Multimodal digital atlas of the rat brain: data integration in a high resolution volumetric template

Understanding of functional brain architecture is fostered by the use of neuroinformatics tools and approaches for combining, analysing, and interpreting information from numerous anatomical datasets. Spatial information in different datasets is, however, trapped in diverse tools, formats and coordinate systems, making it tedious to compare datasets across animals and data modalities. Conventional 2D atlases lack dynamic features, e.g. reslicing of an image set, needed for analysis of 3D spatial distribution patterns in the brain. Definitions of brain structures in such atlases are typically based on single brains, for which delineation criteria are generally not available. To address these challenges, an integrative approach is needed. As part of such an approach, we have initiated a project aimed at producing a novel interactive atlas for the rat brain: a digital 3D atlas based on a volumetric template, featuring an annotation layer with anatomical boundaries of structures and information about the criteria used for the delineations. The volumetric template used to construct the atlas consists of high resolution ex vivo magnetic resonance images, including T₁ anatomical images (44 μm isotropic resolution) and diffusion tensor images (88 μm isotropic resolution) from a Sprague-Dawley rat. Boundaries of brain structures are defined on the basis of contrast in the MRI and DTI data and with use of a number of other imaging modalities and histology taken from other brains and registered to the template.  We report on the status of creating the core template for this atlas,  and on the use of the new Waxholm coordinate system, which will connect the new atlas to a series of interoperable resources and services aimed at facilitating data analysis and multilevel data integration in a common framework.