AI framework decodes how brain regions talk to one another

For decades, neuroscientists have been able to record which parts of the brain show increased electrical activity during a behavior, such as when we form a memory or make a decision. But scientists haven't been able to determine how those regions are communicating with and influencing one another, nor identify which regions serve as the main orchestrators.

A new artificial intelligence framework called Current-Based Decomposition, or CURBD, can now make that distinction. Developed by Harvard Medical School researchers and their colleagues, the tool is described in Neuron.

“No brain region works on its own. They’re all constantly sending signals to each other, shaping each other’s activity,” said Kanaka Rajan, associate professor of neurobiology in the Blavatnik Institute at Harvard Medical School and the study’s co-senior and corresponding author. “CURBD lets us see those conversations like never before—which region is directing which, and how powerfully.”

Sign up for Blog Updates