Effects of dual-task mode brain-computer interface based on motor imagery and virtual reality on balance and attention in patients with stroke: a randomized controlled pilot trial.

Only 30 participants were analysed, and the brain-computer interface group had a higher dropout rate of 26.3%. The primary outcome was changed after registration from Fugl-Meyer Lower Extremity and Montreal Cognitive Assessment to Berg Balance Scale, and the authors say this introduces outcome reporting bias. No correction for multiple outcomes was applied, so the positive change-score findings are exploratory. The brain-computer interface group also received usual rehabilitation activities from physical therapists and occupational therapists, and the intervention combined motor imagery, EEG feedback, robot pedaling and virtual reality, so the contribution of each component cannot be separated. EEG was taken only from the prefrontal cortex, and the system could not objectively verify that participants were consistently doing motor imagery as intended. Participants could walk at least 10 m with or without assistance, had Montreal Cognitive Assessment attention dimension scores of 2 to 5 points, and excluded moderate to severe cognitive impairment, neglect, aphasia, and certain lower-limb pain or spasticity, so results may not apply to more impaired or excluded patients.

Funding was provided by Jiangsu Provincial People's Hospital, the Clinical Diagnosis and Treatment Technology Open bidding for selecting the best candidates Project, the Major sports research projects of Jiangsu Sports Bureau, and the Ministry of Industry and Information Technology and National Health Commission High-end Equipment Promotion and Application Project. The supplied text does not give an author conflict-of-interest declaration.

Do not conclude that dual-task brain-computer interface pedaling is proven to improve balance. The primary between-group Berg Balance Scale scores did not differ, and the positive result came from a change score in a small pilot with a switched primary outcome, multiple outcomes and no correction for multiplicity.