The circuit architecture of cortical multisensory processing: Distinct functions jointly operating within a common anatomical network

Meijer, Guido T.; Mertens, Paul E.C.; Pennartz, Cyriel M. A.; Olcese, Umberto; Lansink, Carien S.

doi:10.1016/j.pneurobio.2019.01.004

Cited by 56 publications

(53 citation statements)

References 148 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These include a hierarchical organization reflecting a well-characterized feedforward-feedback laminar hierarchy 8 , and a spectrum between unimodal regions and transmodal areas. It should be noted however that in the mouse the latter are known to exhibit a much lower degree of regional specialization than in primates, an observation that explains a categorization of latero-posterior visual and auditory territories as polymodal components of the posterior parietal cortex 55,56 . Importantly, our results also revealed that a dominant cortical gradient spatially shapes the emergence of prevailing patterns of cortical co-activation governing spontaneous fMRI dynamics, further relating the topography of the connectome with the structure and temporal evolution of spontaneous cortical activity 26 .…”

Section: Discussionmentioning

confidence: 99%

Network structure of the mouse brain connectome with voxel resolution

Coletta

Pagani

Whitesell

et al. 2020

Preprint

View full text Add to dashboard Cite

Fine-grained descriptions of brain connectivity are fundamental for understanding how neural information is processed and relayed across spatial scales. Prior investigations of the mouse brain connectome have employed discrete anatomical parcellations, limiting spatial resolution and potentially concealing network attributes critical to the organization of the mammalian connectome.Here we provide a voxel-level description of the network and hierarchical structure of the directed mouse connectome, unconstrained by regional partitioning. We show that integrative hub regions can be directionally segregated into neural sinks and sources, defining a hierarchical axis. We describe a set of structural communities that spatially reconstitute previously described fMRI networks of the mouse brain, and document that neuromodulatory nuclei are strategically wired as critical orchestrators of inter-modular and network communicability. Notably, like in primates, the directed mouse connectome is organized along two superimposed cortical gradients reflecting unimodaltransmodal functional processing and a modality-specific sensorimotor axis. These structural features can be related to patterns of intralaminar connectivity and to the spatial topography of dynamic fMRI brain states, respectively. Together, our results reveal a high-resolution structural scaffold linking mesoscale connectome topography to its macroscale functional organization, and create opportunities for identifying targets of interventions to modulate brain function in a physiologicallyaccessible species.

show abstract

Section: Discussionmentioning

confidence: 99%

Network structure of the mouse brain connectome with voxel resolution

Coletta

Pagani

Whitesell

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Consideration should be lent to how and when multiple stimuli in single modality should be grouped together as originating from a single source, or not, before pairing with another modality. As demonstrated above from neuropsychological evidences, and also from recent systems neuroscience evidences (Ghazanfar and Schroeder, 2006;Meijer et al, 2019), the human audio-visual integration system appears to operate in rather complex processing steps, in addition to the traditional thinking of hierarchical processing from single modality. Hence, there is a need for modeling these cognitive processes.…”

Section: Additional Audio-visual Effectsmentioning

confidence: 79%

Optimality and Limitations of Audio-Visual Integration for Cognitive Systems

et al. 2020

View full text Add to dashboard Cite

Multimodal integration is an important process in perceptual decision-making. In humans, this process has often been shown to be statistically optimal, or near optimal: sensory information is combined in a fashion that minimizes the average error in perceptual representation of stimuli. However, sometimes there are costs that come with the optimization, manifesting as illusory percepts. We review audiovisual facilitations and illusions that are products of multisensory integration, and the computational models that account for these phenomena. In particular, the same optimal computational model can lead to illusory percepts, and we suggest that more studies should be needed to detect and mitigate these illusions, as artifacts in artificial cognitive systems. We provide cautionary considerations when designing artificial cognitive systems with the view of avoiding such artifacts. Finally, we suggest avenues of research toward solutions to potential pitfalls in system design. We conclude that detailed understanding of multisensory integration and the mechanisms behind audiovisual illusions can benefit the design of artificial cognitive systems.

