Neurofeedback for Stroke Rehabilitation

The Use of Real-Time fMRI and a Mobile EEG System to Provide Neurofeedback to Stroke Patients to Promote Neural Plasticity for Motor Rehabilitation.

Real-time neurofeedback aims to alter brain activation patterns through online feedback of ongoing brain activity using magnetic resonance imagining (MRI). Stroke survivors will be randomised to receive 3 sessions of real or sham neurofeedback. This study aims to investigate whether: 1) stroke survivors can maintain alterations in brain activity after the feedback is removed, 2) neurofeedback training leads to improvements in movement of the hand and arm, 3) neurofeedback training leads to changes in brain structure and function, 4) variability in response across people can be understood.

Many stroke survivors experience impairment in upper limb function, reducing independence in activities of daily living. These impairments are associated with atypical brain activity patterns. Real-time neurofeedback aims to alter brain activation patterns through online feedback of ongoing brain activity using magnetic resonance imagining (MRI). Patterns of brain activity are displayed to a participant while a task is being performed. The participant is instructed to try to alter the patterns in a particular way, promoting specific brain activity patterns. Previous studies have found that people with and without stroke are capable of utilising the feedback to alter their brain activity. This study aims to investigate whether: stroke survivors can maintain alterations in brain activity after the feedback is removed neurofeedback training leads to improvements in movement of the hand and arm neurofeedback training leads to changes in brain structure and function variability in response across people can be understood. 30 stroke survivors (> 6 months after stroke), with residual upper limb impairment, will be recruited between February 2018 and December 2020. Participants will be randomised to receive 3 sessions of real or sham neurofeedback over one week, taking place at the Wellcome Centre for Integrative Neuroimaging, University of Oxford. Changes in brain activity during affected hand movements will be assessed with and without feedback using functional MRI and after feedback sessions using electroencephalography (EEG). Brain connectivity and structure will also be assessed using MRI at baseline and at a follow-up one week later. Clinical measures of upper limb function and impairment will be performed at baseline and at follow up sessions one week and one month later (Action Research Arm Test, Fugl-Meyer upper limb assessment, Jebsen Taylor hand function test), and in each session following neurofeedback (Jebsen taylor test).

Participants are randomly allocated (1:1 ratio) to either the intervention group (Real Neurofeedback) or the control group (Sham Neurofeedback). Randomisation is done after participants undergo baseline measurements using a computer-generated minimisation method that takes into account baseline upper limb function (Action Research Arm test score) and time since stroke.

A visual representation of the participants brain activity during movement of their affected hand in the MRI scanner.

Lateralisation of brain activity during movement of the affected hand, assessed using functional magnetic resonance imaging (fMRI) blood oxygen level dependent (BOLD) signal. The activation in the region of interest is calculated for each hemisphere and the laterality index calculated as: (ipsilesional hemisphere - contralesional hemisphere) / (ipsilesional hemisphere + contralesional hemisphere). As such, positive values are indicative of greater activation in the ipsilesional hemisphere.

Performance on the Jebsen Taylor hand function test (time, in seconds to complete specified activities reflecting daily living)

Change in lateralisation of brain activity during movement of the affected hand, assessed using functional magnetic resonance imaging (fMRI) blood oxygen level dependent (BOLD) signal.

Lateralisation of brain activity during a visuomotor squeeze task, assessed using functional magnetic resonance imaging (BOLD signal). The activation in the region of interest is calculated for each hemisphere and the laterality index calculated as: (ipsilesional hemisphere - contralesional hemisphere) / (ipsilesional hemisphere + contralesional hemisphere). As such, positive values are indicative of greater activation in the ipsilesional hemisphere.

Action research arm test score (ARAT; upper limb function). Range 0-57, higher numbers indicate better upper limb function

Change in action research arm test score (ARAT; upper limb function). Range 0-57, higher numbers indicate better upper limb function

Upper limb Fugl Meyer assessment score (upper limb impairment). Range 0-66, higher numbers indicate less upper limb impairment

Upper limb Fugl Meyer assessment score (upper limb impairment). Range 0-66, higher numbers indicate less upper limb impairment

Change in integrity of the white matter tracts, assessed using diffusion tensor imaging Analysis nearly completed

Change in white matter microstructure, specifically myelin content, assessed using MRI Multi-Parameter Mapping

The correlation between baseline measures and change in lateralisation of brain activity during movement of the affected hand will be tested in order to identify markers to explain variability in response to real neurofeedback. Analysis still in progress

Inclusion Criteria: Stroke > 6 months previously Unilateral upper limb impairment, but physically able to complete the tasks required Exclusion Criteria: Contraindications to MRI, such as a pacemaker, metallic implants or aneurysm clips Inability to provide informed consent Inability to actively participate in the research procedures

Analysis code may be shared following publication of the study results. There is no set end date for this

Sanders ZB, Fleming MK, Smejka T, Marzolla MC, Zich C, Rieger SW, Luhrs M, Goebel R, Sampaio-Baptista C, Johansen-Berg H. Self-modulation of motor cortex activity after stroke: a randomized controlled trial. Brain. 2022 Oct 21;145(10):3391-3404. doi: 10.1093/brain/awac239.