Changes in electroencephalography complexity and functional magnetic resonance imaging connectivity following robotic...

Introduction: In recent years, robotic training has been utilized for recovery of motor control in patients with motor deficits. Along with clinical assessment, electrical patterns in the brain have emerged as a marker for studying changes in the brain associated with brain injury and rehabilitation. These changes mainly involve an imbalance between the two hemispheres. We aimed to study the effect of brain computer interface (BCI)-based robotic hand training on stroke subjects using clinical assessment, electroencephalographic (EEG) complexity analysis, and functional magnetic resonance imaging (fMRI) connectivity analysis. Method: Resting-state simultaneous EEG-fMRI was conducted on 14 stroke subjects before and after training who underwent 20 sessions robot hand training. Fractal dimension (FD) analysis was used to assess neuronal impairment and functional recovery using the EEG data, and fMRI connectivity analysis was performed to assess changes in the connectivity of brain networks. Results: FD results indicated a significant asymmetric difference between the ipsilesional and contralesional hemispheres before training, which was reduced after robotic hand training. Moreover, a positive correlation between interhemispheric asymmetry change for central brain region and change in Fugl Meyer Assessment (FMA) scores for upper limb was observed. Connectivity results showed a significant difference between pre-training interhemispheric connectivity and post-training interhemispheric connectivity. Moreover, the change in connectivity correlated with the change in FMA scores. Results also indicated a correlation between the increase in connectivity for motor regions and decrease in FD interhemispheric asymmetry for central brain region covering the motor area. Conclusion: In conclusion, robotic hand training significantly facilitated stroke motor recovery, and FD, along with connectivity analysis can detect neuroplasticity changes.