Negative functional brain networks

Abstract

The anticorrelations in fMRI measurements are still not well characterized, but some new evidences point to a possible physiological role. We explored the topology of functional brain networks characterized by negative edgess and their possible alterations in schizophrenia, using functional images of 8 healthy subjects and 8 schizophrenic patients in a resting state condition. In order to minimize the insertion of artifactual negative correlations, the preprocessing of images was carried out by the CompCorr procedure, and the results compared with the Global Signal Regression (GSR) procedure. The degree distribution, the centrality, the efficiency and the rich-club behavior were used to characterize the functional brain network with negative links of healthy controls in comparison with schizophrenic patients. The results show that functional brain networks with both positive and negative values have a truncated power-law degree distribution. Moreover, although functional brain networks characterized by negative values have not small-world topology, they show a specific disassortative configuration: the more connected nodes tend to have fewer connections between them. This feature is lost using the GSR procedure. Finally, the comparison with schizophrenic patients showed a decreased (local and global) efficiency associated to a decreased connectivity among central nodes. As a conclusion, functional brain networks characterized by negative values, despite lacking a well defined topology, show specific features, different from random, and indicate an implication in the alterations associated to schizophrenia.

Dedifferentiation Does Not Account for Hyperconnectivity after Traumatic Brain Injury

“…Finally, given that positive and negative connectivity may represent distinct neural processes (13, 14), to examine dedifferentiation, we separate positive and negative connectivity effects (i.e., positive and negative correlations). This was first observed in a small sample of subjects during the first 6 months post injury (15) and later cross-sectional work demonstrated similar effects.…”

“…Second, we examine connectivity changes both within and between networks (e.g., DMN and the task-positive network). Finally, given that dedifferentiation has implications for how networks relate to one another, we separate positive and negative connectivity effects (i.e., positive and negative correlations), as positive and negative networks have been shown to have different network characteristics from one another (14). …”

Graph theory approaches to functional network organization in brain disorders: A critique for a brave new small-world

“…Given that negative correlations are observable in the absence of GSR, however, others have examined whether negative weights contribute differentially to information processing within the network (Parente et al, 2017). Negative correlations may also reflect NMDA action in cortical inhibition (Anticevic et al, 2012).…”

“…Negative correlations may also reflect NMDA action in cortical inhibition (Anticevic et al, 2012). These connections bear consideration given that brain networks composed of only negative connections do not retain a small-world topology, but do have properties distinct from random networks (Parente et al, 2017;Schwarz & McGonigle, 2011). Altogether, the omission of methodological details about negative FC in empirical reports severely hampers the resolution of this important choice point in defining graphs.…”

Determining the Hierarchical Architecture of the Human Brain Using Subject-Level Clustering of Functional Networks

“…Further, the choice of the threshold is also challenging and can give rise to heterogeneity in the findings (Jalili, 2016;Garrison et al, 2015). Like weak edges, negative edges (i.e., anticorrelations) have been also shown to have a substantial physiological basis (Nielsen et al, 2018;Fox et al, 2005;Zhan et al, 2017;Parente et al, 2017;Kelly et al, 2008).…”