Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Meta-Analysis
. 2013 Dec;34(12):3247-66.
doi: 10.1002/hbm.22138. Epub 2012 Jul 17.

An Investigation of the Structural, Connectional, and Functional Subspecialization in the Human Amygdala

Affiliations
Free PMC article
Meta-Analysis

An Investigation of the Structural, Connectional, and Functional Subspecialization in the Human Amygdala

Danilo Bzdok et al. Hum Brain Mapp. .
Free PMC article

Abstract

Although the amygdala complex is a brain area critical for human behavior, knowledge of its subspecialization is primarily derived from experiments in animals. We here employed methods for large-scale data mining to perform a connectivity-derived parcellation of the human amygdala based on whole-brain coactivation patterns computed for each seed voxel. Voxels within the histologically defined human amygdala were clustered into distinct groups based on their brain-wide coactivation maps. Using this approach, connectivity-based parcellation divided the amygdala into three distinct clusters that are highly consistent with earlier microstructural distinctions. Meta-analytic connectivity modelling then revealed the derived clusters' brain-wide connectivity patterns, while meta-data profiling allowed their functional characterization. These analyses revealed that the amygdala's laterobasal nuclei group was associated with coordinating high-level sensory input, whereas its centromedial nuclei group was linked to mediating attentional, vegetative, and motor responses. The often-neglected superficial nuclei group emerged as particularly sensitive to olfactory and probably social information processing. The results of this model-free approach support the concordance of structural, connectional, and functional organization in the human amygdala and point to the importance of acknowledging the heterogeneity of this region in neuroimaging research.

Keywords: amygdala; behavior; connectivity; data mining; parcellation; social cognition.

Figures

Figure 1
Figure 1
Location of the volume of interest. The seed region was based on probabilistic maps of the amygdala obtained from the Jülich histological atlas [Amunts et al., 2005]. The left image was rendered using Mango (multi-image analysis GUI; http://ric.uthscsa.edu/mango/). Renderings in right column depict coronal, sagittal, and axial sections through the seed region rendered into a T1 weighted MNI single subject template.
Figure 2
Figure 2
Original, reordered, and clustered connectivity cross-correlation matrices. Cross-correlations between coactivation patterns of individual seed voxels (563 in left amygdala, 569 in the right amygdala). Three sets of voxels that feature similar brain-wide coactivation profiles emerged in the reordered cross-correlation matrix and were subsequently grouped by hierarchical cluster analysis.
Figure 3
Figure 3
Renderings of cytoarchitectonic versus connectivity-derived parcellation. Connectivity-derived clusters on the right show spatial continuity and localization in accordance with microscopically observed clusters on the left [Amunts et al., 2005]. Blue = corresponds to laterobasal nuclei group, red = corresponds to centromedial nuclei group, green = corresponds to superficial nuclei group. Images were rendered using Mango.
Figure 4
Figure 4
Difference analyses between coactivation maps of the derived clusters. Difference analyses between the coactivation maps of any combination of two clusters, separately in each hemisphere, to quantify how individual clusters’ coactivation maps relate to each other. For each pair of the three delineated clusters (one pair per column), we performed the following procedure. First, each cluster’s connectivity map was computed using MACM (cf., Supporting Information Fig. 1). Second, the difference analysis on each pair of connectivity maps yielded voxels that were significantly more likely coactivated with either cluster (cf. Method section). Column-wise, the clusters and corresponding coactivations are color coded. Rendered on T1 MNI single subject template. le = left, ri = right, LB = laterobasal nuclei group, CM = centromedial nuclei group, SF = superficial nuclei group.
Figure 5
Figure 5
Connectivity patterns specific to individual derived clusters in the left and right amygdala. Coactivation patterns of individual clusters were initially revealed by meta-analytic connectivity modeling (MACM; Supporting Information Fig. S1). Functional connectivity specific to a given clusters was then determined by computing the AND conjunction between the two difference analyses of coactivation maps with the respective two other clusters. The extent threshold was set to k = 10 voxels. Coordinates in MNI space. leLB: The left cluster identified as laterobasal nuclei group was selectively coactivated with the precuneus, inferior occipital gyrus, cerebellum, superior temporal gyrus/associative auditory cortex, and middle frontal gyrus/frontal eye field on the left side, as well as bilateral temporal pole and right inferior parietal cortex. riLB: The right cluster identified as laterobasal nuclei group was selectively coactivated with the dorsomedial prefrontal cortex, temporal pole, precuneus, and inferior parietal cortex bilaterally, as well as the ventromedial prefrontal cortex, superior temporal gyrus/associative auditory cortex, middle frontal gyrus/frontal eye field, hippocampus, and posterior superior temporal sulcus on the left side. leCM: The left cluster identified as centromedial nuclei group was connected to the supplementary motor cortex (BA 6), pallidum, putamen, cerebellum, insula, and thalamus bilaterally (activity in right pallidum/putamen and bilateral cerebellum not depicted), as well as left posterior mid-cingulate cortex, left primary somatosensory cortex (BA 3), and right occipital lobe (V5). riCM: The right cluster identified as centromedial nuclei group was connected to the primary motor cortex (BA 4), supplementary motor cortex (BA 6), pallidum, putamen, primary somatosensory cortex (BA3), and inferior frontal gyrus bilaterally, as well as the right thalamus and left insula. leSF: The left cluster identified as superficial nuclei group was selectively coactivated with the ventral striatum/nucleus accumbens and olfactory tubercle bilaterally, as well as with the right anterior insula. riSF: The right cluster identified as superficial nuclei group was selectively connected to the left anterior mid-cingulate cortex and bilateral anterior insula, extending into the inferior frontal gyrus.
Figure 6
Figure 6
Comparing different approaches to parcellating the amygdala. The left column shows coronal slices depicting the amygdala parcellation based on the maximum probability map (MPM) from the Jülich atlas using BrainMap experiments activating closest to given seed voxels without further constraints (cf. Method section). The middle column shows slices depicting a supplementary analysis based on the Jülich amygdala MPM, in which only Brain-Map experiments that activated within the histologically identified anygdala were considered for the computation of the voxel-wise coactivation patterns. The observation that 92% of all seed-voxels were assigned to corresponding clusters strongly argues against the possibility that clustering was driven by surrounding nonamygdalar activation. The right column illustrates a second supplementary analysis, in which the amygdala parcellation was based on the spatially more confined MPM from the macroanatomical Harvard–Oxford atlas. For this analysis, we again constrained the computation of coactivation patterns to BrainMap experiments that activated within the amygdala to exclude influences of adjacent regions on the coactivation patterns and hence the connectivity-based parcellation. The two confirmatory analyses provide converging evidence and hence support for the robustness of the original analysis by revealing a very similar parcellation scheme based on a more conservative CBP implementation and in an independent seed region. Blue = laterobasal nuclei group, red = centromedial nuclei group, green = superficial nuclei group.
Figure 7
Figure 7
Meta-data profiling of the derived clusters. Functional characterization by behavioral domain (BD) and paradigm class (PC) meta-data. The colored bars denote the number of foci for the particular BD/PC within the respective derived cluster. The gray bars represent the number of foci that would be expected to hit the particular cluster if all foci with the respective BD/PC were randomly distributed throughout gray-matter. That is, the gray bars indicate the by-chance frequency of that particular label given the size of the clusters. LB = laterobasal nuclei group, CM = centromedial nuclei group, SF = superficial nuclei group. Asterisk indicates significant overrepresentation after Bonferroni correction for multiple comparisons.

Similar articles

See all similar articles

Cited by 107 articles

See all "Cited by" articles

Publication types

Feedback