چكيده به لاتين
Fluid intelligence refers to the ability to solve new problems, use logic, and reason without relying on prior knowledge or experience, and it plays a significant role in cognitive functioning and adapting to new environments. Assessing fluid intelligence using brain signals is particularly important due to its association with the functioning of neural networks. In this study, electroencephalogram (EEG) data were collected from 109 healthy subjects (both men and women) during a resting state with eyes open, using 63 electrodes for a duration of 5 minutes.
After preprocessing the data, frequency band power and three types of functional connectivity matrices were calculated, including the magnitude-squared of coherence (MSC), inter-site phase clustering (ISPC), and phase-lag index (PLI). Graph features were then extracted from these matrices, and their Pearson correlation with intelligence levels across different frequency bands was examined. The results showed that alpha band power and the ISPC and MSC matrices in this band had a significant correlation with fluid intelligence. This correlation was stronger in men than in women, while for women, a weaker correlation was observed in the beta2 band. PLI performed better than other measures in estimating intelligence levels, although it did not show a significant correlation with intelligence.
Additionally, graph metrics such as betweenness centrality, eigenvector centrality, characteristic path length, and modularity in the alpha and beta2 bands showed the highest correlation with intelligence. By using the Wilcoxon rank-sum test to examine differences between various groups in terms of gender and intelligence level, significant gender differences in brain structure were observed. Finally, a method for estimating intelligence based on the connectome was proposed, yielding satisfactory results.
Keywords: fluid intelligence, Resting-state electroencephalogram, functional connectivity, graph theory, Wilcoxon rank-sum test, Pearson correlation.
