Motifs analysis-based indices to discover brain network architecture

Motifs are small recurring meso-scale structures characterizing real networks. Studies in literature investigated network motifs focusing only on their frequency spectrum. In the present work, we propose to study functional brain networks by means of motifs analysis, going beyond the identification of brain network-characterizing motifs. Indeed, we defined and implemented new indices based on motifs analysis and we applied this approach to better understand the architecture of brain networks after stroke. In particular, 3-node motifs analysis was performed on resting state connectivity patterns estimated starting from EEG signals of 45 subacute unilateral stroke patients. The new indices were extracted from these networks focusing on intra- and inter-hemispheric connectivity and revealed that: i) after stroke the communication in the affected hemisphere is better characterized by complex “building blocks” with respect to the unaffected hemisphere; ii) a positive correlation exists between specific inter-hemispheric connectivity patterns at rest before a rehabilitative intervention and the clinical outcome in theta frequency band.