomorphic robotic arms.

Spinal Cord Injury (SCI) causes alterations in brain organization and structure, including changes in brain connectivity (BC), such as abnormal alterations and pathological brain activations (Freund et al. 2013; Nardone et al. 2013). Those have been shown even during the early stages of SCI and they correlate with the degree of neurological impairment (Nardone et al. 2013). fMRI studies within the first month post-injury depicted altered spontaneous resting-state brain activation in almost all cortical and sub-cortical sensorimotor areas and at mean two months post-injury showed structural changes and network alterations (Hou et al. 2014a; Hou et al. 2014b; Hou et al. 2016). Decreased inter-hemispheric resting-state functional connectivity (FC) and increased intra-hemispheric resting-state FC within the sensorimotor cortex, premotor area, supplementary motor area (SMA) and other nodes of the motor pathways have been calculated (Min et al 2015a; Min et al 2015b). Somatosensory networks showed decreased connectivity in SCI patients compared to healthy controls while predictors of good versus poor neurological recovery at 6 months post-injury has been theorized that can be deduced from such reorganization (Hou et al. 2014b). During chronic phases of complete SCI, the continued disruption of sensorimotor pathways causes reorganization of resting-state functional networks (Astolfi et al. 2007; Mattia et al. 2009).

These alterations have not been systematically investigated and are often not taken into consideration in neurophysiological experiments or rehabilitation practices, while the most important questions regard whether and when such reorganization is adaptive (and should be promoted) or maladaptive. In light of very recent research findings reported by Donati et al. (2016), that show that BCI use for neurorehabilitation by SCI patients promotes an initially unexpected plasticity-induced neurological recovery, research on brain network plasticity following SCI should be considered timely and of great importance to the field. While studies on the field have been published, we aim to be among the first to systematically produce classifiable results regarding reorganization that occurs after SCI and BCI training.

Eligible participants for the study include female & male healthy individuals and SCI patients. Inclusion criteria are: clinical diagnosis of SCI (evaluated by ASIA scores & Impairment Scale) or healthy participants (age and gender matched to SCI patients), sufficient documentation of the injury (recorded neurological examination and history, imaging of the injury by MRI or other means). Exclusion criteria are: other neurological hat has a possibility to significantly affect the neurological status of the participants (or) the ability to control a BCI (or) the neurophysiological recordings (such as traumatic brain injury, CNS tumors, Multiple Sclerosis, Amyotrophic Lateral Sclerosis, Parkinson's disease or refractory epilepsy), other grave medical condition that could affect the participation (or) the safety of the participants (such as cardiac deficiency, pulmonary deficiency, hearing and visual impairments that can affect the participant's understanding of the intervention and performance, illegal drug use or chronic alcoholism).

Estimated enrollment is 30 SCI patient participants. Participants are classified into four categories: a) Experimental: Complete Injury. Patients suffering from complete injury at the cervical spinal cord level (ASIA Impairment Scale A). b) Experimental: Incomplete Injury, patients suffering from incomplete injury at the cervical spinal cord level (ASIA Impairment Scale B,C,D,E). c) Active Comparator: Non-cervical injury, patients suffering from complete or incomplete injury of the spinal cord at a level other than the cervical (thoracic or lumbar). d) Active Comparator: Healthy participants, age and sex matched to those of the other arms. Assessment of neurological condition, as well as of alterations (or possible recovery) of function will be made with the ASIA scores and impairment scale, the greek translation of the Spinal Cord Independence Measure (g-SCIM-III) and where appropriate with the Walking Index for Spinal Cord Injury (WISCI). Functional status will be re-assessed at 6 months and 1 year after the intervention.

Participants, after initial interview and examination will also be subjected to a brain MRI scan (to be used as an individual model in subsequent brain networks analysis). The experimental procedure consists of two parts. During the first part, they will be performing Kinesthetic and Visual (while presented with relevant videos) Motor Imagery (MI) tasks of the upper extremities (both arms, 7 degrees-of-freedom) and a MI walking task, while under high-resolution (128channels) EEG recording. During the second part, they will be controlling the bimanual robot with the 14channel Emotiv EEG and the home-built BCI program. All neurophysiological recordings (both of 128channels and of 14channels) will be subjected to an offline brain network analysis (Athanasiou et al. 2016b). This will be performed in Matlab using appropriate toolboxes (brainstorm, EEGLAB, eConnectome) using the participants own individual brain model (extracted from base MRI scan). Connectivity algorithms used will also include time-adaptive techniques. Subject to sufficient funding, the experimental procedures and analysis will be repeated at the 1-year post-intervention period.


Project news

Distinction for CSI: Brainwave research project of the Medical Physics Lab of AUTH

Dec 05, 2016 / 0 Comments

Distinction for CSI:Brainwave research project of the Medical Physics Lab of AUTH

Project Type: