Fluid intelligence is arguably the defining feature of human cognition. It predicts educational and professional success, social mobility, health, and longevity. It also correlates with many cognitive abilities such as memory. Yet the nature of its relationship with the brain is a contentious topic. New research, led by University College London and the UK’s National Hospital for Neurology and Neurosurgery, shows that a set of predominantly right frontal regions of the brain is critical to the high-level functions involved in fluid intelligence.
Network modeling of fluid intelligence. Image credit: Cipolotti et al., doi: 10.1093/brain/awac304.
Fluid intelligence refers to the ability to solve challenging novel problems when prior learning or accumulated experience are of limited use.
It ranks amongst the most important features of cognition, correlates with many cognitive abilities (e.g. memory), and predicts educational and professional success, social mobility, health and longevity.
It is thought to be a key mental capacity involved in ‘active thinking,’ fluid intelligence declines dramatically in various types of dementia and reflects the degree of executive impairment in older patients with frontal involvement.
Despite the importance of fluid intelligence in defining human behavior, it remains contentious whether this is a single or a cluster of cognitive abilities and the nature of its relationship with the brain.
To establish which parts of the brain are necessary for a certain ability, researchers must study patients in whom that part is either missing or damaged.
Such ‘lesion-deficit mapping’ studies are difficult to conduct owing to the challenge of identifying and testing patients with focal brain injury.
Consequently, previous studies have mainly used functional imaging (fMRI) techniques — which can be misleading.
In the new study, University College London’s Professor Lisa Cipolotti and colleagues investigated 227 patients who had suffered either a brain tumor or stroke to specific parts of the brain, using the Raven Advanced Progressive Matrices (APM): the best-established test of fluid intelligence.
The test contains multiple choice visual pattern problems of increasing difficulty.
Each problem presents an incomplete pattern of geometric figures and requires selection of the missing piece from a set of multiple possible choices.
The researchers then introduced a novel ‘lesion-deficit mapping’ approach to disentangle the intricate anatomical patterns of common forms of brain injury, such as stroke.
Their approach treated the relations between brain regions as a mathematical network whose connections describe the tendency of regions to be affected together, either because of the disease process or in reflection of common cognitive ability.
This enabled the authors to disentangle the brain map of cognitive abilities from the patterns of damage — allowing them to map the different parts of the brain and determine which patients did worse in the fluid intelligence task according to their injuries.
They found that fluid intelligence impaired performance was largely confined to patients with right frontal lesions — rather than a wide set of regions distributed across the brain.
“Our findings indicate for the first time that the right frontal regions of the brain are critical to the high-level functions involved in fluid intelligence, such as problem solving and reasoning,” Professor Cipolotti said.
“This supports the use of APM in a clinical setting, as a way of assessing fluid intelligence and identifying right frontal lobe dysfunction.”
“Our approach of combining novel lesion-deficit mapping with detailed investigation of APM performance in a large sample of patients provides crucial information about the neural basis of fluid intelligence.”
“More attention to lesion studies is essential to uncover the relationship between the brain and cognition, which often determines how neurological disorders are treated.”
The results appear in the journal Brain.
_____
Lisa Cipolotti et al. Graph lesion-deficit mapping of fluid intelligence. Brain, published online December 28, 2022; doi: 10.1093/brain/awac304
