German scientists are to utilise the power of hardware manufacturer Nvidia's graphical processing units (GPU) to discover the secrets of the brain.

Scientists at the research centre Forschungszenstrum Julich, which houses one of the most powerful supercomputing resources in Europe, are hoping that the GPUs will be a catalyst for the advanced neurological research.

The Nvidia Tesla GPUs are to accelerate the reconstruction of sections of the brain by around 50 times, creating high-definition, precisely accurate and realistic images never before seen by neuroscientists.

It will allow researchers at the Julich Institute of Neuroscience and Medicine (INM-1) to look at the brain and its functions and interconnections, with the hope that it will lead to discoveries to help understand and cure afflictions such as autism, multiple sclerosis and Alzheimer's, as well as other degenerative neurological conditions.

The model is being created through a huge collection of data sets of microscopic tissue structure, magnetic resonance images and representations of the brain using advanced 3D polarised light imaging (3D-PLI).

In order to trace the tracts of the brain in fine detail, very high performance GPUs are required.

"3D-PLI is the only way to achieve highly detailed images of nerve fibers in adult human brains, but reconstructing and rendering them in real time into the world's first micro-atlas of the human brain poses a major computational problem," said Professor Katrin Amunts, director of INM-1.

"Imagine the billions of nerve cells inside the human brain, connected via fibers. This gives you a sense of the complexity and intricacy needed to accurately model the network within the human brain."

In addition, Nvidia is working with the Julich Supercomputing Centre to provide GPU-accelerated scientific research for other areas, including astronomy, astrophysics, material science, particle physics and protein folding.

Together the two organisations are working to open the Nvidia Application Lab to help further the studies.ADNFCR-1220-ID-801390513-ADNFCR
Enjoyed this article? share it!