Relational Integration Demands Are Tracked by Temporally Delayed Neural Representations in Alpha and Beta Rhythms...

Area of Science:

• Cognitive Neuroscience
• Neuroimaging
• Human Cognition

Background:

• Relational reasoning underpins higher cognitive functions and fluid intelligence.
• Relational complexity (RC) classifies reasoning problem difficulty.
• Previous research linked RC to static frontoparietal activity, with limited understanding of spatiotemporal dynamics.

Purpose of the Study:

• To investigate the spatiotemporal neural dynamics underlying relational complexity (RC).
• To examine how RC is represented across brain networks using neuroimaging.
• To differentiate neural coding of RC from cognitive effort (CE).

Main Methods:

• Representational Similarity Analysis (RSA) applied to fMRI and EEG data.
• Participants completed a visuospatial reasoning task with varying RC levels (Latin Square Task).
• Multimodal fusion analysis combined EEG and fMRI signals.

Main Results:

• RC representations were spatially widespread, prominent in frontoparietal, dorsal-attention, and cingulo-opercular networks.
• Temporal dynamics showed RC representation between 2.5-4.1s post-stimulus, in alpha and beta frequency bands.
• Shared EEG-fMRI variability in higher-order networks better supported the RC model over the cognitive effort model.

Conclusions:

• Relational processing involves spatially distributed neural coding across cortical networks.
• Late-stage, frequency-specific neural dynamics are critical for resolving relational complexity.
• Findings advance understanding of the neural basis of relational reasoning and cognitive effort.