In my current Marie Sklodowska-Curie Fellowship, I aim to map the interface between the semantic representational and control networks by estimating the dynamical causal networks (DCNs) underlying on-going resting-state fMRI. This also includes determining the structural (based on DW-MRI) and functional connectivity information to use them as priors to improve specificity and accuracy of DCNs discovery. This will be instrumental in the development of imaging-based diagnostic classification and predictive models (e.g., for patients with post-stroke aphasia). Specifically, the reorganization of DCNs underpins the early stages of numerous pathological conditions, even before any observable change in behaviour emerges. Therefore, predictive models such as those assessing status after a stroke (where the last symptoms appear after months) can help improve effective early rehabilitation and intervention strategies.