Efficacy of the Conventional Rehabilitation Robot and bio-Signal Feedback-Based Rehabilitation Robot on Upper-Limb Function in Patients with Stroke: A Systematic Review and Network Meta-Analysis

BackgroundWith the development of modern biomedical engineering, bio-signal feedback-based robots, such as electromyography (EMG)-based and brain-computer interface (BCI)-based rehabilitation robot, have emerged beyond conventional designs. However, their comparative effectiveness for improving upper limb function in stroke patients remains unassessed.ObjectiveTo evaluate the comparative effectiveness and ranking of the conventional rehabilitation robot and bio-signal feedback-based rehabilitation robot in improving upper limb function in stroke patients.MethodsPubMed, EMBASE, Cochrane Library, CINAHL, PEDro, EI, IEEEXplore, ClinicalTrials.gov, ICTRP, and ISRCTN Registry were searched for randomized controlled trials (RCTs) from their inception to December 25, 2024. The risk of bias was assessed using the Cochrane Risk of Bias tool (RoB 2.0) and evidence certainty with the GRADE (Grading of Recommendations Assessment, Development and Evaluation) approach. Network meta-analyses were performed using a random-effects model within a frequentist framework.Results59 RCTs with 3,387 participants were included. Based on the surface under the cumulative ranking curve (SUCRA), the BCI-based rehabilitation robot demonstrated the highest overall effects (SUCRA: 99.9%), short-term effects (SUCRA: 99.4%), and long-term effects (SUCRA: 85.1%), though its long-term effects were not significant (mean difference: 2.21; 95% confidence interval: -0.79, 5.21). The EMG-based rehabilitation robot outperformed the conventional rehabilitation robot in short-term interventions (SUCRA: 59.8% vs. 40.3%), but it did not have the same advantage in long-term interventions (SUCRA: 27.1% vs. 66.8%).ConclusionsThe BCI-based rehabilitation robot might be the best choice for improving upper limb function in stroke patients. Future studies should focus on the intervention time for the EMG-based rehabilitation robot.