Generative Modeling of Hidden Functional Brain Networks

TL;DR

This paper presents a Hidden Markov Random Field framework to infer and analyze the latent functional brain networks from fMRI data, aiming to understand the underlying computational principles of observed network features.

Contribution

It introduces a novel probabilistic model for estimating hidden neuronal functional relationships from fMRI data, addressing a gap in understanding brain network structure.

Findings

Effective inference of latent brain networks from fMRI data.

Insights into the computational principles underlying network features.

Potential for improved understanding of brain connectivity patterns.

Abstract

Functional connectivity refers to the temporal statistical relationship between spatially distinct brain regions and is usually inferred from the time series coherence/correlation in brain activity between regions of interest. In human functional brain networks, the network structure is often inferred from functional magnetic resonance imaging (fMRI) blood oxygen level dependent (BOLD) signal. Since the BOLD signal is a proxy for neuronal activity, it is of interest to learn the latent functional network structure. Additionally, despite a core set of observations about functional networks such as small-worldness, modularity, exponentially truncated degree distributions, and presence of various types of hubs, very little is known about the computational principles which can give rise to these observations. This paper introduces a Hidden Markov Random Field framework for the purpose of representing, estimating, and evaluating latent neuronal functional relationships between different brain regions using fMRI data.