Learning occurs quicker than thought, according to brain imaging

Researchers supported by the U.S. National Science Foundation have provided a new understanding of how and where learning occurs in the brain. The two-part finding has implications for understanding and treating neurodegenerative diseases like Alzheimer's and other dementias, which impact more than 7 million people in the United States and account for $384 billion in health and long-term care costs, as well as for enhancing neural networks.

"Identifying how the brain actually forms new connections and learns is a question at the frontier of neuroscience," said Paul Forlano, program officer in the NSF Directorate for Biological Sciences. "Knowing that influences our understanding of how we interact with our environment and pick up on and respond to cues, which opens the door to a range of new fundamental and applied research."

The researchers, led by Kishore Kuchibhotla, assistant professor at Johns Hopkins University, used brain imaging to determine when mice learned a new skill. The imaging reinforced previous work, showing that mice learned quickly and that those that continued to make errors weren't still learning; they were experimenting. The difference between mistakes and testing the rules was evident in changes in the neural activity that the researchers saw in the mice.

Kuchibhotla said the distinction between the brain dynamics in learning and the dynamics involved in using that skill could be mimicked in having a memory and being able to retrieve it. If a similar paradigm exists in humans, it could alter how scientists approach questions about neurodegenerative diseases like dementia and Alzheimer's, as well as how those conditions are treated.

The other surprising outcome of the research was that learning occurs in the sensory cortex, a region of the brain generally associated with interpreting (for example, "this stove is hot") but not having input on behavior (like removing one's hand from the stove). The team argues that the cortex is better described as a sensory-enriched associative cortex, wherein sensory and associative learning functions are intrinsically intermingled. The parallel functions and how the brain accomplishes them could lead to advances in how neural networks, which are modelled on the brain, process information.