According to a recent study, the synchronized activity of brain cells, similar to birds flying together, enables us to act intelligently in unfamiliar situations. This research is pioneering in revealing the neurological processes associated with abstraction and inference in the human brain.
A groundbreaking study by Cedars-Sinai researchers suggests that the synchronized activity of brain cells—like birds flying in formation—enables intelligent behavior in new environments. This research, published in the peer-reviewed journal Nature, is the first to explore the neurological mechanisms of abstraction and inference in human cognition.
“Abstraction helps us disregard unnecessary details and concentrate on the essential information needed for our actions, while inference involves using our knowledge to make informed guesses about our surroundings,” explained Ueli Rutishauser, PhD, professor and Board of Governors Chair in Neurosciences at Cedars-Sinai, who also co-authored the study. “Both processes are crucial for learning and reasoning.”
Humans frequently utilize these cognitive skills together to swiftly understand and respond appropriately in new contexts, like an American driver renting a car in London for the first time.
“In the UK, driving involves steering on the right side of the car while traveling on the left side of the road, which is the opposite of what Americans are used to,” Rutishauser noted. “An American driver needs to reverse many learned driving rules, which necessitates abstraction to focus on which side to drive on and inference to avoid heading into oncoming traffic.”
For this study, researchers worked with 17 hospitalized patients who had electrodes implanted in their brains for epilepsy diagnosis. They recorded the activity of thousands of brain cells as participants completed an inference task on a computer.
Analyzing thousands of brain cells required artificial intelligence to identify relevant patterns, enabling researchers to observe the neurons’ coordination during successful inference tasks.
“These form intricate high-dimensional geometrical shapes that we can’t easily visualize on a computer,” stated Stefano Fusi, PhD, a principal investigator at Columbia University’s Zuckerman Mind Brain Behavior Institute and co-author of the study. “However, mathematical methods allow us to represent simplified 3D versions.”
During the experiment, participants saw four images—a person, a monkey, a car, and a watermelon—and were instructed to press a button with their left or right hand in response. They subsequently received feedback indicating whether they were “correct” or “incorrect.”
As the participants repeated the task, they learned the correct responses for each image. Eventually, the rules were reversed without their knowledge, requiring them to give the opposite responses to be correct.
After this change, some participants quickly adjusted to the new rules and inferred the right responses without needing to relearn, demonstrating their inference skills.
Researchers observed notable geometric patterns in the brain activity of these successful participants. Groups of neurons were firing synchronously, reminiscent of birds in formation or a group of fans chanting at an event. The coordinated neuron activity signified that these participants had acquired the conceptual understanding necessary for the task; no such patterns were evident in those who struggled with inference.
“Developing conceptual understanding is vital for effective learning,” said Hristos Courellis, PhD, a Cedars-Sinai researcher and the study’s first author. “Our research highlights a neural foundation for this process known as abstraction in cognitive psychology.”
Some participants initially couldn’t deduce the correct responses through experiential learning alone. For these individuals, researchers provided verbal instructions enabling them to infer the right answers.
“A significant finding was that similar neural patterns surfaced in participants who received verbal guidance as in those who learned through experience,” commented Adam Mamelak, MD, director of the Functional Neurosurgery Program at Cedars-Sinai and co-author of the study. “This discovery reveals that verbal cues can establish neural representations that might typically take longer to learn through experience.”
This study, conducted by Cedars-Sinai and involving data from the University of Toronto, was part of a multi-institutional collaboration funded by the National Institutes of Health’s The Brain Research Through Advancing Innovative Neurotechnologies Initiative, commonly known as The BRAIN Initiative.
“This research offers fresh insights into how our brains enable adaptable learning and task execution in varied conditions,” stated Merav Sabri, PhD, program director for The BRAIN Initiative. “These findings contribute to a growing knowledge base that could inform future interventions for neurological and psychiatric conditions marked by memory and decision-making deficits.”
Interestingly, researchers found that these distinctive brain activity patterns appeared exclusively in the hippocampus, a brain region pivotal for forming new long-term memories.
“Our discovery enhances our understanding of the hippocampus’s role in learning,” Rutishauser added. “This marks the first direct evidence of the human hippocampus’s involvement in acquiring abstract knowledge and inference skills. Many neurological disorders, including Alzheimer’s, OCD, and schizophrenia, affect this area, potentially shedding light on the impaired decision-making observed in these patients.”
