The cell-type underpinnings of regional brain entropy (BEN) in human brain

Abstract

Brain entropy (BEN) indicates the irregularity, unpredictability and complexity of brain activity. Research using regional brain entropy (rBEN) based on fMRI has established its associations with structural and functional brain networks, as well as their coupling. These relationships are influenced by the sensorimotor-association (S-A) axis and cytoarchitectural organization. Recent studies have also revealed links between functional connectome gradients and cell-type, suggesting that rBEN may have a biological basis at the cellular level. However, this possibility remains unexplored at specifically cellular level. In this study, we analyzed mean rBEN maps derived from 176 participants using HCP 7T data and correlated these with publicly available cell-type datasets. Our findings rBEN exhibited a positive correlation with oligodendrocytes (Oligo) and a negative correlation with parvalbumin-positive interneurons (PVALB) under both resting-state and movie-watching conditions at the whole-brain level. Furthermore, we observed that the S-A axis modulates the relationship between rBEN and cell-type. Specifically, L5 extratelencephalic neurons (L5 ET) showed a positive correlation with rBEN in unimodal cortex but a negative correlation in multimodal cortex and interneurons somatostatin (SST) was correlated with rBEN only in multimodal cortex. These results provide further evidence that rBEN has a biological foundation at the cellular level, with spatial heterogeneity in its associations with different cell types. Moreover, rBEN appears to capture information beyond functional connectivity network.