Neuroscientists gain understanding through reconstructed brain simulation

The complexities of the brain and how it works have stumped researchers for ages. Neuroscientists have searched for any possible method in many different fields of study to learn just a little more about what takes place inside our heads.

Researchers at the Ecole Polytechnique Federale de Lausanne in Sweden published an animal study on October 8 in the journal, Cell, which explained how they deconstructed a piece of a young rat brain to simulate on a supercomputer. The research was a part of the Blue Brain Project, an initiative under the Human Brain Project that focuses on computational simulation of the brain.

“The breadth of experimental data, both anatomical and physiological, and our digital reconstruction, is consistent yet unprecedented in the neuroscience community,” said Eilif Muller, a computational neuroscientist at the Blue Brain Project. “It has taken us 10 years to get to this point.”

The reason these findings are important, and should lead to further simulation studies, is due to the effectiveness in mapping different types of neurons in a sample of the brain and distinguishing the properties of those neurons. These scientists performed thousands of experiments during the course of a decade to determine ideal situations in modeling and representing these cells.

The project team was able to do this research by taking the extracted portion of a newborn rat brain and model it on IBM supercomputers. The choice to use supercomputers was due to the high amounts of data from a small piece of brain that must be processed on the computer.

Muller said if each neuron took a second to be modeled - there were over 31 thousand cells in the sample - it would take almost 10 hours to fully process all of that data. With all that time spent processing a model, the supercomputer becomess ideal because it is energy efficient and reliable.

Alongside the supercomputer, the Blue Brain Project used NEURON, “a well-known and very widely used simulation package in the field of neuroscience,” Muller said. It is open source, meaning it is open to anyone in the field. For this study, the software had to be modified for use on the supercomputer.

After the data was processed on the system, NEURON was used to help parse the information and categorize connections between brain cells. One major finding in this study is that calcium ions are key components to connectivity between synapses and allow higher electrical activity in neurons.

As soon as these findings were published, this information was immediately made available online by the opening of a new website with images, videos, and discussions about the simulation. The neocortical microcircuit collaboration portal, or NMC portal, “provides models and tools for researchers to download and use for their own research,” Muller said. The Blue Brain Project aims to have more collaborators and further studies from other research groups in order to validate this one.

Muller said potential applications of this modeling program could be beneficial in diagnostics and finding treatments in humans, especially in areas such as depression, epilepsy, and migraines.

“I see a bright future in [computational] neuroscience research to enable simulation-based studies of brain diseases, and through such studies to inspire new treatments,” he said. “Like any medical application, the road of treatment development is long and unpredictable, but modeling approaches to treatment development do hold the potential to complement current empirical approaches.”