A framework based on sulcal constraints to align preterm, infant and adult human brain images acquired in vivo and post mortem.

Robust spatial alignment of post mortem data and in vivo MRI acquisitions from different ages, especially from the early developmental stages, into standard spaces is still a bottleneck hampering easy comparison with the mainstream neuroimaging results. In this paper, we test a landmark-based spatial normalization strategy as a framework for the seamless integration of any macroscopic dataset in the context of the Human Brain Project (HBP). This strategy stems from an approach called DISCO embedding sulcal constraints in a registration framework used to initialize DARTEL, the widely used spatial normalization approach proposed in the SPM software. We show that this strategy is efficient with a heterogeneous dataset including challenging data as preterm newborns, infants, post mortem histological data and a synthetic atlas computed from averaging the ICBM database, as well as more commonly studied data acquired in vivo in adults. We then describe some perspectives for a research program aiming at improving folding pattern matching for atlas inference in the context of the future HBP's portal.

Spatial normalization of brain images and beyond.

The deformable atlas paradigm has been at the core of computational anatomy during the last two decades. Spatial normalization is the variant endowing the atlas with a coordinate system used for voxel-based aggregation of images across subjects and studies. This framework has largely contributed to the success of brain mapping. Brain spatial normalization, however, is still ill-posed because of the complexity of the human brain architecture and the lack of architectural landmarks in standard morphological MRI. Multi-atlas strategies have been developed during the last decade to overcome some difficulties in the context of segmentation. A new generation of registration algorithms embedding architectural features inferred for instance from diffusion or functional MRI is on the verge to improve the architectural value of spatial normalization. A better understanding of the architectural meaning of the cortical folding pattern will lead to use some sulci as complementary constraints. Improving the architectural compliance of spatial normalization may impose to relax the diffeomorphic constraint usually underlying atlas warping. A two-level strategy could be designed: in each region, a dictionary of templates of incompatible folding patterns would be collected and matched in a way or another using rare architectural information, while individual subjects would be aligned using diffeomorphisms to the closest template. Manifold learning could help to aggregate subjects according to their morphology. Connectivity-based strategies could emerge as an alternative to deformation-based alignment leading to match the connectomes of the subjects rather than images.

Cytoarchitecture, probability maps and functions of the human frontal pole.

The frontal pole has more expanded than any other part in the human brain as compared to our ancestors. It plays an important role for specifically human behavior and cognitive abilities, e.g. action selection (Kovach et al., 2012). Evidence about divergent functions of its medial and lateral part has been provided, both in the healthy brain and in psychiatric disorders. The anatomical correlates of such functional segregation, however, are still unknown due to a lack of stereotaxic, microstructural maps obtained in a representative sample of brains. Here we show that the human frontopolar cortex consists of two cytoarchitectonically and functionally distinct areas: lateral frontopolar area 1 (Fp1) and medial frontopolar area 2 (Fp2). Based on observer-independent mapping in serial, cell-body stained sections of 10 brains, three-dimensional, probabilistic maps of areas Fp1 and Fp2 were created. They show, for each position of the reference space, the probability with which each area was found in a particular voxel. Applying these maps as seed regions for a meta-analysis revealed that Fp1 and Fp2 differentially contribute to functional networks: Fp1 was involved in cognition, working memory and perception, whereas Fp2 was part of brain networks underlying affective processing and social cognition. The present study thus disclosed cortical correlates of a functional segregation of the human frontopolar cortex. The probabilistic maps provide a sound anatomical basis for interpreting neuroimaging data in the living human brain, and open new perspectives for analyzing structure-function relationships in the prefrontal cortex. The new data will also serve as a starting point for further comparative studies between human and non-human primate brains. This allows finding similarities and differences in the organizational principles of the frontal lobe during evolution as neurobiological basis for our behavior and cognitive abilities.

Stereotaxic probabilistic maps of the magnocellular cell groups in human basal forebrain.

