Waxholm space (WHS) is a coordinate-based canonical space currently implemented for the C57BL/6J mouse strain, which provides a spatial standard that aids in translation between different digital atlasing efforts. WHS was created by the International Neuroinformatics Coordinating Facility (INCF) Digital Brain Atlasing Program and was named after Waxholm, Sweden, the site of the first INCF Digital Atlasing Working Group. The idea of using a reference space for the canonical mapping and registration of anatomic based data is not new, but the goal is for WHS to act as the center of a digital atlasing architecture for sharing data. Currently WHS is defined by a high-quality, multi-modal dataset created specifically for this purpose, but it is expected to evolve as higher-resolution data are added to this canonical space.
WHS is simply a continuous Cartesian coordinate system and similar in concept to Talairach space for the human brain. While the position of the origin is arbitrary and readily transformable, in WHS, it is placed at the intersection of the midsagittal plane (Z), a plane tangentinal to the rostral (Y) and another to the dorsal anterior commissure (X). It is currently determined by the intersection of the mid-sagittal plane and the median rostral/dorsal edges of the anterior commissure (as they appear in horizontal/coronal sections).
WHS was created using a multi-spectral dataset containing 3 different MR microscopic volumes and a Nissl volume. MR data were acquired at 9.4T in specimens perfused with formalin/Prohance (Johnson GA 2007). Specimens were imaged with T1, T2, and T2* weighted sequences at 21.5 μm³ resolution with the brains in the skull. The brains were later frozen and cryosectioned using a low-distortion tape collection protocol (Bertrand 2008). Each 20 μm section was then collected and Nissl-stained. Following its 3D reconstruction, the Nissl volume was aligned to the MR volumes. Since the MR data are acquired without moving, all three MR datasets are co-registered. By registering the Nissl volume to this space there are five different potential registration entry points into WHS: three MRI volumes with different tissue contrasts, a Nissl volume, and a volume of delineated structures. The WHS datasets are freely available in NIfTI-1 format at the INCF Software Center.
A standardized space by itself does not create a digital atlasing framework that allows the linking of multiple data types from different sources for processing, upload, analysis, data retrieval, and sharing. This INCF program is also in the process of fleshing out such a framework, the INCF Digital Atlasing Infrastructure (INCF-DAI). An INCF-DAI prototype currently includes a collection of distributed services that support the publication, discovery, and invocation of heterogeneous atlas resources. Central servers reference remote and autonomously supported and updated resources, and host or mirror some resources that are critical for operation of this infrastructure. The current version of the INCF-DAI prototype for the mouse brain supports mapping between WHS reference space and the following:
While this prototype demonstrates several key operations of the INCF-DAI, a full implementation of a digital atlasing system will require the support and active participation of the larger scientific community to further refine and expand it.