Automated claustrum segmentation in human brain MRI using deep learning.

In the last two decades, neuroscience has produced intriguing evidence for a central role of the claustrum in mammalian forebrain structure and function. However, relatively few in vivo studies of the claustrum exist in humans. A reason for this may be the delicate and sheet-like structure of the claustrum lying between the insular cortex and the putamen, which makes it not amenable to conventional segmentation methods. Recently, Deep Learning (DL) based approaches have been successfully introduced for automated segmentation of complex, subcortical brain structures. In the following, we present a multi-view DL-based approach to segment the claustrum in T1-weighted MRI scans. We trained and evaluated the proposed method in 181 individuals, using bilateral manual claustrum annotations by an expert neuroradiologist as reference standard. Cross-validation experiments yielded median volumetric similarity, robust Hausdorff distance, and Dice score of 93.3%, 1.41 mm, and 71.8%, respectively, representing equal or superior segmentation performance compared to human intra-rater reliability. The leave-one-scanner-out evaluation showed good transferability of the algorithm to images from unseen scanners at slightly inferior performance. Furthermore, we found that DL-based claustrum segmentation benefits from multi-view information and requires a sample size of around 75 MRI scans in the training set. We conclude that the developed algorithm allows for robust automated claustrum segmentation and thus yields considerable potential for facilitating MRI-based research of the human claustrum. The software and models of our method are made publicly available.

Spatially patterned excitatory neuron subtypes and projections of the claustrum.

The claustrum is a functionally and structurally complex brain region, whose very spatial extent remains debated. Histochemical-based approaches typically treat the claustrum as a relatively narrow anatomical region that primarily projects to the neocortex, whereas circuit-based approaches can suggest a broader claustrum region containing projections to the neocortex and other regions. Here, in the mouse, we took a bottom-up and cell-type-specific approach to complement and possibly unite these seemingly disparate conclusions. Using single-cell RNA-sequencing, we found that the claustrum comprises two excitatory neuron subtypes that are differentiable from the surrounding cortex. Multicolor retrograde tracing in conjunction with 12-channel multiplexed in situ hybridization revealed a core-shell spatial arrangement of these subtypes, as well as differential downstream targets. Thus, the claustrum comprises excitatory neuron subtypes with distinct molecular and projection properties, whose spatial patterns reflect the narrower and broader claustral extents debated in previous research. This subtype-specific heterogeneity likely shapes the functional complexity of the claustrum.

The claustrum of the sheep and its connections to the visual cortex.

The present study analyses the organization and selected neurochemical features of the claustrum and visual cortex of the sheep, based on the patterns of calcium-binding proteins expression. Connections of the claustrum with the visual cortex have been studied by tractography. Parvalbumin-immunoreactive (PV-ir) and Calbindin-immunoreactive (CB-ir) cell bodies increased along the rostro-caudal axis of the nucleus. Calretinin (CR)-labeled somata were few and evenly distributed along the rostro-caudal axis. PV and CB distribution in the visual cortex was characterized by larger round and multipolar cells for PV, and more bitufted neurons for CB. The staining pattern for PV was the opposite of that of CR, which showed densely stained but rare cell bodies. Tractography shows the existence of connections with the caudal visual cortex. However, we detected no contralateral projection in the visuo-claustral interconnections. Since sheep and goats have laterally placed eyes and a limited binocular vision, the absence of contralateral projections could be of prime importance if confirmed by other studies, to rule out the role of the claustrum in stereopsis.

Topographic gradients define the projection patterns of the claustrum core and shell in mice.

The claustrum is densely connected to the cortex and participates in brain functions such as attention and sleep. Although some studies have reported the widely divergent organization of claustrum projections, others describe parallel claustrocortical connections to different cortical regions. Therefore, the details underlying how claustrum neurons broadcast information to cortical networks remain incompletely understood. Using multicolor retrograde tracing we determined the density, topography, and co-projection pattern of 14 claustrocortical pathways, in mice. We spatially registered these pathways to a common coordinate space and found that the claustrocortical system is topographically organized as a series of overlapping spatial modules, continuously distributed across the dorsoventral claustrum axis. The claustrum core projects predominantly to frontal-midline cortical regions, whereas the dorsal and ventral shell project to the cortical motor system and temporal lobe, respectively. Anatomically connected cortical regions receive common input from a subset of claustrum neurons shared by neighboring modules, whereas spatially separated regions of cortex are innervated by different claustrum modules. Therefore, each output module exhibits a unique position within the claustrum and overlaps substantially with other modules projecting to functionally related cortical regions. Claustrum inhibitory cells containing parvalbumin, somatostatin, and neuropeptide Y also show unique topographical distributions, suggesting different output modules are controlled by distinct inhibitory circuit motifs. The topographic organization of excitatory and inhibitory cell types may enable parallel claustrum outputs to independently coordinate distinct cortical networks.

The claustrum.

