Recovery-related indicators of motor network plasticity according to impairment severity after stroke

Background and purpose: Brain connectivity analysis has been widely used to investigate brain plasticity and recovery-related indicators of patients with stroke. However, results remain controversial because of interindividual variability of initial impairment and subsequent recovery of function. In this study, we aimed to investigate the differences in network plasticity and motor recovery-related indicators according to initial severity. Methods: We divided participants (16 males and 14 females, aged 54.2 +/- 12.0 years) into groups of different severity by Fugl-Mayer Assessment score, i.e. moderate (50-84), severe (20-49) and extremely severe (<20) impairment groups. Longitudinal resting-state functional magnetic resonance imaging data were acquired at 2 weeks and 3 months after onset. The differences in network plasticity and recovery-related indicators between groups were investigated using network distance and graph measurements. Results: As the level of impairment increased, the network balance was more disrupted. Network balance, interhemispheric connectivity and network efficiency were recovered at 3 months only in the moderate impairment group. However, this was not the case in the extremely severe impairment group. A single connection strength between the ipsilesional primary motor cortex and ventral premotor cortex was implicated in the recovery of motor function for the extremely severe impairment group. The connections of the ipsilesional primary motor cortex-ventral premotor cortex were positively associated with motor recovery as the patients were more severely impaired. Conclusions: Differences in plasticity and recovery-related indicators of motor networks were noted according to impairment severity. Our results may suggest meaningful implications for recovery prediction and treatment strategies in future stroke research.