Enhanced Functional Connectivity of Cortical-Subcortical Sensorimotor Areas After Brain-Controlled Supernumerary Robotic Finger Training: A Longitudinal fMRI Study

Brain computer interface (BCI) technology can decode motion intentions to manipulate supernumerary robotic finger (SRF), facilitating direct human-machine control and enhancing limb motion capabilities. However, the longitudinal neuroplasticity mechanism of BCI-controlled SRF training using the sixth finger motor imagery paradigm remains unclear. In this study, we observed concurrent changes in behavior and brain activity following 4 weeks of BCI-controlled SRF training (BCI_SRF) or sham SRF training (Sham). The results demonstrated a 132% improvement in SRF-finger opposition accuracy with BCI_SRF training compared to Sham training. Resting-state fMRI analysis revealed a significant enhancement in functional connectivity between M1_ R and Cerebellum_8_L due to BCI-SRF training. These findings suggest that active brain involvement enhances the cortical-subcortical interaction of the sensorimotor network, thereby improving human-computer interaction performance. Our research provides valuable insights for enhancing human-computer interaction and exhibits significant potential in hand rehabilitation post-stroke.