Combined MEG and fMRI Exponential Random Graph Modeling for inferring functional Brain Connectivity

Abstract

Estimated connectomes by the means of neuroimaging techniques have enriched our knowledge of the organizational properties of the brain leading to the development of network-based clinical diagnostics. Unfortunately, to date, many of those network-based clinical diagnostics tools, based on the mere description of isolated instances of observed connectomes are noisy estimates of the true connectivity network. Modeling brain connectivity networks is therefore important to better explain the functional organization of the brain and allow inference of specific brain properties. In this report, we present pilot results on the modeling of combined MEG and fMRI neuroimaging data acquired during an n-back memory task experiment. We adopted a pooled Exponential Random Graph Model (ERGM) as a network statistical model to capture the underlying process in functional brain networks of 9 subjects MEG and fMRI data out of 32 during a 0-back vs 2-back memory task experiment. Our results suggested strong evidence that all the functional connectomes of the 9 subjects have small world properties. A group level comparison using comparing the conditions pairwise showed no significant difference in the functional connectomes across the subjects. Our pooled ERGMs successfully reproduced important brain properties such as functional segregation and functional integration. However, the ERGMs reproducing the functional segregation of the brain networks discriminated between the 0-back and 2-back conditions while the models reproducing both properties failed to successfully discriminate between both conditions. Our results are promising and would improve in robustness with a larger sample size. Nevertheless, our pilot results tend to support previous findings that functional segregation and integration are sufficient to statistically reproduce the main properties of brain network.