Pitt, CMU researchers shed light on how learning works

You hear it all the time — if you speak Spanish, French is easier to learn. If you skateboard, snowboarding isn't as difficult. Anything you want to learn is a little bit easier if you know how to do a similar thing.

Now researchers know more about what happens in the brain when that kind of learning occurs.

Scientists at the University of Pittsburgh and Carnegie Mellon University have shown for the first time how groups of neurons work together when learning to do one thing, then when learning something similar. Their findings appear in this week's issue of Nature.

“What's easier or harder for you to learn depends on what you already know. Whether you talk about language or sports, it's going to be easier for a tennis player to learn to play squash, rather than learn to play a new melody on the piano,” said Byron Yu, an assistant professor of electrical and computer engineering at Carnegie Mellon University, and a main author of the study.

To show this, researchers used a brain-computer interface, which allows a subject to move something via computer just by thinking about it. A similar tool is in clinical trials to help people with paralysis regain motor function.

The animals in the study had to use their minds to move the cursor toward a target. The researchers had programmed the cursor to move in a certain way, so when the animal's thought pattern matched the program, regardless of what the animal was actually thinking, the cursor moved to the target and the animal got a reward.

It took the animal a few hours to get it right, said Aaron Batista, an assistant professor of bioengineering at Pitt and the other main author of the study. And then, each time, as the animal thought to move the cursor to the target to get the reward, similar patterns of neurons fired, even if whatever the animal was thinking about might have been different, said Batista.

“Some types of patterns were achievable in a few hours. Others were not,” said Batista. “It's completely reasonable, but no one has shown this in the brain.”

From their findings, the researchers want to try to expand what humans can learn.

Lee Miller, a professor of neuroscience at Northwestern University, said the research could have an immediate impact on how humans use their brains to control devices such as prosthetics, but it also could reveal why children seem to learn different things so easily. Miller, who was not involved in the research, said it's possible the study could inform how, after a stroke, newfound flexibility in the brain can offset stroke damage.

“It gives us the ability to learn something about how the brain learns and its limitations. We don't create things completely from scratch,” Miller said of neuron activity patterns.

The Pittsburgh-based researchers and their work will be part of Brain Hub, a $75 million, five-year neuroscience project CMU officials announced Tuesday. The National Institutes of Health, National Science Foundation and the Burroughs Wellcome Fund contributed funding.

Megha Satyanarayana is a staff writer at Trib Total Media. She can be reached at 412-320-7991 or megha@tribweb.com.