Thursday, May 29, 2014

BRAIN Project Proposals are Getting a Reality Check

Neuroscientists were over the moon in April 2013 when President Barack Obama announced a bold new initiative to study the human brain in action. But in their heady excitement, some may have forgotten to check the math in their first proposals. At least, that's the contention of a group of physicists, engineers, and neuroscientists meeting this week in Arlington, Virginia, to discuss which ideas are likely to succeed and which may fall flat.

Key to the success of the roughly $100 million Brain Research through Advancing Innovative Neurotechnologies (BRAIN) Initiative is crafting new tools or methods to measure neural activity either from inside or outside the brain. Unfortunately, some ideas “violated either a physical law or some very significant engineering constraint or biological constraint,” says neurophysicist Partha Mitra of Cold Spring Harbor Laboratory in New York, who helped organize the meeting, sponsored by the National Science Foundation.

The goal is to have a realistic discussion of what the physical limits are, he says, so “scientists who want to make devices will not make crazy proposals,” or, “if a proposal is crazy, one could recognize it as such” and look for other ways to make the idea work.

One such “fanciful” idea is to build nanosized radios that could snuggle up to individual neurons to record and transmit information about their activity, says physicist Peter Littlewood, director of Argonne National Laboratory in Lemont, Illinois. But any radio small enough to be injected into the brain without causing significant harm would not be able to transmit any information out through tissue and bone, he says. Make the devices any more powerful, he adds, and they'd likely cook the surrounding brain. Another aspiration that is likely doomed is to get microscopes that probe the brain with pulses of light to penetrate much further than they already do, Mitra says. A little more than 1 mm is possible, he adds, but even 1 cm is “out of the question, since the signal to background [noise] ratio decreases exponentially with depth.”

But physicists and engineers shouldn't simply shoot down outlandish proposals—or gripe about the intrinsic messiness of the brain's biology. They should model themselves as “fancy technicians” who can help develop revolutionary tools, Littlewood says. There are precedents for such collaboration, he notes: He, Mitra, and their colleagues at Bell Labs, for example, helped develop functional magnetic resonance imaging in the 1990s.

No comments: