Non-invasive Brain-Computer Interfaces (BCIs) can exploit the ability of subjects to voluntary modulate their brain activity through mental imagery. Despite its clinical applications, controlling a BCI appears to be a learned skill that requires several weeks to reach relatively high-performance in control, without being sufficient for 15 to 30 % of the users [1]. This gap has motivated a deeper understanding of mechanisms associated with motor imagery (MI) tasks. Here, we investigated dynamical changes in multimodal network recruitment. We hypothesized that integrating information from EEG and MEG data, show a better description of the core-periphery changes occurring during a motor imagery-based BCI training. Such an enriched description could reveal fresh insights into learning processes that are difficult to observe at the signle layer level and eventually improve the prediction of future BCI performance.multimodal brain network properties could be considered as a potential marker of BCI learning.