Establishing Clinical Protocols for BCI-Based Motor Rehabilitation in Individuals Post Stroke - The Impact of Feedback Type and Selected Outcome Measures: A Systematic Review

Abstract

Stroke is a major cause of disability resulting in multiple system impairments. Limited extended care resulted in prioritizing high level repetitions of task-specific activities to improve function. One such modality is BCI to drive motor rehabilitation. While several systematic reviews and meta-analyses highlight the benefits of utilizing BCIs to enhance motor recovery, it is still unclear how these interventions facilitate rehabilitation of motor function in individuals post-stroke. This systematic review analyzed outcome measures and type of feedback during BCI interventions to inform future protocol development. Included articles were held to rigorous criteria, and potential studies were assessed for methodological quality using the PEDro Scale. Only articles that scored six or greater were included for analysis, and nine randomized controlled trials were included. In brief, the randomized controlled trials demonstrated that BCI enhanced the motor function of the upper extremity as measured by the FMA UE, however no other consistent outcome measures of function or self-efficacy were reported. EEG and ERD of the affected sensorimotor cortices were significantly enhanced in the BCI groups (p < 0.05). For those studies that measured retention of function, long-lasting improvements were noted, and BCI coupled to FES elicited significant, clinically relevant motor recovery. Somatosensory/motor and visual feedback were the most common across reviewed studies. While each of the studies had a wide variety of methods, all the evidence suggested that subjects improved. These findings suggest that the most important concept in protocol development may have been the incorporation of principles of motor learning.