Activity in the human brain displays features of localization, allowing for multiple tasks at the same time, and slow relaxation, which means that in localized regions activity is sustained for long times. This phenomenology can be explained starting from the structure of brain connectivity patterns, which is that of hierarchical modular network.
In this seminar I will address the study of these two features, localization and anomalous dynamics, using minimal computational models for diffusion and activity propagation, highlighting their emergence in low-dimensional networks. Since several brain pathologies are associated with a dimension reduction (a breaking process), our results may serve as the foundation for topology-based diagnostic tools.