Network communication in the brain

Brain networks can be inferred using a process called tractography and diffusion MRI scans. The leftmost image shows two candidate paths in the brain network.

Healthy brain activity relies on the constant exchange of electrical signals between grey matter regions. The traffic of information across the brain allows distant areas to work together to create the neural dynamics that underpin perception, behaviour, and cognition.

Cutting-edge brain imaging techniques can be used to map the human connectome – the complex network of nerve fibres that interlinks all brain regions and supports their communication. It is easy to see that, for any pair of communicating regions, there exist an astounding number of paths through which signalling can take place. What propagation and routing strategies are used to navigate the brain’s complex wiring and establish communication between regions?

To answer this question, we develop models of network communication that approximate biological neural signalling. To do this, we combine innovative methods from computer science, physics and statistics with multimodal neuroimaging datasets comprising thousands of participants. Our research has revealed new fundamental insight into how the brain’s anatomical wiring shapes functional interactions between regions. Applications of these insights include mapping disrupted neural communication in the aftermath of stroke, machine-learning predictions of human behaviour from imaging data, and the refinement of brain stimulation protocols used for treatment of clinical conditions.

Further research and key questions
  • Develop computational models of network communication that accurately approximate patterns of neural signalling
  • Understanding disrupted neural communication in disease states
  • Investigate how external stimuli are propagated through the brain, with emphasis on clinical brain stimulation
Project leaders

Caio Seguin & Andrew Zalesky

Further reading: General interest pieces
Further reading: Journal articles

Back to other projects