Modelling information flow along the human connectome using maximum flow.

Area of Science:

• Neuroscience
• Network Science
• Computational Biology

Background:

• The human connectome is a complex information-processing network.
• Traditional graph theory assumes information flow along shortest paths, which may not reflect real-world network dynamics.
• Real-world networks, including biological ones, utilize non-shortest paths for information transmission.

Purpose of the Study:

• To present a novel framework using maximum flow to quantify information flow across all possible paths in the brain.
• To expand existing network topological measures by incorporating non-shortest path information flow.
• To better integrate weighted connectivity data for a more accurate representation of brain information flow.

Main Methods:

• Developed a novel framework based on the concept of maximum flow, analogous to network traffic.
• Hypothesized that connection strengths limit information flow per unit time.
• Computed maximum information flow between brain regions considering all possible paths.

Main Results:

• The maximum flow framework quantifies information flow through both shortest and non-shortest paths.
• This approach integrates weighted connectivity data, providing a more realistic model of brain network function.
• Expanded network topological measures to account for comprehensive information flow.

Conclusions:

• The maximum flow framework offers insights into how network structure shapes information flow in the brain.
• This approach provides a more accurate representation of brain network dynamics compared to traditional graph theory.
• Suggests future applications in analyzing structural and functional connectomes at the neuronal level.