Brain-machine interface of upper limb recovery in stroke patients rehabilitation: A systematic review

Abstract

Background: Technologies such as brain-computer interfaces are able to guide mental practice, in particular motor imagery performance, to promote recovery in stroke patients, as a combined approach to conventional therapy.

Objective: The aim of this systematic review was to provide a status report regarding advances in brain-computer interface, focusing in particular in upper limb motor recovery.

Methods: The databases PubMed, Scopus, and PEDro were systematically searched for articles published between January 2010 and December 2017. The selected studies were randomized controlled trials involving brain-computer interface interventions in stroke patients, with upper limb assessment as primary outcome measures. Reviewers independently extracted data and assessed the methodological quality of the trials, using the PEDro methodologic rating scale.

Results: From 309 titles, we included nine studies with high quality (PEDro ≥ 6). We found that the most common interface used was non-invasive electroencephalography, and the main neurofeedback, in stroke rehabilitation, was usually visual abstract or a combination with the control of an orthosis/robotic limb. Moreover, the Fugl-Meyer Assessment Scale was a major outcome measure in eight out of nine studies. In addition, the benefits of functional electric stimulation associated to an interface were found in three studies.

Conclusions: Neurofeedback training with brain-computer interface systems seem to promote clinical and neurophysiologic changes in stroke patients, in particular those with long-term efficacy.

Keywords: arm; brain-computer interface; hemiplegia; neurofeedback.