siibra-explorer is built around an interactive 3D view of the brain displaying a unique selection of detailed templates and parcellation maps for the human, macaque, rat or mouse brain, including BigBrain as a microscopic resolution human brain model at its full resolution of 20 micrometres.
A Python library for using EBRAINS brain atlases in scripts, notebooks and computational workflows for reproducible analysis and modeling.
- Access preconfigured parcellations, maps, regions and templates.
- Define locations in the brain as Python objects, and perform probabilistic assignment of brain regions to locations.
- Retrieve multimodal datasets assigned to brain regions as common Python structures.
- Sample from ultrahigh-resolution 2D and 3D image data.
The Python library siibra-python is designed for integrating EBRAINS atlases into scripts and computational workflows. Besides providing programmatic access to all functionalities available in the interactive viewer siibra-explorer, it enables more advanced utilization of atlas information.
Access to atlases
siibra-python provides streamlined access to preconfigured parcellations, 3D maps, region masks and reference templates supported in EBRAINS. This includes searching parcellations and region hierarchies, fetching volumetric and surface maps in different reference spaces, and safe handling of the links between brain regions and their corresponding voxel and vertex labels. Maps and templates are modeled as commonly used data types, such as Nifti1Images and numpy arrays. This way, they are well supported by popular Python libraries such as nilearn, and can be easily visualized and processed.
Anatomical assignment of locations in the brain
siibra-python provides data structures to model locations in the brain as images, points, or bounding boxes. Locations are clearly linked to reference coordinate systems, can be safely compared across different reference spaces, and can be explicitly modelled with location accuracy. The library provides comprehensive functionality to perform probabilistic assignment of brain regions to locations, to tag objects with brain structures. Such assignments are automatically qualified by relationships such as incidence, overlap, or containedness.
Using brain regions or locations in the brains, siibra-python provides efficient mechanisms to retrieve related data measurements, and thus programmatically collect multimodal characterizations of brain architecture. The data features include measures of molecular, cellular and fiber architecture, structural and functional connectivity matrices, and many forms of image data linked to the brain atlases.
Big data sampling
siibra-python models small and large image data in a coherent fashion. In particular, besides downloading MRI volumes as NIfTI images, you can also fetch custom samples from ultrahigh-resolution image data such as the microscopic BigBrain model or micrometer-resolution image sections stored in the cloud to obtain properly localized images of regions of interest.
Other softwareAll software
siibra-api is an HTTP API for querying and retrieving contents of EBRAINS atlases. Originally built as a backend service for the interactive atlas viewer siibra-explorer, the API has been documented for connecting the brain atlases to other applications and web services.