New laboratory to advance research in 3D brain modelling

Canadian-German collaboration will build “the equivalent of Google Earth for the brain”
Alan Evans from The Neuro (Montreal Neurological Institute-Hospital) and Katrin Amunts, from Forschungszentrum Jülich will co-lead the Helmholtz International BigBrain Analytics Learning Laboratory

Launched on June 26, 2020, the Helmholtz International BigBrain Analytics Learning Laboratory (HIBALL) will work to create 3D brain models at an unprecedented cellular-level resolution. Co-led by Katrin Amunts, from Forschungszentrum Jülich (FJZ) in Germany and Alan Evans from The Neuro (Montreal Neurological Institute-Hospital) and McGill University in Canada, the project builds on more than two decades of collaborations between the Amunts and Evans labs. HIBALL is jointly funded for five years by the Helmholtz Association of German Research Centres (€3,000,000 Euro) and McGill University ($4,500,000 CAD). Of McGill University’s funding, Healthy Brains, Healthy Lives (HBHL) has contributed $3,375,000 CAD.

Using the latest artificial intelligence and high-performance computing technologies, HIBALL will build on previous work done by Amunts and Evans on the BigBrain dataset, currently the highest-resolution 3D model of the human brain.

“We will be building the equivalent of Google Earth for the brain and exploring how artificial intelligence can model the functional organization of the brain at an unmatched 3D spatial scale,” Evans explains. “Quite apart from the scientific insights we can expect, HIBALL involves brilliant young faculty and trainees who have come to McGill through HBHL. We are eager to get going on this innovative project with our long-term partners at Jülich.”

The project’s research approach is divided into six scientific themes, each with designated theme leaders:

  • Brain parcellation using AI-based segmentation (Timo Discksheid, FJZ; Alan Evans, McGill)
  • Multimodal data integration into BigBrain (Boris Bernhardt, McGill; Markus Axer, FJZ)
  • Cellular BigBrain (Katrin Amunts, FZJ ; Timo Dickscheid, FZJ)
  • Modular computing architecture (Thomas Lippert, FZJ; Tristan Glatard, Concordia/McGill)
  • The Virtual BigBrain (Alan Evans, McGill; Randy MacIntosh, U. of Toronto)
  • Brain-Inspired Artificial Intelligence (Blake Richards, McGill/MILA; Doina Precup, McGill/MILA)

Evans says that HIBALL touches on many key aspects of HBHL’s mission, with multidisciplinary collaboration playing an important role in the project.

The project will offer opportunities for collaboration on a global scale, and involves several international partner organizations, including the Canadian Institute for Advanced Research (CIFAR), the Virtual Brain, the Montreal Institute for Learning Algorithms (MILA), The Neuro, Western University (BrainsCAN), the Helmholtz Artificial Intelligence Cooperation Unit (HAICU),the Helmholtz Centre Munich (HMGU), and the Joint Cuba China Lab for Neurotechnology and the Cuban Human Brain Mapping Project, both coordinated by Pedro Valdes-Sosa. Across these in stitutions, HIBALL includes faculty members, postdoctoral fellows, students and software developers working on the project in tandem. McGill University faculty members involved in the project include Blake Richards, Doina Precup, Boris Bernhardt, Louis Collins, Tristan Glatard, Bratislav Misic and Pedro Valdes-Sosa.