Graph-theory based analyses of human motor learning: A resting-state fMRI study

Resting-state functional magnetic resonance imaging (rs-fMRI) is a promising method to investigate brain functional connectivity (FC) changes resulted from interventional studies. Graph-theory based analyses quantity multiple aspects of the brain network by computing different metrics. We used a graph-theory approach to investigate neural plasticity of the human brain due to a novel motor learning protocol that has not been addressed before using rs-fMRI. Thirteen young and healthy participants practiced seven sessions of motor learning exercises. rs-fMRI data were collected before the first and after the 7th session of exercises. We performed ROI-to-ROI analyses along with the computation of different graph metrics on rs-fMRI data after preprocessing steps. The graph networks were constructed using correlation signals originating from 164 parcellated regions (Harvard-Oxford atlas) of the brain. In further analyses, different metrics of the constructed graph were computed and compared between pre- and post-training sessions using proper statistical tests. Thirteen significant regions of interest (Left Dorsal Attention.IPS, Left Angular Gyrus, and Right Postcentral Gyrus, to name a few) were obtained with p-FDR <; 0.05, which had essential roles in the motor learning procedure.