In neural networks, the interaction between network regions was described in terms of network connectivity. Connectivity measures included the correlation between signals, i.e., if one region began signal transmission and a second region did the same, these two regions were considered connected and working together. Regions could also be negatively correlated, meaning that one region increases its signalling while a second decreases, or that there is a temporal delay between changes. Dynamic functional connectivity addresses frequency specific interactions between network regions. For example, if two region simultaneously switch their signalling from low frequencies, associated with resting-state of the brain to higher frequencies, related to cognitive processing, these two regions are supposed to collaborate in a cognitive task. This talk will introduce the concept of frequency-specific dynamic functional connectivity at the example of normal and pathological attention processing.