Network Neuroscience Theory – The Best Predictor of Intelligence

Researchers have been working for years to understand the relationship between brain structure, functional connectivity and intelligence. Recent research provides the most comprehensive understanding to date of how different regions of the brain and neural networks contribute to a person’s ability to solve problems in a variety of contexts, a trait known as general intelligence.

The researchers recently published their findings in a journal. human brain mapping.

The study, led by Aron Barbey, professor of psychology, bioengineering, and neuroscience at the University of Illinois at Urbana-Champaign, and first author Evan Anderson, a research fellow at Ball Aerospace and Technologies Corp. at the Air Force Research Laboratory, explores how the brain leads to intelligence. A “connectome-based predictive modeling” technique to evaluate the five theories.

“To understand the amazing cognitive abilities that underlie intelligence, neuroscientists look at the biological basis of the brain,” said Barbey. “Modern theories attempt to explain how our ability to solve problems is activated by information-processing structures in the brain.”

Anderson said that understanding the biology of these cognitive abilities requires “characterizing how individual differences in intelligence and problem-solving relate to the underlying structure and neural mechanisms of brain networks.”

Historically, intelligence theories have focused on localized brain regions, such as the prefrontal cortex, that play important roles in cognitive processes such as planning, problem solving, and decision making. While more recent theories emphasize specific brain networks, other theories examine how different networks overlap and interact, Barbey said. He and Anderson tested these established theories against their own “network neuroscience theory,” which posits that intelligence emerges from the overall structure of the brain, which contains both strong and weak connections.

“Strong connections involve highly connected information-processing hubs that are built as we learn about the world and become adept at solving familiar problems,” Anderson said. “Weak connections have fewer neural connections, but allow for flexibility and adaptive problem solving.” Together, these connections “provide the network architecture needed to solve the many problems we face in life.”

To test their idea, the team recruited a demographically diverse pool of 297 undergraduate students, first asking each participant to take a battery of comprehensive tests designed to measure their problem-solving abilities and adaptability in a variety of situations. Similarly, various tests are routinely used to measure general intelligence, Barbey said.

Researchers next collected each participant’s resting-state functional MRI scans.

“One of the really interesting properties of the human brain is how it implements a rich set of networks that are active even when we are at rest,” said Barbey. “These networks create the biological infrastructure of the mind and are thought to be intrinsic properties of the brain.”

These include frontal lobe networks that enable cognitive control and goal-directed decision-making. a network of dorsal attention that aids visual and spatial perception; and a salient network that pays attention to the most relevant stimuli. Previous research has shown that activity in these and other networks “reliably predicts our cognitive skills and abilities,” Barbey said, when people are awake but not engaged in tasks or paying attention to external events.

Using cognitive tests and fMRI data, researchers were able to evaluate which theory best predicted how participants performed on intelligence tests.

“We can systematically examine how well a theory predicts general intelligence based on the connectivity of the brain regions or networks that the theory entails,” Anderson said. “This approach allowed us to directly compare the evidence for neuroscience predictions made by current theories.”

Researchers have found that considering features of the whole brain is the most accurate predictor of a person’s problem-solving aptitude and adaptability. This was true even when considering the number of brain regions included in the analysis.

Other theories also predicted intelligence, but network neuroscience theories in many respects outperformed theories limited to local brain regions or networks, the researchers said.

The findings reveal that the brain’s “global information processing” is fundamental to how well individuals overcome cognitive challenges, Barbey said.

“Intelligence, rather than stemming from a specific region or network, appears to emerge from the global architecture of the brain and reflects the efficiency and flexibility of system-wide network functions,” he said.

Reference: Evan D. Anderson and Aron K. Barbey, 20 Dec. 2022, “Investigating the Cognitive Neuroscience Theory of Human Intelligence: A Connectome-Based Predictive Modeling Approach”, human brain mapping.
DOI: 10.1002/hbm.26164

This study was funded by the Office of the Director of National Intelligence, Intelligence Advanced Research Project Activity and the Department of Defense, Defense Advanced Research Project Activity.