The EBRAINS multilevel human brain atlas integrates maps that capture different facets of human brain organization into a common framework. It is defined across multiple reference spaces, zooming up to the micrometer level of the BigBrain. The atlas currently links cytoarchitectonic areas with ultra-high resolution BigBrain data, fiber bundles from MRI tractography, maps of functional modules, and a whole range of multimodal data features associated to brain regions. The range and scope of available maps and features is constantly increasing. As the amount of data constituting the human atlas exceeds the disk capacity of most computers, it is designed as an online framework that you can access in a number of ways. The main entry point is our interactive 3D viewer, which allows you to explore the different facets of the atlas in 3D and to select specific data features for download. Programmatic interfaces will soon be available.
Julich-Brain cytoarchitectonic maps represent microstructural reference data of the multilevel atlas. They define cortical and subcortical brain regions based on reproducible border delineations, performed at microscopic resolution in serial, cell-body stained histological sections of ten post-mortem brains. The probabilistic maps capture variations of areas between examined brains in size and extent. They are available in different reference spaces. A maximum probability map summarizes all maps by assigning each voxel to the most likely histological area.
The atlas features probabilistic maps of deep and superficial white matter fibre bundles, capturing variability and structure of fibre architecture at the macroscopic scale. These maps are based on diffusion MRI and tractography, initially inferred from MRI scans of the ARCHI dataset, and more recently from the Human Connectome Project resulting in more than 700 short bundles. As for the cytoarchitectonic maps, a maximum probability parcellation summarizes the maps by assigning to each voxel the most likely bundle.
The 3D BigBrain model represents the microscopic resolution reference space of the multilevel atlas. It is a 3D-reconstruction from 7404 digitized histological brain sections at nearly cellular resolution of 20 micrometer. The multilevel atlas provides detailed 3D maps of brain areas, which link directly to corresponding maps in MNI space, as well as maps of cortical layers.
Maps of dictionaries of functional modes (DiFuMo) at different granularities ranging from 64 to 1024 networks are being offered. These functional modes are sampled on millions of functional brain (fMRI) volumes, spanned over 27 studies and many research groups. These maps help to solve classic analysis problems on a wide variety of functional images: stimuli decoding, standard GLM analysis of fMRI across sessions and individuals, resting-state functional- connectomes biomarkers.
Use our interactive 3D viewer to navigate multiple human brain template spaces that have been integrated into a common framework. Select different modalities and map granularities of maps, and explore these atlases in 3D, navigate through them using the brain region hierarchies provided, zoom to microscopic resolution and examine brain function and connectivity. Select specific components, maps, and data features for download. Through interactive plugins, perform simple data analysis tasks, such as comparing gene expressions in different atlas regions.
EBRAINS is developing a Python client, HTTP API and command line interface (CLI) to provide programmatic access to all the maps, data features and functions of the Multi-level Human Brain Atlas. These will allow you to build automated workflows for retrieving information about regions, maps, and multi-modal data features, and to make use of the EBRAINS Human Brain Atlas for data analysis and modelling. First releases of these interfaces will be available in 2021. Stay connected, or get in touch with us if you are interested in testing an early release.