Poster No:
1752
Submission Type:
Abstract Submission
Authors:
Yezhou Wang1, Jordan DeKraker2, Raúl Rodriguez-Cruces2, Donna Gift Cabalo1, Bin Wan3, Casey Paquola4, Sofie Valk3, Seok Jun Hong5, Alan Evans6, Boris Bernhardt2
Institutions:
1McGill University, Montreal, Quebec, 2Montreal Neurological Institute and Hospital, Montreal, Quebec, 3Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany, 4INM-7, Jülich, Jülich, 5Center for Neuroscience Imaging Research, Institute for Basic Science, Suwon, Gyeonggi-do, 6McGill Centre for Integrative Neuroscience (MCIN), Montreal, Quebec
First Author:
Co-Author(s):
Jordan DeKraker
Montreal Neurological Institute and Hospital
Montreal, Quebec
Bin Wan
Max Planck Institute for Human Cognitive and Brain Sciences
Leipzig, Germany
Sofie Valk
Max Planck Institute for Human Cognitive and Brain Sciences
Leipzig, Germany
Seok Jun Hong
Center for Neuroscience Imaging Research, Institute for Basic Science
Suwon, Gyeonggi-do
Alan Evans
McGill Centre for Integrative Neuroscience (MCIN)
Montreal, Quebec
Boris Bernhardt
Montreal Neurological Institute and Hospital
Montreal, Quebec
Introduction:
Understanding the functional processing hierarchy of the human cerebral cortex is a longstanding challenge in neuroscience. Foundational hierarchical models of the cortex, formulated in non-human primates, presume the interaction of feedforward and feedback processes [1-3]. There is increasing support for a dual stream architecture, in which the laminar origin of projection patterns determines feedback vs feedforward signaling, especially for long-range connectivity patterns. Here, we expand this research to the human cortex, capitalizing on ultra-high-resolution 7 Tesla magnetic resonance imaging (7T MRI) [4]. Combining directional functional connectivity modelling with geodesic distance mapping across cortex, we specifically tested whether feedback signalling to primary sensory systems derives from deeper layers with increasing anatomical distance between different cortical areas.
Methods:
We analyzed 7T MRI data of 10 unrelated healthy adults (age: 26.80±4.61 years, 5 females). Each participant underwent three separate sessions, during which structural (resolution 0.5mm) and resting-state functional MRI (rs-fMRI; 1.9mm) scans were acquired. MRI data were processed using micapipe [5]. We constructed 14 equivolumetric surfaces between the pial and white matter boundaries based on structural MRI and registered them to the rs-fMRI data to sample deep- and superficial-layer timeseries. Timeseries from the three sessions were concatenated and mapped to a parcellation [6]. Regression dynamic causal models (rDCM) [7] were applied to these timeseries, resulting in effective connectivity matrices. To evaluate patterns of feedback streams, we defined the deep source ratio (DSR). For area V1, we assessed DSR of the feedback stream from regions in visual network (VN), dorsal attention network (dAN), and default mode network (DMN). For each region, we calculated Spearman's correlation coefficients between its DSR and its geodesic distance to the regions of interest (ROI). We performed similar analyses for primary somatosensory cortex (S1) within the somatomotor network (SMN) and for A1 within the auditory network (AN). We integrated all ROIs together and examined associations by controlling inter-ROI differences using a mixed linear model. To further enhance the precision of defining superficial and deep layers, we reconstructed the effective connectivity matrix by incorporating cortical layer information from a histological dataset (Figure 2.A) [8]. Subsequently, we reran all main analyses based on the updated effective connectivity matrix. Spatial autocorrelation permutation tests were conducted for all correlations.
Results:
We derived effective connectivity matrices for each participant (Figure 1.A). Individual matrices were averaged to obtain the group-level matrix for subsequent analyses. Our initial investigation focused on V1, revealing a significant association between DSR and geodesic distance, especially in VN+dAN+DMN (r=0.45, p=0.002; Figure 1.B). Similar findings were observed for S1 in SMN+dAN+DMN (r=0.37, p=0.042; Figure 1.C) and for A1 in AN (r=0.60, p=0.034; Figure 1.D). Combining all regions using a mixed model revealed significant associations (r=0.49, p<0.001; Figure 1.E). Finally, we cross-validated findings after adjusting definitions of deep and superficial surfaces by incorporating cortical layer information from a histological dataset [8] (Figure 2.A). We repeated the main analyses and obtained similar results in VN (r=0.43, p=0.002), SMN (r=0.43, p=0.016), AN (r=0.62, p=0.027), and all regions together (r=0.46, p<0.001; Figure 2.B).
Conclusions:
Our findings reveal a clear hierarchical organization within the human cortex, particularly in the feedback stream from unimodal and heteromodal association cortex to V1, S1, and A1. These results indicate the presence of well-defined, distance-dependent feedback pathways in both the superficial and deep layers, consistent with the notion of feedback connections providing a generative network [9].
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural) 2
fMRI Connectivity and Network Modeling 1
Methods Development
Keywords:
Computational Neuroscience
Cortex
FUNCTIONAL MRI
MRI
STRUCTURAL MRI
1|2Indicates the priority used for review
Provide references using author date format
1. Felleman DJ, Van Essen DC. (1991), Distributed Hierarchical Processing in the Primate Cerebral Cortex. Cereb Cortex. 1(1):1-47. doi: 10.1093/cercor/1.1.1-a.
2. Mesulam MM. (1998), From sensation to cognition. Brain. 121(6):1013-52. doi: 10.1093/brain/121.6.1013.
3. Pascal B, Alexandre B, Kenneth K, Henry K. (2000), Laminar Distribution of Neurons in Extrastriate Areas Projecting to Visual Areas V1 and V4 Correlates with the Hierarchical Rank and Indicates the Operation of a Distance Rule. J Neurosci. 20(9):3263. doi: 10.1523/JNEUROSCI.20-09-03263.2000.
4. van der Kolk AG, Hendrikse J, Zwanenburg JJM, Visser F, Luijten PR. (2013), Clinical applications of 7T MRI in the brain. Eur J Radiol. 82(5):708-18. doi: https://doi.org/10.1016/j.ejrad.2011.07.007.
5. Cruces RR, Royer J, Herholz P, Larivière S, Vos de Wael R, Paquola C, et al. (2022), Micapipe: A pipeline for multimodal neuroimaging and connectome analysis. Neuroimage. 263:119612. doi: https://doi.org/10.1016/j.neuroimage.2022.119612.
6. Glasser MF, Coalson TS, Robinson EC, Hacker CD, Harwell J, Yacoub E, et al. (2016), A multi-modal parcellation of human cerebral cortex. Nature. 536(7615):171-8. doi: 10.1038/nature18933.
7. Frässle S, Lomakina EI, Razi A, Friston KJ, Buhmann JM, Stephan KE. (2017), Regression DCM for fMRI. Neuroimage. 155:406-21. doi: https://doi.org/10.1016/j.neuroimage.2017.02.090.
8. Wagstyl K, Larocque S, Cucurull G, Lepage C, Cohen JP, Bludau S, et al. BigBrain 3D atlas of cortical layers: (2020),Cortical and laminar thickness gradients diverge in sensory and motor cortices. PLoS Biol. 18(4):e3000678. doi: 10.1371/journal.pbio.3000678.
9. Vezoli J, Magrou L, Goebel R, Wang X-J, Knoblauch K, Vinck M, et al. (2021), Cortical hierarchy, dual counterstream architecture and the importance of top-down generative networks. Neuroimage. 225:117479. doi: https://doi.org/10.1016/j.neuroimage.2020.117479.