Smarter with EBRAINS

Schlauer dank EBRAINS — 38 Mio. Euro erhält EBRAINS 2.0 von der Europäischen Kommission bis Ende38 million euros will be given to EBRAINS 2.0 by the European Commission until the end of 2026. | Copyrights: Mareen Fischinger / Forschungszentrum Jülich

The digital user infrastructure EBRAINS, which was born out of the Human Brain Project, has given researchers worldwide access to brain atlases, brain models, around 200 tools, and countless data. In the three-year EU project EBRAINS 2.0, the platform is now being developed further, with Jülich playing a major role. Scientists are supported through the platform if they want to use supercomputers, store large amounts of data, and analyse these data. But clinical applications are also benefitting.

LOOKING INTO THE BRAIN: a selection of different imaging techniques in the EBRAINS brain atlas

3D Polarized Light Imaging

The 3D PLI microscopy method developed at Jülich uses polarized light to show the distribution of nerve fibres in the brain, where they connect around 86 billion nerve cells. You can’t tell from the colourful pictures, but a single brain slice contains over 250,000 individual microscopic images.

Neurons

In order to see them under the microscope and determine their arrangement and number, specially prepared brain slices placed on glass are typically required. Such images are important for mapping areas of the brain. Their boundaries can be recognized here up close, for example, by the different density of neurons.

Probability maps

The data from several brains are linked in a map and show where a certain area extends to in most of the brains examined (warm colours) and where it only extends to in just a few brains (cold colours).

BigBrain

With this digital model developed at Jülich and in Canada, researchers can take a close look at the human brain in high resolution: to an accuracy of 20 micrometres, which is as thin as a hair. INM is currently working on a digital 1-micrometre brain. This should make it possible to record all 86 billion neurons in the brain for the first time and gain an even better understanding of neural networks.

Receptors for neurotransmitters

This view of the EBRAINS brain atlas shows which receptors for certain neurotransmitters are present in which areas of the brain and how frequently. If researchers find out more about this, they can better understand information processing in the brain and better classify findings from PET imaging. In the long term, this will allow more targeted treatment methods to be developed.

Three EBRAINS highlights:

EBRAINS brain atlas

The interactive EBRAINS human brain atlas is at the heart of the new digital infrastructure, and is a huge asset for brain research. It was made possible through FZJ contributions such as the Julich Brain Atlas, the high-resolution BigBrain model, and many other findings from Jülich and Europe. Serving as a kind of Google Maps for the brain, very different aspects of brain organization – such as neurons, brain areas, and nerve fibre connections – can be captured using a wide variety of imaging techniques, and linked to aspects such as ageing processes. All of this increasingly benefits from artificial intelligence methods, which are taking the brain atlas to a new level.

EBRAINS data

The research data made available through EBRAINS are particularly valuable. These data troves, which grow each day, stem from a wide variety of experiments across Europe. In EBRAINS, they can be linked with many other functions, for example to record and visualize cell numbers in the atlas or to develop new models and simulations. The data can also be read out and optimally reused in other software, as they are labeled in detail and processed according to the FAIR principles. EBRAINS has also made it technically possible to share really large data for the first time.

EBRAINS simulations

Another highlight of EBRAINS are tools that can be used to simulate the brain in various ways. They go into particular detail on the molecular level. Through EBRAINS, Jülich scientists such as Prof. Paolo Carloni and Prof. Giulia Rossetti (IAS-5) use a lot of computing power to simulate exactly how certain molecules change their orientation when another molecule approaches. In order to simulate the activity of whole cells as accurately as possible, Sandra Diaz and Wouter Klijn (JSC) in turn use the specially developed Arbor tool, which will be one of the first applications to run on the new exascale supercomputer JUPITER. Prof. Markus Diesmann (IAS-6) also developed the network simulator NEST at Jülich to simulate large networks consisting of millions of neurons. Simulations that attempt to recreate the electrical activity of brains are also possible. The Simulation and Data Lab Neuroscience also actively supports users in transferring the simulations to the large JSC computers.

Originally published in the employee magazine "intern" of Forschungszentrum Jülich.
Author: Hanno Schiffer