The basal forebrain contains several interdigitating anatomical structures, including the diagonal band of Broca, the basal nucleus of Meynert, the ventral striatum, and also cell groups underneath the globus pallidus that bridge the centromedial amygdala to the bed nucleus of the stria terminalis. Among the cell populations, the magnocellular, cholinergic corticopetal projection neurons have received particular attention due to their loss in Alzheimer's disease. In MRI images, the precise delineation of these structures is difficult due to limited spatial resolution and contrast. Here, using microscopic delineations in ten human postmortem brains, we present stereotaxic probabilistic maps of the basal forebrain areas containing the magnocellular cell groups. Cytoarchitectonic mapping was performed in silver stained histological serial sections. The positions and the extent of the magnocellular cell groups within the septum (Ch1-2), the horizontal limb of the diagonal band (Ch3), and in the sublenticular part of the basal forebrain (Ch4) were traced in high-resolution digitized histological sections, 3D reconstructed, and warped to the reference space of the MNI single subject brain. The superposition of the cytoarchitectonic maps in the MNI brain shows the intersubject variability of the various Ch compartments and their stereotaxic position relative to other brain structures. Both the right and left Ch4 regions showed significantly smaller volumes when age was considered as a covariate. Probabilistic maps of compartments of the basal forebrain magnocellular system are now available as an open source reference for correlation with fMRI, PET, and structural MRI data of the living human brain.

Cytoarchitectonic mapping of the human amygdala, hippocampal region and entorhinal cortex: intersubject variability and probability maps.

Probabilistic maps of neocortical areas and subcortical fiber tracts, warped to a common reference brain, have been published using microscopic architectonic parcellations in ten human postmortem brains. The maps have been successfully applied as topographical references for the anatomical localization of activations observed in functional imaging studies. Here, for the first time, we present stereotaxic, probabilistic maps of the hippocampus, the amygdala and the entorhinal cortex and some of their subdivisions. Cytoarchitectonic mapping was performed in serial, cell-body stained histological sections. The positions and the extent of cytoarchitectonically defined structures were traced in digitized histological sections, 3-D reconstructed and warped to the reference space of the MNI single subject brain using both linear and non-linear elastic tools of alignment. The probability maps and volumes of all structures were calculated. The precise localization of the borders of the mapped regions cannot be predicted consistently by macroanatomical landmarks. Many borders, e.g. between the subiculum and entorhinal cortex, subiculum and Cornu ammonis, and amygdala and hippocampus, do not match sulcal landmarks such as the bottom of a sulcus. Only microscopic observation enables the precise localization of the borders of these brain regions. The superposition of the cytoarchitectonic maps in the common spatial reference system shows a considerably lower degree of intersubject variability in size and position of the allocortical structures and nuclei than the previously delineated neocortical areas. For the first time, the present observations provide cytoarchitectonically verified maps of the human amygdala, hippocampus and entorhinal cortex, which take into account the stereotaxic position of the brain structures as well as intersubject variability. We believe that these maps are efficient tools for the precise microstructural localization of fMRI, PET and anatomical MR data, both in healthy and pathologically altered brains.

Linking retinotopic fMRI mapping and anatomical probability maps of human occipital areas V1 and V2.

Using functional MRI, we characterized field sign maps of the occipital cortex and created three-dimensional maps of these areas. By averaging the individual maps into group maps, probability maps of functionally defined V1 or V2 were determined and compared to anatomical probability maps of Brodmann areas BA17 and BA18 derived from cytoarchitectonic analysis (Amunts, K., Malikovic, A., Mohlberg, H., Schormann, T., Zilles, K., 2000. Brodmann's areas 17 and 18 brought into stereotaxic space-where and how variable? NeuroImage 11, 66-84). Comparison of areas BA17/V1 and BA18/V2 revealed good agreement of the anatomical and functional probability maps. Taking into account that our functional stimulation (due to constraints of the visual angle of stimulation achievable in the MR scanner) only identified parts of V1 and V2, for statistical evaluation of the spatial correlation of V1 and BA17, or V2 and BA18, respectively, the a priori measure kappa was calculated testing the hypothesis that a region can only be part of functionally defined V1 or V2 if it is also in anatomically defined BA17 or BA18, respectively. kappa = 1 means the hypothesis is fully true, kappa = 0 means functionally and anatomically defined visual areas are independent. When applying this measure to the probability maps, kappa was equal to 0.84 for both V1/BA17 and V2/BA18. The data thus show a good correspondence of functionally and anatomically derived segregations of early visual processing areas and serve as a basis for employing anatomical probability maps of V1 and V2 in group analyses to characterize functional activations of early visual processing areas.

The importance of seeing it coming: a functional magnetic resonance imaging study of motion-in-depth towards the human observer.

