This website uses cookies primarily for visitor analytics. Certain pages will ask you to fill in contact details to receive additional information. On these pages you have the option of having the site log your details for future visits. Indicating you want the site to remember your details will place a cookie on your device. To view our full cookie policy, please click here. You can also view it at any time by going to our Contact Us page.

'BigBrain' project makes Terabyte map of a human brain

21 June 2013

For the first time ever a complete 3D digital map of a post mortem human brain will be available online for neuroscientists and the general public.

The project involved slicing up the brain of a once healthy 65-year-old woman into over 7,000 segments, each thinner than a human hair, and then digitising the findings (photo: Amunts, Zilles, Evan et al)

The new ultra-detailed model, consisting of a terabyte of data, is part of the European Human Brain Project, created in a joint effort by Canadian and German neuroscientists.

Writing in the IEEE Spectrum blog, TechTalk, Mia Feldman says that with a resolution of 20 micrometers it’s the only model yet to go beyond the macroscopic level. At this degree of resolution, she adds, cells 20 micrometres in diameter are visible.

Although individual smaller cells can’t be seen, it’s possible to identify and analyse the distribution of cells into cortical areas and sub-layers. Previous brain mapping efforts had resolutions one-fiftieth as fine. 

“The whole point of such a modelling project is that you can then start to simulate what the brain does in normal development in children or in degeneration,” says Dr Alan Evans, a professor of biomedical engineering at the McGill University, in Montreal. “If you wanted to look to Alzheimer’s Disease, you can examine how that brain might perform computationally in a computational model if you remove certain key structures or key connections.” 

Collecting images for the project involved slicing up the brain of a once healthy 65-year-old woman into over 7,000 segments, each thinner than a human hair, and then digitising the findings.

This was an especially challenging task, because, once digitised, ruptures created in the slicing process had to be detected and then corrected to develop the final model; a task done both by large amount computer analysis and by manually shifting pieces of data to their proper locations. 

BigBrain is just one of many large-scale brain mapping projects including President Obama’s recently proposed BRAIN Initiative, Paul Allen’s Brain Atlas, and the Human Connectome Project.

BigBrain is the only one to provide a complete map of an individual brain. The Human Connectome Project and BRAIN Initiative focus more on brain activity. The latter will map the connections of small groups of neurons. The former compiles thousands of MRI images from 68 volunteers to map activity, look at how individual brains vary, and see which parts of the brain are involved in specific tasks. Paul Allen’s Brain Atlas focuses more on gene expression in the brain.

An in-depth model of a single post mortem brain can’t really say much about brain activity nor can it account for slight variances in the structures of individual brains, says Dr.- Katrin Amunts, a professor of structural functional brain mapping at Aachen University. Think of it as a general model into which data collected from in vivo brains can be put into context. 

The project is “a common basis for scientific discussions because everybody can work with this brain model and we speak about the same basic findings and we can develop new methodical aspects based on these common model of the human brain,” says Dr Karl Zilles, a senior professor at the Jülich-Aachen Research Alliance.

BigBrain pushes the limits of today’s technology, as software doesn’t yet exist to place data from multiple brains into a single model at 20-micrometer resolution. A 1-micrometer model could take up 20 to 22 petabytes of data, an amount that no computer today would be able to process, according to Amunts.

Mia Feldman's article first appeared in the IEEE Spectrum TechTalk blog on June 20 2013.


Print this page | E-mail this page

Hammond White Paper