American scientists have created the full symmetric three-dimensional model of the mouse brain with a resolution of 10 micrometers. To generate the Atlas, the researchers used averaged 1675 laboratory mice, a model of the brain. Scans were obtained using two-photon tomography and then correlated with their structures in three-dimensional space, focusing on the histological data and the existing anatomical atlases. Atlas, you can view on the official website, and the scientists described in the article published in Cell.
Data obtained with the help of modern neuroimaging techniques now allow us to make fairly detailed atlases of the brain — both human and laboratory animals. Most of them, however, focus only on one modality: it may be, for example, the complete connectome of all neural connections (for example, this is for round worm), or an anatomical Atlas with the distribution of each plot (same thing for fruit flies).
Each Atlas individually is very important for understanding the brain of the animal, and versatile tool for this may be their combination into a single Atlas, ideally three-dimensional. Scientists led by Lydia Eun (Lydia Ng) of Ulanovskaja Institute for the study of the brain in the frame of the project of the Allen Brain Atlas has created such an Atlas for the laboratory mouse.
To generate the Atlas, the scientists used the brain of a young mouse line C57BL/6 — they are used as laboratory mice most often. Scientists have received one image of the brain 1675 mice, scans were made using two-photon imaging with a resolution of 100 micrometers each. To the final Atlas was symmetrical, each image is mirrored at the sagittal section, thus increasing their number in half. Then the images are averaged: small deviations in the coordinates used to increase the resolution first to 50 micrometers, then to 25 and finally to 10 micrometers. Thus, each voxel datetimecontrol averaged 335 images of the hemispheres, and in the Atlas of around 506 million: high resolution allowed precise separation of the layers of the cortex, the white matter tracts and subcortical areas of gray matter.
Next, the researchers spaced voxels, taking advantage of the existing atlases, as well as data obtained during histological examination, depending on what kind of structure it was necessary to mark it. For example, in order to separate the parts of the cortex and separated them from the subcortical structures, the researchers used data about gene expression in different tissues. After that, each structure is visualized in three-dimensional space averaged and made the final three-dimensional model.