# Network neuroscience - Author(s): Danielle S. Bassett & Olaf Sporns - Date: 2017 - Publication: Nature Neuroscience - [Link](https://www.nature.com/articles/nn.4502) --- ## Summary This is a review paper which attempts to discuss the fundamentals of network neuroscience and outline new tools that will be useful for addressing long-standing complex problems and research frontiers. ### Network mapping Brain imaging and EEG measurements are utilized in various ways to construct brain networks that are — functional, correlational, structural, or causal — and can be used to study the interactions of different brain regions at different scales. Highlight a move from understanding individual brain regions as having one behavior (univariate) to a multi-variate perspective in which behavior is enabled by the connections of different regions. Generally, they discuss that network mapping takes place at different levels. For example, at the molecular level (systems biology) all the way up to the social level (computational social science). ### Network analysis They highlight that their has been a converging understanding that the brain represents a small-world-like network — that is, it has a high clustering coefficient (lots of subregions where most nodes/neurons are highly connected) but also a relatively small number of random-like connections (neurons) to different areas in the brain. This construction allows for the efficient specialization of specific regions for specific tasks, while also allowing efficient communication between different regions. They emphasize a movement towards understanding higher-order brain network dynamics in the future. ## Frontiers #### Network Dynamics - Dynamics **of** networks -> relates to dynamics on **fixed networks** - Dynamics **on** networks -> relates to dynamics on networks that can change — whether that is edges can rewire or nodes are not permanent within the network They also highlight that **these areas will likely merge** in the future and that there is a lot of potential for **multi-layer networks** to capture the different network scales discussed above. #### Prediction & Control The general idea here is that a good understanding of network dynamics should allow for prediction, which should then allow for control. They posit that control is the highest form of understanding and if this is achieved then we can begin to intervene positively on the most problematic brain disorders. --- #### Related [[neuroscience]]