Despite their crucial biological relevance, the neural structures differentially activated by the detection of optic flow towards the observer remain to be elucidated. Here, we deploy functional magnetic resonance imaging with normal volunteers to locate the areas differentially activated when motion towards the observer is detected. Motion towards the observer, compared with motion away, showed significant activations (P<0.05, corrected for multiple comparisons), as assessed using statistical parametric mapping, in the lateral inferior occipital cortex bilaterally and in right lateral superior occipital cortex. The areas implicated do not extend into area V5 or subdivisions thereof.Our data suggest that the representations of motion towards the observer implicate perceptual and attentional mechanisms acting at early stages of visual processing in extrastriate cortex. From the standpoint of efficient biological engineering, it makes sense that such crucially important functions as object motion towards the observer would be computed in early visual processing areas. Further studies will be required to determine the extent to which the effects we observed in lateral occipital cortex reflect differential attention to different types of motion, as contrasted with the derivation of explicit representations of motion towards the observer.

Integration of microstructural and functional aspects of human somatosensory areas 3a, 3b, and 1 on the basis of a computerized brain atlas.

In this study we analyzed structural and functional aspects of the human primary somatosensory areas 3a, 3b, and 1 on the basis of a computerized brain atlas. The approach overcomes many of the problems associated with subjective architectonic parcellations of the cortex and with 'classical" brain maps published in a "rigid" print format. Magnetic resonance (MR) scans were obtained from ten postmortem brains. The brains were serially sectioned at 20 microm, and sections were stained for cell bodies. Areas 3a, 3b, and 1 were delineated statistically on the basis of differences in the laminar densities of neuronal cell bodies. The borders of the areas were topographically variable across different brains and did not match macroanatomical landmarks of the postcentral gyrus. After correction of the sections for deformations due to histological processing, each brain's 3-D reconstructed histological volume and the volume representations of areas 3a, 3b, and 1 were adapted to the reference brain of a computerized atlas and superimposed in 3-D space. For each area, a population map was generated that described, for each voxel, how many brains had a representation of that area. Despite considerable interindividual variability, representations of areas 3a, 3b, and 1 in > or = 50% of the brains were found in the fundus of the central sulcus, in the rostral bank, and on the crown of the postcentral gyrus, respectively. For each area, a volume of interest (VOI) was defined that encompassed that area's representation in > or = 50% of the brains. Despite close spatial relationship in the postcentral gyrus, the three VOIs overlapped by < 1% of their volumes. Changes in regional cerebral blood flow (rCBF) were measured with positron emission tomography when six right-handed subjects discriminated differences in the speed of a rotating brush stimulating the palmar surface of the right hand. With co-registered MR images, the rCBF data were adapted to the same reference brain and superimposed with the microstructural VOIs. Discrimination of moving stimuli, contrasted to rest, increased the rCBF in the VOIs of areas 3b and 1, but not in area 3a. This approach opens up the possibility of (1) defining VOIs of cortical areas which are not based on macroanatomical landmarks but instead on observer-independent cytoarchitectonic mapping of postmortem brains and of (2) determining in these VOIs changes in rCBF data obtained from functional imaging experiments.

Areas 3a, 3b, and 1 of human primary somatosensory cortex. Part 2. Spatial normalization to standard anatomical space.

Interindividual topographical variability of cytoarchitectonically defined somatosensory areas 3a, 3b, and 1 was analyzed in the standard anatomical format of a computerized brain atlas. T1-weighted magnetic resonance images were obtained from 10 postmortem brains. The brains were serially sectioned at 20 mcm, sections were stained for cell bodies, and areas 3a, 3b, and 1 were defined with an observer-independent cytoarchitectonic technique. After correction of the sections for deformations due to histological processing, the 3-D reconstructed histological volumes of the individual brains and the volume representations of the cytoarchitectonic areas were adapted to the reference brain of a computerized atlas. Corresponding areas were superimposed in the 3-D space of the reference brain. These population maps describe, for each voxel, how many brains have a representation of one particular cytoarchitectonic area. Each area's extent is very variable across different brains, but representations of areas 3a, 3b, and 1 in >/=50% of the brains were found in the fundus of the central sulcus, its caudal bank, and on the crown of the postcentral gyrus, respectively. Volumes of interest (VOIs) were defined for each area in which >/=50% of the brains have a representation of that area. Despite close spatial relationship of areas 3a, 3b, and 1 in the postcentral gyrus, the three VOIs overlap by <1% of their volumes. Functional imaging data can now be brought into the same standard anatomical format, and changes in regional cerebral blood flow can be calculated in VOIs of areas 3a, 3b, and 1, which are derived from genuine cytoarchitectonic data.

Interhemispheric asymmetry of the human motor cortex related to handedness and gender.