The claustrum is a brain region that has been investigated for over 200 years, yet its precise function remains unknown. In the final posthumously released article of Francis Crick, written with Christof Koch, the claustrum was suggested to be critically linked to consciousness. Though the claustrum remained relatively obscure throughout the last half century, it has enjoyed a renewed interest in the last 15 years since Crick and Koch's article. During this time, the claustrum, like many other brain regions, has been studied with the myriad of modern systems neuroscience tools that have been made available by the intersection of genetic and viral technologies. This has uncovered new information about its anatomical connectivity and physiological properties and begun to reveal aspects of its function. From these studies, one clear consensus has emerged which supports Crick and Koch's primary interest in the claustrum: the claustrum has widespread extensive connectivity with the entire cerebral cortex, suggesting a prominent role in 'higher order processes'.

A claustrum in reptiles and its role in slow-wave sleep.

The mammalian claustrum, owing to its widespread connectivity with other forebrain structures, has been hypothesized to mediate functions that range from decision-making to consciousness1. Here we report that a homologue of the claustrum, identified by single-cell transcriptomics and viral tracing of connectivity, also exists in a reptile-the Australian bearded dragon Pogona vitticeps. In Pogona, the claustrum underlies the generation of sharp waves during slow-wave sleep. The sharp waves, together with superimposed high-frequency ripples2, propagate to the entire neighbouring pallial dorsal ventricular ridge (DVR). Unilateral or bilateral lesions of the claustrum suppress the production of sharp-wave ripples during slow-wave sleep in a unilateral or bilateral manner, respectively, but do not affect the regular and rapidly alternating sleep rhythm that is characteristic of sleep in this species3. The claustrum is thus not involved in the generation of the sleep rhythm itself. Tract tracing revealed that the reptilian claustrum projects widely to a variety of forebrain areas, including the cortex, and that it receives converging inputs from, among others, areas of the mid- and hindbrain that are known to be involved in wake-sleep control in mammals4-6. Periodically modulating the concentration of serotonin in the claustrum, for example, caused a matching modulation of sharp-wave production there and in the neighbouring DVR. Using transcriptomic approaches, we also identified a claustrum in the turtle Trachemys scripta, a distant reptilian relative of lizards. The claustrum is therefore an ancient structure that was probably already present in the brain of the common vertebrate ancestor of reptiles and mammals. It may have an important role in the control of brain states owing to the ascending input it receives from the mid- and hindbrain, its widespread projections to the forebrain and its role in sharp-wave generation during slow-wave sleep.

The Anatomy and Physiology of Claustrum-Cortex Interactions.

The claustrum is one of the most widely connected regions of the forebrain, yet its function has remained obscure, largely due to the experimentally challenging nature of targeting this small, thin, and elongated brain area. However, recent advances in molecular techniques have enabled the anatomy and physiology of the claustrum to be studied with the spatiotemporal and cell type-specific precision required to eventually converge on what this area does. Here we review early anatomical and electrophysiological results from cats and primates, as well as recent work in the rodent, identifying the connectivity, cell types, and physiological circuit mechanisms underlying the communication between the claustrum and the cortex. The emerging picture is one in which the rodent claustrum is closely tied to frontal/limbic regions and plays a role in processes, such as attention, that are associated with these areas.

Microneuroanatomy of the Anterior Frontal Laser Trajectory to the Insula.

BACKGROUND: Magnetic resonance imaging-guided laser interstitial thermal therapy (LITT) is an emerging minimally invasive procedure for the treatment of deep intracranial lesions. Insular lesions are challenging to treat because of the risk of damaging important surrounding structures. The precise knowledge of the neural structures that are at risk along the trajectory and during the ablation is essential to reduce associated complications. This study aims to describe the relevant anatomy of the anterior frontal LITT trajectory to the insular region by using sectional anatomy and fiber dissection technique. METHODS: Three silicone-injected cadaveric heads were used to implant laser catheters bilaterally to the insular region by using a frameless stereotactic technique from a frontal approach. Sections were cut in both the oblique axial plane parallel to the trajectory and in the coronal plane. White matter fiber dissections were used to establish the tracts related to the laser trajectory from lateral to medial and medial to lateral. RESULTS: Supraorbital regions were selected as entry points. After crossing the frontal bone, the track intersected the inferior frontal lobe. The catheter was illustrated reaching the insular region medial to the inferior fronto-occipital fasciculus and insular cortex, and superior to the uncinate fasciculus. The uncinate fasciculus, extreme capsule, claustrum, external capsule, and putamen were traversed, preserving the major vascular structures. CONCLUSIONS: Independent of the insular area treated, an understanding of the neuroanatomy related to the anterior frontal laser trajectory is essential to improve the ability to perform LITT of this challenging region.

Resting state functional connectivity and cognitive task-related activation of the human claustrum.

