Theraputic Effects of Brain-Computer Interface on Motor Recovery of Stroke Patients: A Meta-analysis

Abstract

Purpose<break>The aim of this study was to evaluate the therapeutic effects of brain-computer interface (BCI) on improving motor functions of stroke patients.<break>Methods<break>We conducted a meta-analysis on randomized controlled trials (RCTs) on BCI training for post-stroke motor rehabilitation. Relevant publications were identified from the databases of PubMed, Embase, ScienceDirect, and Cochrane Library. The standardized mean difference (SMD) with 95% confidence intervals (CI) were calculated as the pooled effect size of the motor outcome.<break>Results<break>Thirty-five of the 43 candidate articles involving 749 participants were included in this meta-analysis. Overall, both the significant immediate effect size of 0.53 and long-lasting effect size of 0.26 were found for motor outcome measured by Fugl-Meyer Assessment. A further subgroup-analysis observed larger therapeutic effects on lower-limb than upper-limb. A subgroup-analysis also indicated that stroke patients may gain better functional outcome in the subacute phase than in the chronic phase. Superior effect of BCI training was also detected for distal function of upper-limb over proximal function. BCI training combined with functional electrical stimulation (FES) was more effective than BCI combined with robot . No significant effect was found in other combined interventional methods, especially the use of transcranial direct current stimulation, which cannot potentiate the effects of BCI training. In addition, subgroup-analysis also indicated a greater effect for longer durations of intervention. And the dosage between 15 min and 180 min was found to be optimal.<break>Conclusions<break>BCI has significant immediate and long-lasting effects on improving motor function of both upper-limb and lower-limb of stroke patients. Superior therapeutic could be delivered to patients in the subacute phase and clearer benefits are evident in distal functions of upper extremity. When combined with FES, BCI seems to be more effective than when combined with robot and other external devices. Longer durations of intervention could provide better effects. But bigger is not always better for weekly dosage.