Most people are right-handed, preferring the right hand for skilled as well as unskilled activities, but a notable proportion are mixed-handed, preferring to use the right hand for some actions and the left hand for others. Assuming a structural/functional correlation in the motor system we tested whether asymmetries in hand performance in consistent right and left handers as well as in mixed handers are associated with anatomical asymmetries in the motor cortex. In vivo MR morphometry was used for analyzing interhemispheric asymmetry in the depth of the central sulcus in the region of cortical hand representation of 103 healthy subjects. Subjects were tested both for hand preference and hand performance. As expected, left-right differences in hand performance differed significantly between consistent right, consistent left and mixed handers and were independent on gender. Male consistent right handers showed a significant deeper central sulcus on the left hemisphere than on the right. Anatomical asymmetries decreased significantly from male consistent right over mixed to consistent left handers. Sixty two per cent of consistent left handers revealed a deeper central sulcus on the right than on the left hemisphere, but for the group as a whole this rightward asymmetry was not significant. No interhemispheric asymmetry was found in females. Thus, anatomical asymmetry was associated with handedness only in males, but not in females, suggesting sex differences in the cortical organization of hand movements.

Brodmann's areas 17 and 18 brought into stereotaxic space-where and how variable?

Studies on structural-functional associations in the visual system require precise information on the location and variability of Brodmann's areas 17 and 18. Usually, these studies are based on the Talairach atlas, which does not rely on cytoarchitectonic observations, but on comparisons of macroscopic features in the Talairach brain and Brodmann's drawing. In addition, in this atlas are found only the approximate positions of cytoarchitectonic areas and not the exact borders. We have cytoarchitectonically mapped both areas in 10 human brains and marked their borders in corresponding computerized images. Borders were defined on the basis of quantitative cytoarchitecture and multivariate statistics. In addition to borders of areas 17 and 18, subparcellations within both areas were found. The cytoarchitectonically defined areas were 3-D reconstructed and transferred into the stereotaxic space of the standard reference brain. Surface rendering of the brains revealed high individual variability in size and shape of the areas and in the relationship to the free surface and sulci. Ranges and centers of gravity of both areas were calculated in Talairach coordinates. The positions of areas 17 and 18 in the stereotaxic space differed between the hemispheres. Both areas reached significantly more caudal and medial positions on the left than on the right. Probability maps were created in which the degree of overlap in each stereotaxic position was quantified. These maps of areas 17 and 18 are the first of their kind and contain precise stereotaxic information on both interhemispheric and interindividual differences.

Broca's region revisited: cytoarchitecture and intersubject variability.

The sizes of Brodmann's areas 44 and 45 (Broca's speech region) and their extent in relation to macroscopic landmarks and surrounding areas differ considerably among the available cytoarchitectonic maps. Such variability may be due to intersubject differences in anatomy, observer-dependent discrepancies in cytoarchitectonic mapping, or both. Because a reliable definition of cytoarchitectonic borders is important for interpreting functional imaging data, we mapped areas 44 and 45 by means of an observer-independent technique. In 10 human brains, the laminar distributions of cell densities were measured vertical to the cortical surface in serial coronal sections stained for perikarya. Thousands of density profiles were obtained. Cytoarchitectonic borders were defined as statistically significant changes in laminar patterns. The analysis of the three-dimensional reconstructed brains and the two areas showed that cytoarchitectonic borders did not consistently coincide with sulcal contours. Therefore, macroscopic features are not reliable landmarks of cytoarchitectonic borders. Intersubject variability in the cytoarchitecture of areas 44 and 45 was significantly greater than cytoarchitectonic differences between these areas in individual brains. Although the volumes of area 44 differed across subjects by up to a factor of 10, area 44 but not area 45 was left-over-right asymmetrical in all brains. All five male but only three of five female brains had significantly higher cell densities on the left than on the right side. Such hemispheric and gender differences were not detected in area 45. These morphologic asymmetries of area 44 provide a putative correlate of the functional lateralization of speech production.

Quantitative analysis of sulci in the human cerebral cortex: development, regional heterogeneity, gender difference, asymmetry, intersubject variability and cortical architecture.

The degree of cortical folding (GI) and the relation between sulci and borders of cyto- and receptorarchitectonically defined areas were analyzed in postmortem human brains. The GI reaches adult levels (with highest values in the association cortices) around birth and does not decrease during aging. It shows a sex-dependent left-over-right asymmetry. Sulci and borders of architectonical areas coincide only in a few examples; thus, sulci are not generally valid landmarks of the microstructural organization of the cortex. Individual sulci were studied in 3D-reconstructed MRI sequences of living brains. A considerable intersubject variability of the distances between the sulcal surfaces of individual brains and their mean sulcal surfaces is apparent. The depth of the central sulcus varies with manual skill and handedness.