RESUMO
Objective:To systematically evaluate the effect of brain-computer interface (BCI) on upper-limb motor function after stroke, and compare the effects under different interfaces. Methods:Randomized controlled trials (RCTs) about BCI for upper-limb motor function after stroke were retrieved from databases of PubMed, Embase, Cochrane Library, Web of Science, CNKI, Wanfang Data and CBM, from inception to October, 2020. The quality of the trials was assessed and the data were extracted according to the Cochrane Handbook of Systematic Review. A meta-analysis was carried out with RevMan 5.3 and ADDIS 1.16.8. Results:Ultimately, 14 RCTs involving 504 patients were included. Meta-analysis showed that BCI could obviously improve the Fugl-Meyer Assessment-Upper Extremities (FMA-UE) score (MD = 6.81, 95%CI 1.51 to 12.11, P < 0.05), Action Research Arm Test score (MD = 7.68, 95%CI 0.49 to 14.88, P < 0.05) and modified Barthel Index score (MD = 8.91, 95%CI 5.57 to 12.25, P < 0.001) after stroke. Subgroup analysis showed that FMA-UE score could be improved by BCI for both more than four weeks (MD = 9.44, 95%CI 1.83 to 17.04, P < 0.05) and less than four weeks (MD = 5.18, 95%CI 2.84 to 7.51, P < 0.001). For the types of interface, the probabilities of the best effects from network meta-analysis ranked as electrical stimulator (P = 0.53), visual feedback (P = 0.41) and machine assistance (P = 0.06). Conclusion:BCI, especially with electrical stimulator interface, could obviously improve upper-limb motor function and activities of daily living for stroke patients.