Resting state changes in functional connectivity correlate with movement recovery for BCI and robot-Assisted upper-extremity training after stroke

Background. Robot-Assisted training may improve motor function in some hemiparetic patients after stroke, but no physiological predictor of rehabilitation progress is reliable. Resting state functional magnetic resonance imaging (RS-fMRI) may serve as a method to assess and predict changes in the motor network. Objective. The authors examined the effects of upper-extremity robot-Assisted rehabilitation (MANUS) versus an electroencephalography-based brain computer interface setup with motor imagery (MI EEG-BCI) and compared pretreatment and posttreatment RS-fMRI. Methods. In all, 9 adults with upper-extremity paresis were trained for 4 weeks with a MANUS shoulder-elbow robotic rehabilitation paradigm. In 3 participants, robot-Assisted movement began if no voluntary movement was initiated within 2 s. In 6 participants, MI-BCI-based movement was initiated if motor imagery was detected. RS-fMRI and Fugl-Meyer (FM) upper-extremity motor score were assessed before and after training. Results. The individual gain in FM scores over 12 weeks could be predicted from functional connectivity changes (FCCs) based on the pre-post differences in RS-fMRI measurements. Both the FM gain and FCC were numerically higher in the MI-BCI group. Increases in FC of the supplementary motor area, the contralesional and ipsilesional motor cortex, and parts of the visuospatial system with mostly association cortex regions and the cerebellum correlated with individual upper-extremity function improvement. Conclusion. FCC may predict the steepness of individual motor gains. Future training could therefore focus on directly inducing these beneficial increases in FC. Evaluation of the treatment groups suggests that MI is a potential facilitator of such neuroplasticity.