Sept. 16, 2022

HBP and EBRAINS are present at the Bernstein Network conference on computational neuroscience


The Human Brain Project (HBP) and EBRAINS are present at the Bernstein Conference 2022, which takes place in Berlin from 13 to 17 September. EBRAINS is one of the sponsors of the conference, held annually by the Bernstein Network to bring together the international computational neuroscience community for scientific exchange.

Researchers attending the event have a chance to visit the HBP and EBRAINS booth and learn about how their research infrastructure supports brain science in Europe. Visitors who want to know more about the cutting-edge tools and resources available for the neuroscience community are welcome to stop by the booth at the poster area.

“At the moment, Berlin is the place in Europe for interdisciplinary discussion on computational neuroscience,” said Geraldine Rauch, president of the Technische Universität Berlin, which is hosting the conference, as she welcomed participants in the opening session on Wednesday.
 
HBP researchers are present in several activities of the programme, including workshops, lectures, symposia and poster sessions. In one of the workshops, “Large-scale cortical networks: from dynamics to cognition”, theoretical neuroscientist researcher Hiba Sheheitli, Aix-Marseille University, spoke about her group’s efforts towards the development of digital twins of the brain, aiming for applications in personalised medicine.
 
“The Virtual Brain is an open-source neuroinformatics platform that makes it easy for experts, but also to non-experts, to build brain models,” explained Sheheitli. The goal of these models is not only to have predictive powers but also to help disentangle brain mechanisms.
 
Computational neuroscientist Sacha van Albada, Forschungszentrum Jülich, talked about cellular-resolution simulations of macaque and human cortices and what can be learned from modelling the resting state. She also highlighted recent research from her group that reviews how connectivity is specified for neural network models and proposes unified descriptions and depictions, in order to make reproducibility easier.

In a contributed talk, Paul Heider, PhD student at the University of Bern, spoke about a newly developed learning theory that shows how the brain can learn extremely fast sequences of sensory stimuli. He presented the study “Latent Equilibrium: A unified learning theory for arbitrarily fast computation with arbitrarily slow neurons”, by the groups of HBP researchers Mihai Petrovici and Walter Senn, which not only helps to understand complex neural networks and learning in the brain, but is also relevant to research into artificial intelligence and neuromorphic hardware.

The Bernstein Network is an international research network in the field of computational neuroscience launched in 2004 in Germany. The field of computational neuroscience brings together experimental approaches in neurobiology with theoretical models and computer simulations.