show abstract

“…The 28 somatosensory, visual, auditory, gustatory and other sensory streams, come together (integrate) and separate 29 (diverge) to be processed in different parts of the cortex. The neural basis of how these sensory streams 30 are processed and how they generate a coherent perceptual state remains elusive [55]. It is likely that this 31 state is created and regulated by multiple structures distributed throughout the cortex working together in 32 concert [13].…”

mentioning

confidence: 99%

“…The remaining four sensory sources are: visual (VISp), auditory (AUDp), gustatory (GU) and olfactory (MOB). different parts of the cortex [30] -this type of processing however cannot be modeled without further sub-120 divisions of the thalamus and other subcortical regions and without at least some cell-type specificity. Second, 121 modeling the interactions between cortical and sub-cortical regions, especially in the context of MSI, would 122 require more elaborate communication models that can capture feedback mechanisms.…”

mentioning

confidence: 99%

“…This study built a suite of tools to quantify and characterize multisensory integration from the perspective of 470 communication dynamics. We chose to analyze the mouse connectome due to the availability of a detailed 471 meso-scale anatomical map and because the problem of multisensory integration is relatively well-studied 472 in rodents [30]. In particular, we analyzed the cortical brain network from the Allen Mouse Connectivity 473 Atlas [42] and focused on "early" dynamics of sensory integration.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Multi-sensory integration in the mouse cortical connectome using a network diffusion model

Shadi

Dyer

Dovrolis

2019

Preprint

View full text Add to dashboard Cite

8Having a structural network representation of connectivity in the brain is instrumental in analyzing 9 communication dynamics and information processing in the brain. In this work, we make steps towards 10 understanding multi-sensory information flow and integration using a network diffusion approach. In par-11 ticular, we model the flow of evoked activity, initiated by stimuli at primary sensory regions, using the 12 Asynchronous Linear Threshold (ALT) diffusion model. The ALT model captures how evoked activity that 13 originates at a given region of the cortex "ripples through" other brain regions (referred to as an activation 14 cascade). By comparing the model results to functional datasets based on Voltage Sensitive Dye (VSD) 15 imaging, we find that in most cases the ALT model predicts the temporal ordering of an activation cascade 16 correctly. Our results on the Mouse Connectivity Atlas from the Allen Institute for Brain Science show 17 that a small number of brain regions are involved in many primary sensory streams -the claustrum and the 18 parietal temporal cortex being at the top of the list. This suggests that the cortex relies on an hourglass ar-19 chitecture to first integrate and compress multi-sensory information from multiple sensory regions, before 20 utilizing that lower-dimensionality representation in higher-level association regions and more complex 21 cognitive tasks. 22The final network, denoted as N c , consists of 617 directed edges between 67 nodes -each node corre-126 sponds to one of the ROIs in our model. The density of N c is 13.9%. The distribution of edge weights is 127 skewed, with 80% of the edges having a weight of less than 5 and few edges having a weight of up to 40. The 128 distribution of edge lengths is almost uniform in the 1-7mm range. The diameter of the network (maximum 129 shortest path length between any two nodes) is 7 hops, while a node is on the average about 4 hops away from 130 any other node. The average in-degree of each node is 9.2 connections (σ=3.1), while the out-degree distri-131 bution has the same mean but larger variability (σ=8.3). Additionally, the network N c is strongly clustered, 132 with an average clustering coefficient of 60% [10]. 133The ten primary sensory regions associated with the visual, auditory, gustatory, olfactory systems, as well 134 as six somatosensory regions for different body parts (see Table SI-1), have a special role in our analysis: 135 they are viewed as sources of sensory information in the cortex [4,61]. The location of these ROIs in the 136 Allen Mouse Brain Atlas is shown in Figure 2. 137 6

show abstract

The circuit architecture of cortical multisensory processing: Distinct functions jointly operating within a common anatomical network

Cited by 56 publications

References 148 publications

Network structure of the mouse brain connectome with voxel resolution

Network structure of the mouse brain connectome with voxel resolution

Optimality and Limitations of Audio-Visual Integration for Cognitive Systems

Multi-sensory integration in the mouse cortical connectome using a network diffusion model

Contact Info

Product

Resources

About