BRAIN will attempt to create a functional model of the brain - a "connectome" - mapping its billions of neuronal connections and firing patterns. This would enable scientists to create both a "static" and "active" model
of the brain, mapping the physical location and connections of these neurons, as well as how they work and fire together between and within different
regions. At the moment, we have small snap-shots into some of these circuits, exposing the function of different brain areas and how these cells
communicate, but on only a fraction of the scale of the entire brain. This process would first be done on much smaller models, such as a fruit fly and a
mouse, before working up to the complexities of a human brain version.
BRAIN proposes to create this model by measuring the activity of every single neuron in a circuit. At the moment, this is done using deep brain techniques,
a highly invasive process that involves opening up the skull to implant electrodes onto individual cells to read and record their outputs. Understandably,
this is only done in patients already undergoing brain surgery, and is a slow and expensive process. Thus, the first task of BRAIN would be to develop
better techniques to acquire this information. Research into this field is already underway, and exciting proposals have included nanoparticles and lasers that
could measure electrical outputs from these cells less invasively, or even using DNA to map neural connections.
Neither project has directly acknowledged the other, but it is thought that the recent announcement of the U.S. proposal is a response to the initial
European scheme launched earlier this year. And while there are distinct differences between the two initiatives in how they will acquire and store the raw
information, as well as how they plan to build their subsequent models, the two projects overlap significantly. Both have the potential to better
illuminate how exactly the brain works, and each ultimately hopes to provide us with a clearer picture of not only normal brain functioning, but also what
happens when these processes are disrupted. Scientists and doctors could then use computer models to simulate dysfunction involved in neurological or
psychiatric disorders, such as Alzheimer's or schizophrenia. This would also open up possibilities for investigating better treatment options, as well as
drastically cutting down on the expense and risk currently involved in clinical drug trials for psychiatric and neurological disorders.
Dr. Francis Collins, NIH director, noted that such disorders "collectively affect 100 million Americans and cost us $500 billion each year in terms of health-care costs."
However, there is a
long list of obstacles
these projects must overcome before we get too excited, not the least of which are the 100,000,000,000,000 connections that need to be measured and
modeled. That's over one million times as many neurons as there were genes to map in the Human Genome Project, the closest approximation to the current
endeavors. Additionally, while there was a clear end to the human genome, the ambition of making a human connectome is both much larger and much less well
defined. Indeed, neither proposal yet has a definitive end-goal, and no one is clear on what the final product will look like.