Mechanisms of fMRI (dys)connectivity

Abstract

Modern neuroimaging methods have been widely used to map interareal communication in health and disease. However, many fundamental questions regarding the mechanisms governing large-scale functional connectivity in the mammalian brain remain unanswered. For one, how does activity in one region causally affect whole-brain patterns of brain activity? And what neural mechanism underlie the fMRI connectivity alterations observed in brain disorders?

To address these questions, my lab has pioneered methods to map and manipulate fMRI connectivity in the mouse brain. Using this approach, we recently found that the relationship between neural activity and large-scale fMRI coupling is non-monotonic and critically biased by local excitatory/inhibitory ratio. Electrophysiological recordings further revealed that large-scale fMRI coupling is supported by electrophysiological coherence in infraslow/slow rhythms, but not in higher frequency bands.

These findings shed light on the general principles underlying fMRI coupling in the mammalian brain, and support a simple framework whereby fMRIhyper- andhypo-connectivity observed in brain disease may counterintuitively reflectreducedandincreasedinterareal activity, respectively. Future extensions of this framework may offer opportunities to physiologically-decode fMRI dysconnectivity in human disorders.

Alessandro Gozzi, PhD, is a senior scientist and director of the functional neuroimaging laboratory at the Italian Institute of Technology, in Rovereto, Italy. The Gozzi lab focuses on the study of the functional organization of the mammalian brain at the macroscale. A major goal of Gozzi’s research is to unravel the neural basis and dynamic organization of large-scale functional connectivity as measured with fMRI, and the underpinnings of its disruption in developmental disorders like autism. The Gozzi lab addresses these questions via the combination of cutting-edge functional MRI mapping with genetic and neural manipulations in the living mouse brain. This approach has laid the foundations for a new field of research in which large-scale neuroimaging readouts are used as probe of network activity and dysfunction across species.