Structural and functional analyses of the human claustrum, a poorly understood telencephalic gray matter structure, are hampered by its sheet-like anatomical arrangement. Here, we first describe a functional magnetic resonance imaging (fMRI) method to reveal claustrum signal with no linear relationship with adjacent regions in human subjects. We applied this approach to resting state functional connectivity (RSFC) analysis of the claustrum at high resolution (1.5 mm isotropic voxels) using a 7T dataset (n = 20) and a separate 3T dataset for replication (n = 35). We then assessed claustrum activation during performance of a cognitive task, the multi-source interference task, at 3T (n = 33). Extensive functional connectivity was observed between claustrum and cortical regions associated with cognitive control, including anterior cingulate, prefrontal and parietal cortices. Cognitive task performance was associated with widespread activation and deactivation that overlapped with the cortical areas showing functional connectivity to the claustrum. Furthermore, during high cognitive conflict conditions of the task, the claustrum was significantly activated at the onset of the task, but not during the remainder of the difficult condition. Both of these findings suggest that the human claustrum can be functionally isolated with fMRI, and that it may play a role in cognitive control, and specifically task switching, independent of sensorimotor processing.

The Claustrum of the Pig: An Immunohistochemical and a Quantitative Golgi Study.

Despite increasing interest in the claustrum (Cl) over the last decades, its function is still a puzzling problem. Among the experimental species of potential use in Cl research, the pig is considered an interesting model, because of the similarities of its brain with the corresponding cortical and subcortical human structures. The swine Cl presents a peculiar morphology, characterized by a wide posterior enlargement, ideal for physiological investigations. There is a wealth of data on general anatomy, cytoarchitecture, and chemo architecture of the Cl, but much less is known about the dendritic morphometry of its neurons. Dendritic length and branching pattern are key features to understand the organization of the microcircuitry, and thus the delineation of the structure-function relationships of the Cl. To this effect, we undertook (a) a quantitative study of the dendrites of the spiny neurons of the swine Cl, employing the Golgi staining; and (b) an immunohistochemical analysis to describe the distribution of the parvalbumin (PV)-immunoreactive interneurons throughout the same nucleus. Taken together, the results that we report here show that the dendritic architecture and the distribution of the PV expressing interneurons change when the Cl of this species changes its shape along the rostro-caudal axis, thus suggesting a potentially specific function for the large posterior puddle. Anat Rec, 302:1638-1646, 2019. © 2019 American Association for Anatomy.

The relationship between the claustrum and endopiriform nucleus: A perspective towards consensus on cross-species homology.

With the emergence of interest in studying the claustrum, a recent special issue of the Journal of Comparative Neurology dedicated to the claustrum (Volume 525, Issue 6, pp. 1313-1513) brought to light questions concerning the relationship between the claustrum (CLA) and a region immediately ventral known as the endopiriform nucleus (En). These structures have been identified as separate entities in rodents but appear as a single continuous structure in primates. During the recent Society for Claustrum Research meeting, a panel of experts presented data pertaining to the relationship of these regions and held a discussion on whether the CLA and En should be considered (a) separate unrelated structures, (b) separate nuclei within the same formation, or (c) subregions of a continuous structure. This review article summarizes that discussion, presenting comparisons of the cytoarchitecture, neurochemical profiles, genetic markers, and anatomical connectivity of the CLA and En across several mammalian species. In rodents, we conclude that the CLA and the dorsal endopiriform nucleus (DEn) are subregions of a larger complex, which likely performs analogous computations and exert similar effects on their respective cortical targets (e.g., sensorimotor versus limbic). Moving forward, we recommend that the field retain the nomenclature currently employed for this region but should continue to examine the delineation of these structures across different species. Using thorough descriptions of a variety of anatomical features, this review offers a clear definition of the CLA and En in rodents, which provides a framework for identifying homologous structures in primates.

Input-output organization of the mouse claustrum.

Progress in determining the precise organization and function of the claustrum (CLA) has been hindered by the difficulty in reliably targeting these neurons. To overcome this, we used a projection-based targeting strategy to selectively label CLA principal neurons. Combined with adeno-associated virus (AAV) and monosynaptic rabies tracing techniques, we systematically examined the pre-synaptic input and axonal output of this structure. We found that CLA neurons projecting to retrosplenial cortex (RSP) collateralize extensively to innervate a variety of higher-order cortical regions. No subcortical labeling was found, with the exception of sparse terminals in the basolateral amygdala (BLA). This pattern of output was similar to cingulate- and visual cortex-projecting CLA neurons, suggesting a common targeting scheme among these projection-defined populations. Rabies virus tracing directly demonstrated widespread synaptic inputs to RSP-projecting CLA neurons from both cortical and subcortical areas. The strongest inputs arose from classically defined limbic regions, including medial prefrontal cortex, anterior cingulate, BLA, ventral hippocampus, and neuromodulatory systems such as the dorsal raphe and cholinergic basal forebrain. These results suggest that the CLA may integrate information related to the emotional salience of stimuli and may globally modulate cortical state by broadcasting its output uniformly across a variety of higher cognitive centers.