Abstract:The neuroimaging basis of intelligence remains elusive; however, there is a growing body of research employing connectome‐based predictive modeling to estimate individual intelligence scores, aiming to identify the optimal set of neuroimaging features for accurately predicting an individual's cognitive abilities. Compared to adults, the disparities in cognitive performance among children and adolescents are more likely to captivate individuals' interest and attention. Limited research has been dedicated to exploring neuroimaging markers of intelligence specifically in the pediatric population. In this study, we utilized resting‐state functional magnetic resonance imaging (fMRI) and intelligence quotient (IQ) scores of 170 healthy children and adolescents obtained from a public database to identify brain functional connectivity markers associated with individual intellectual behavior. Initially, we extracted and summarized relevant resting‐state features from whole‐brain or functional network connectivity that were most pertinent to IQ scores. Subsequently, these features were employed to establish prediction models for both performance and verbal IQ scores. Within a 10‐fold cross‐validation framework, our findings revealed that prediction models based on whole‐brain functional connectivity effectively predicted performance IQ scores( R=0.35,P=2.2×10−4 ) but not verbal IQ scores( R=0.12,P=0.20 ). Results of prediction models based on brain functional network connectivity further demonstrated the exceptional predictive ability of the default mode network (DMN) and fronto‐parietal task control network (FTPN) for performance IQ scores ( R=0.71,P=2.2×10−18 ). The above findings have also been validated using an independent dataset. Our findings suggest that the performance IQ of children and adolescents primarily relies on the connectivity of brain regions associated with DMN and FTPN. Moreover, variations in intellectual performance during childhood and adolescences are closely linked to alterations in brain functional network connectivity.

Brain Networks and Intelligence: A Graph Neural Network Based Approach to Resting State fMRI Data

Brain Networks and Intelligence: A Graph Neural Network Based Approach to Resting State fMRI Data

Intelligence Quotient Scores Prediction in rs-fMRI via Graph Convolutional Regression Network

BrainGNN: Interpretable Brain Graph Neural Network for fMRI Analysis

A deep learning based approach identifies regions more relevant than resting‐state networks to the prediction of general intelligence from resting‐state fMRI

A Hierarchical Graph Learning Model for Brain Network Regression Analysis.

Brain Functional Connectivity Analysis via Graphical Deep Learning

Whole-brain Prediction Analysis with GraphNet

NeuroGraph: Benchmarks for Graph Machine Learning in Brain Connectomics

BrainNet with Connectivity Attention for Individualized Predictions Based on Multi-Facet Connections Extracted from Resting-State Fmri Data

Weighted Ensemble-model and Network Analysis: A method to predict fluid intelligence via naturalistic functional connectivity

Graph-based network analysis of resting-state functional MRI

A brain structure learning-guided multi-view graph representation learning for brain network analysis

EEG connectome-based predictive modeling of nonverbal intelligence level in healthy subjects

Prediction of individual performance and verbal intelligence scores from resting‐state fMRI in children and adolescents

Meta-RegGNN: Predicting Verbal and Full-Scale Intelligence Scores using Graph Neural Networks and Meta-Learning

A Deep Graph Neural Network Architecture for Modelling Spatio-temporal Dynamics in resting-state functional MRI Data

Integrated Brain Connectivity Analysis with fMRI, DTI, and sMRI Powered by Interpretable Graph Neural Networks

Brain Network Classification Based on Dynamic Graph Attention Information Bottleneck.

Fmri-Based Brain Disease Diagnosis: A Graph Network Approach

Learnable Brain Connectivity Structures for Identifying Neurological Disorders