Distinct structure-function relationships across cortical regions and connectivity scales in the rat brain

Abstract

An improved understanding of the structure-function relationship in the brain is necessary to know to what degree structural connectivity underpins abnormal functional connectivity seen in many disorders. We integrated high-field resting-state fMRI-based functional connectivity with highresolution macro-scale diffusion-based and meso-scale neuronal tracer-based structural connectivity, to obtain an accurate depiction of the structure-function relationship in the rat brain. Our main goal was to identify to what extent structural and functional connectivity strengths are correlated, macroand meso-scopically, across the cortex. Correlation analyses revealed a positive correspondence between functional connectivity and macro-scale diffusion-based structural connectivity, but no correspondence between functional connectivity and meso-scale neuronal tracer-based structural connectivity. Locally, strong functional connectivity was found in two well-known resting-state networks: the sensorimotor and default mode network. Strong functional connectivity within these networks coincided with strong short-range intrahemispheric structural connectivity, but with weak heterotopic interhemispheric and long-range intrahemispheric structural connectivity. Our study indicates the importance of combining measures of connectivity at distinct hierarchical levels to accurately determine connectivity across networks in the healthy and diseased brain. Distinct structure-function relationships across the brain can explain the organization of networks and may underlie variations in the impact of structural damage on functional networks and behavior. The brain is a complex organ that can be regarded as a structural and functional network of interacting regions at the micro-, meso-and macroscopic level. At the macro-scale, whole-brain functional networks can be non-invasively mapped with resting-state functional MRI (resting-state fMRI). In resting-state fMRI data the inter-regional temporal correlations of spontaneous lowfrequency blood oxygenation level-dependent (BOLD) fluctuations reflect functional connectivity (Biswal et al. 1995; Fox and Raichle 2007). Based on clusters of functionally connected regions, various resting-state networks have been identified, such as the default mode network (Greicius et al. 2003). These networks have been related to behavioral functioning in health and disease, and abnormalities partially explain pathophysiological processes and disease progression (van den Heuvel and Hulshoff Pol 2010; Zhang and Raichle 2010; Raichle 2015). The exact nature of functional connectivity is nonetheless not yet fully established. Since functional connectivity measured with resting-state fMRI relies on synchronous BOLD signals, understanding functional connectivity starts with understanding the origin of BOLD signals. The BOLD signal captures hemodynamic changes, such as blood flow, in response to neural activity.Although it is clear that BOLD signals reflect aspects of neural responses, it is still unclear which processes a...

Whole-Brain Causal Discovery Using fMRI

“…When applying CaLLTiF to resting-state HCP data, we also found a significant effect of Euclidean distance on causal connections, both in terms of their strength and their end-point degree similarity (Figure 7). This finding is in line with several prior studies that have shown distance-dependence in structural and functional connections [91][92][93], and provides new evidence for the presence of a similar distance dependence in causal connections. Since causal connections are more mechanistically meaningful than functional connections, our findings also open the door to potentially more accurate investigations of how variations in distance-dependent patterns of connectivity can relate to different states of health and disease [91].…”

Networked Federated Multi-Task Learning

“…Many important application domains generate distributed collections of local datasets that are related via an intrinsic network structure [1]. Such a network structure can arise from spatio-temporal proximity, statistical dependencies or functional relations [2,3,4]. The network structure of data might also arise from a distributed computing infrastructure that generates the data [5].…”

MRI-based Measurement of Structural and Functional Connectivity in Health and Disease

“…However, previous research has shown a clear relationship between different levels of structural connectivity (i.e. from neuronal tracer and diffusion-based tractography studies) and functional connectivity (Straathof et al, 2020), which we have recently described in a review (Straathof et al, 2019). How structural and functional connectivity exactly relate across different levels of investigation (i.e.…”