Effect of electroencephalography-based motor imagery neurofeedback on mu suppression during motor attempt in patients with stroke

Abstract

Objective: The primary aim of this study was to explore the neurophysiological effects of motor imagery neurofeedback using electroencephalography (EEG), specifically focusing on mu suppression during serial motor attempts, and to assess its potential benefits in patients with subacute stroke.

Methods: A total of 15 patients with hemiplegia following subacute ischemic stroke were prospectively enrolled in this randomized cross-over study. This study comprised two experiments: neurofeedback and sham. Each experiment included four blocks: three blocks of resting, grasp, resting, and an interventional task, followed by one block of resting and grasp. During the resting sessions, participants fixated on a white cross on a black background for 2 min without moving their upper extremities. In the grasp sessions, participants were instructed to grasp and release their paretic hand at a frequency of about 1 Hz for 3 min while maintaining fixation on the white cross. During the interventional task, the neurofeedback presented a punching image using the affected upper limb, corresponding to the mu suppression induced by imagined movement for 3 min. In contrast, the sham presented an image based on mu suppression data from randomly selected participants. EEG data were recorded throughout the experiment, and data from electrodes C3/C4 and P3/P4 were analyzed to compare the degree of mu suppression between the neurofeedback and sham experiments.

Results: Significant mu suppression was observed in the bilateral motor and parietal cortices during the neurofeedback experiment compared with the sham across serial sessions (p < 0.001). Following neurofeedback, real grasping sessions showed progressive strengthening of mu suppression in the ipsilesional motor cortex and bilateral parietal cortices compared to sessions following sham (p < 0.05). This effect was not observed in the contralesional motor cortex.

Conclusions: Motor imagery neurofeedback significantly enhances mu suppression in the ipsilesional motor and bilateral parietal cortices during motor attempts in patients with subacute stroke. These findings suggest that motor imagery neurofeedback could serve as a promising adjunctive therapy to enhance motor-related cortical activity and support motor rehabilitation in patients with stroke.

Keywords: Electroencephalography; Motor imagery; Mu suppression; Neurofeedback; Rehabilitation; Stroke; Upper extremity.

Fig. 1

(A) A participant engaged in the neurofeedback intervention. (B) Study design

Fig. 2

(A) Topography of mu suppression in electroencephalography during the intervention. The lesion side was standardized to the left hemisphere. (B) Line charts depicting mu suppression in the ipsilesional and contralesional motor cortices, as well as the ipsilesional and contralesional parietal cortices during the intervention. Results of repeated measures ANOVA are shown below the charts. Statistical significance from post-hoc analysis between neurofeedback and sham at each intervention is presented as *p < 0.05 or **p < 0.01

Fig. 3

(A) Topography of mu suppression in electroencephalography during grasp. The lesion side was standardized to the left hemisphere. (B) Line charts depicting mu suppression in the ipsilesional and contralesional motor cortices, as well as the ipsilesional and contralesional parietal cortices during grasp. Results of repeated measures ANOVA are shown below the charts. Statistical significance from post-hoc analysis between neurofeedback and sham at each intervention is presented as *p < 0.05 or **p < 0.01

Fig. 4

Scatter plots showing the relationship between the change in mu suppression (fourth grasp minus first grasp) and mu suppression in (A) ipsilesional and contralesional motor cortices and (B) ipsilesional and contralesional parietal cortices. Spearman correlation coefficients and p-values are presented