Monolingual and bilingual language networks in healthy subjects using functional MRI and graph theory

We wish to model brain networks by analyzing tasked-based fMRI data of healthy monolingual and bilingual subjects who performed a specific language task designed for clinical applications. We aim to study the architecture of the functional networks in each group and characterize any differences arising from network centrality measurements. Particularly, we sought to assess the k core, which is emerging as an important topological measure of networks and may provide useful insights in addition to the functional connectivity map. We analyzed fMRI scans from 8 healthy bilinguals and 8 monolinguals. For every bilingual subject, two scans were acquired - Spanish (L1) and in English (L2). We unveiled a persistent functional architecture, the “common network”, beyond inter-subject variability, which wires together with the left Broca’s area, the left Wernicke's area (WA), the left PreMotor area, and the pre-Supplementary Motor Area. This structure displays the differential connectivity of WA between groups. The k core centrality measure showed several areas belong to the maximum k core while WA’ shell occupancy varies across groups.