Jülich Supercomputing Center
Unlocking the Secrets of the Human Brain

Background
Germany’s Forschungszentrum Jülich is a large interdisciplinary research center focused on developing solutions that will have a direct impact on society in the areas of health, energy and information technology. Forschungszentrum Jülich hosts the Jülich Supercomputing Centre, one of Europe’s most powerful supercomputing resources, enabling scientists and engineers around the world to solve highly challenging and complex problems across a broad range of fields and disciplines.

Neuroscience is among the increasingly important research focus areas at Forschungszentrum Jülich, and researchers at Jülich are undertaking a new approach to advanced neuroscience research aimed at combating a number of neurological diseases.

Challenge
Autism, multiple sclerosis and Alzheimer’s disease are among the most debilitating neurological diseases affecting millions of individuals worldwide. While much progress has been made to advance understanding of these serious ailments, researchers continue to search for better treatments and cures.

Unlocking the mysteries behind these neurological disorders requires an intimate understanding of the structure and inner-workings of the human brain, its network of 100 billion neurons, and the short- and long-distance interconnections of each of these neurons throughout all the different sections of the brain.

One way to study the structure and function of the brain is to develop sophisticated computer models. However, creating a realistic model with the required precise level of detail

is an extremely complex and intricate process requiring tremendous computational resources.

Solution
Researchers from the Jülich Institute of Neuroscience and Medicine (INM-1) took a unique approach to creating an accurate highly detailed computer model of the human brain by combining a vast collection of data sets. These included images of histologic sections (microscopic tissue) and magnetic resonance images.

They also utilized images from an advanced 3D polarized light imaging (3D-PLI) technique developed at INM-1, which provides highly detailed information about the direction and the inclination of nerve fiber tracts in the brain at a sub-millimeter resolution.

INM-1 leveraged Jülich’s GPU-accelerated supercomputers to build the computer model, including the 206-node Jülich Dedicated GPU Environment (JuDGE) system, which is equipped with NVIDIA® Tesla® GPUs.

GPUs enabled researchers to accelerate by as much as 50x the reconstruction of the histologic brain sections, and to render a high-definition, structurally accurate and realistic model of the human brain.

Impact
Once fully developed, the 3D model of the brain developed by Jülich will give researchers a previously unattainable level of visibility into brain architecture, function and interconnections, with levels of detail never before available to neuroscientists.

It is expected to shed new light on the causes of some of the most serious and debilitating neurological diseases, which may ultimately lead to new and more effective treatments.

Jülich plans to leverage key learnings from its neuroscience  research as a blueprint for other advanced GPU-accelerated research across a range of other scientific fields.