Markov Processes and Brain Network Hubs

TL;DR

This paper models brain neural networks as directed graphs with Markov processes to identify hubs and derive bounds on graph diameter, integrating neurophysiology with graph theory.

Contribution

It introduces a novel Markov process framework for brain networks and establishes a new upper bound for graph diameter based on eigenvalues.

Findings

Hubs are identified as nodes with highest stationary probabilities.

Derived a new upper bound for graph diameter using Markov matrix eigenvalues.

Provides a mathematical link between neurophysiological data and graph theoretical analysis.

Abstract

Current concepts of neural networks have emerged over two centuries of progress beginning with the neural doctrine to the idea of neural cell assemblies. Presently the model of neural networks involves distributed neural circuits of nodes, hubs, and connections that are dynamic in different states of brain function. Advances in neurophysiology, neuroimaging and the field of connectomics have given impetus to the application of mathematical concepts of graph theory. Current approaches do carry limitations and inconsistency in results achieved. We model the neural network of the brain as a directed graph and attach a matrix (called the Markov matrix) of transition probabilities (determined by the synaptic strengths) to every pair of distinct nodes giving rise to a (continuous) Markov process. We postulate that the network hubs are the nodes with the highest probabilities given by the stationary distribution of Markov theory. We also derive a new upper bound for the diameter of a graph in terms of the eigenvalues of the Markov matrix.