BigBrain: an ultra-high-resolution 3D roadmap of the human brain

June 21, 2013

BigBrain (credit: Montreal Neurological Institute and Forschungszentrum Jülich)

A landmark three-dimensional (3-D) digital reconstruction of a complete human brain, called the BigBrain, shows for the first time the brain anatomy in microscopic detail — at a spatial resolution of 20 microns, smaller than the size of one fine strand of hair — exceeding that of existing reference brains presently in the public domain.

The new tool is made freely available to the broader scientific community to advance the field of neuroscience.

Researchers from Germany and Canada, who collaborated on the ultra-high resolution brain model, present their work in the 21 June issue of the journal Science.

“The authors pushed the limits of current technology,” said Science‘s senior editor Peter Stern about the international scientific effort. “Such spatial resolution exceeds that of presently available reference brains by a factor of 50 in each of the three spatial dimensions.”

Better resolution than MRI

Illustration of tissue and image processing. (A) Photographs of the fixed brain. Lateral left (top), lateral right (middle), and dorsal (bottom) views. (B) Magnetic resonance image (coronal view) and (C) 3D-reconstructed MRI volume of the fixed brain. (D) Histological sectioning. (E) Blockface image of a section (pseudocolored) resting on the mounting grid that served for alignment of the blockface images. (F) Series of blockface images. (G) Cell body–stained histological sections with the region of interest denoted by a red box. This area is shown with higher magnification in (H). (I) Series of histological images, which were 3D-reconstructed using the blockface images (F) and the MRI (C). (Credit: K. Amunts et al./Science)

The sophisticated modern image processing methods reveal an unprecedented look at the very fine details of the human brain’s microstructure, or cellular level. The anatomical tool will allow for three-dimensional cytoarchitectonic mapping of the human brain and serve as an atlas for small cellular circuit data, or single layers or sublayers of the cerebral cortex, explained the researchers.

Until recently, reference brains did not probe further than the macroscopic, or visible, components of the brain. Now, the BigBrain provides a resolution much finer than the typical 1 mm resolution from MRI studies.

The project “has been a tour-de-force to assemble images of over 7,400 individual histological sections, each with its own distortions, rips and tears, into a coherent 3-D volume,” said senior author Dr. Alan Evans, a professor at the Montreal Neurological Institute at McGill University in Montreal, Canada. “This dataset allows for the first time a 3-D exploration of human cytoarchitectural anatomy.”

Thin sections of a 65-year-old human female brain, which was embedded in paraffin wax, were cut with a special large-scale tool called a microtome. Then, the 20-micrometer thick histological sections were mounted on slides, stained to detect cell structures and finally digitized with a high-resolution flatbed scanner so researchers could reconstruct the high-resolution 3-D brain model. It took approximately 1,000 hours to collect the data. The resulting images reveal differences in the laminar pattern between brain areas.

Part of the European Human Brain Project

The new reference brain, which is part of the European Human Brain Project, serves as a powerful tool to facilitate neuroscience research and “redefines traditional maps from the beginning of the 20th century,” explained lead author Dr. Katrin Amunts from the Research Centre Jülich and director of the Cecile and Oskar Vogt Institute for Brain Research at the Heinrich Heine University Düsseldorf in Germany.

“The famous cytoarchitectural atlases of the early 1900’s were simplified drawings of a brain and were based on pure visual analysis of cellular organization patterns,” added Dr. Amunts.

Because of the sheer volume of this dataset, the researchers say that there will be a push by those who want to use it to develop new and valuable tools for visualization, data management and analysis.

“We plan to repeat this process in a sample of brains so that we can quantify cytoarchitectural variability,” said Dr. Evans. “We will also integrate this dataset with high-resolution maps of white matter connectivity in post-mortem brains. This will allow us to explore the relationship between cortical microanatomy and fiber connectivity,” said Dr. Amunts.

“We are planning to integrate our receptor data of the human brain in the reference frame provided by the BigBrain,” continued senior co-author Dr. Karl Zilles, who is senior professor of the Jülich Aachen Research Alliance and former director of the Cecile and Oskar Vogt Institute for Brain Research at the Heinrich Heine University Düsseldorf in Germany.

“We will also transfer high-resolution maps of quantitative data on the regional and laminar distribution of native receptor complexes to the BigBrain. This will allow us to explore the relationship between cortical microanatomy and key molecules of neurotransmission.”

The fine-grained anatomical resolution will allow scientists to gain insights into the neurobiological basis of cognition, language, emotions and other processes, according to the study. The researchers also stated that they plan to extract measurements of cortical thickness to gain insights into understanding aging and neurodegenerative disorders; create cortical thickness maps to compare data from in vivo imaging; integrate gene expression data from the Allen Institute; and generate a brain model with a resolution of 1 micron to capture details of single cell morphology.

Public access of the BigBrain dataset will be provided through the CBRAIN Portal with free registration, stated the researchers.