A new analysis of the brain’s intrinsic communication networks suggests a layered organization behind human intelligence. The research, published in NeuroImage, indicates that while efficient information flow within certain brain systems is associated with higher cognitive ability, excessive communication from networks involved in internal thought may hinder performance. These findings paint a more nuanced picture of how the brain’s architecture supports both general and specific intellectual skills.
The human brain is an extraordinarily complex network, and understanding how its structure and function give rise to cognitive abilities like intelligence remains a central challenge in neuroscience. Intelligence is not a single entity; it is often described as a hierarchy, with a general factor influencing broad abilities such as problem-solving, verbal understanding, and processing speed.
Researchers have long sought to identify the neural signatures that correspond to these different layers of cognition. A team of scientists from the G. D’Annunzio University of Chieti-Pescara in Italy designed a study to investigate this relationship by examining the brain’s functional connectivity, which maps the synchronized activity between different brain regions.
The researchers conceptualized the brain as a collection of specialized modules, or communities of brain regions that communicate intensely with one another. To get a comprehensive view, they analyzed this modular structure at multiple scales, or resolutions.
A low-resolution view reveals large, sprawling networks, while a high-resolution view pinpoints smaller, more specialized circuits. They hypothesized that this multi-scale brain organization would mirror the hierarchical structure of intelligence itself, with general abilities relating to large-scale networks and specific skills relating to finer, more localized modules.
To test this, the team conducted three separate studies, each using a different set of cognitive assessments and participants. The first study involved 33 university students who underwent resting-state functional magnetic resonance imaging (fMRI), a technique that measures brain activity while a person is at rest.
This provides a snapshot of the brain’s baseline communication patterns. Participants also completed the Wechsler Adult Intelligence Scale (WAIS-IV), a comprehensive test that yields a full-scale intelligence quotient (IQ) score as well as indices for perceptual reasoning, processing speed, working memory, and verbal comprehension.
The analysis focused on two key network properties. The first was nodal efficiency, a measure of how effectively a brain region can exchange information with the rest of the network. The second was the participation coefficient, which quantifies how much a region communicates across different modules, acting as a bridge between specialized communities.
The results showed that higher full-scale IQ scores were associated with greater nodal efficiency in a widespread “task-positive” network, which includes regions in the frontal, parietal, and occipital lobes that are typically active during goal-oriented tasks. This connection was most evident at a low-resolution scale, suggesting that general intelligence relies on the efficient functioning of a large, integrated brain system.
When the researchers examined the brain’s network at a higher resolution, they found that specific cognitive abilities were linked to the efficiency of smaller, more specialized modules. Better processing speed, for example, was associated with greater efficiency within the visual network. Superior perceptual reasoning was connected to efficiency in sensorimotor and fronto-parietal networks, which are involved in integrating sensory information and executing cognitive control. These findings support the idea that the brain encodes intelligence hierarchically, with broad networks supporting general ability and distinct subsystems supporting specialized skills.
The study also produced another consistent pattern. Higher scores on full-scale IQ, perceptual reasoning, and processing speed were all associated with lower participation coefficients in the default mode and subcortical networks. These networks are typically active during inward-focused thought, such as daydreaming or recalling memories. The negative association suggests that better cognitive performance is linked to less cross-talk from these internally oriented systems. This may reflect an ability to shield task-focused processing from irrelevant internal distractions.
The second study aimed to replicate and extend these findings using a larger, publicly available dataset from the Human Connectome Project, which included 500 participants. These individuals completed a different set of cognitive tests. The researchers first grouped the test scores into four broad cognitive categories: Language, Executive Functions, Fluid Reasoning, and Memory. They then performed the same brain network analysis on the participants’ resting-state fMRI scans.
The results from this larger dataset were consistent with the first study. Higher scores in fluid reasoning, which involves abstract problem-solving, were linked to greater nodal efficiency in sensorimotor and fronto-parietal networks. Better memory performance was associated with higher efficiency in the visual network.
The analysis again revealed that better performance in language and memory was linked to a lower degree of cross-network communication from the default mode network. The convergence of these findings across two independent datasets and different cognitive tests strengthens the conclusions.
For the third study, the research team recruited another group of 37 university students. These participants completed the Raven’s Progressive Matrices, a well-known test designed specifically to measure fluid reasoning. The analysis of their brain scans confirmed the patterns observed in the previous two studies. Higher fluid reasoning scores were associated with greater efficiency in sensorimotor and fronto-parietal executive networks. At the same time, better performance was linked to reduced cross-network communication involving subcortical, limbic, and fronto-parietal modules.
Taken together, the three studies build a cohesive account of how the brain’s intrinsic network architecture relates to intelligence. The efficiency of information processing within key sensory and executive networks appears to be a positive contributor to cognitive ability. Conversely, the degree to which internally focused networks integrate with the rest of the brain seems to have a negative influence on performance, possibly by creating interference. The multi-resolution approach also provides novel evidence for a parallel between the brain’s layered network organization and the hierarchical structure of human intelligence.
The study is not without its limitations. Because three independent groups of participants were used, a direct, one-to-one comparison of the results from the different cognitive tests was not possible. Another point of consideration is the use of resting-state fMRI, which measures the brain’s baseline activity rather than its activity during an active cognitive task.
Future research could explore how these brain networks reconfigure when individuals are actively engaged in problem-solving. Such studies could help clarify the dynamic interplay between network efficiency and segregation during real-time cognitive processing. Future work might also investigate whether interventions aimed at modulating these network properties could lead to changes in cognitive performance.
The study, “Intrinsic brain mapping of cognitive abilities: A multiple-dataset study on intelligence and its components,” was authored by Simone Di Plinio, Mauro Gianni Perrucci, Grazia Ferrara, Maria Rita Sergi, Marco Tommasi, Mariavittoria Martino, Aristide Saggino, and Sjoerd JH Ebisch.
