Human Brain Mapping of Homotopic Functional Affinity

Abstract

Homotopic positions are defined as the two areas with opposite but equal horizonal coordinates in the standard symmetric brain space. Characterizing similarity between two homotopic areas, brain homotopy represents a typical feature of the two hemispheres for both structure and function. Functional homotopy provides important perspectives for understanding neural correlates of cognition and behavior. Despite the decisive role of spatial geometric constraints and homophilic attachment on the human connectome, traditional practices in mapping functional homotopy only considered the temporal correlations of functional timeseries between homotopic areas, but ignored the homophily factors in generative connectivity models. Here, we proposed a novel method for functional homotopy analysis, namely Homotopic Functional Affinity (HFA). This method quantifies the homotopic affinity as the cosine distance of the full-brain functional connectivity profiles or fingerprints between the homotopic areas. HFA captures both geometric constraints (homotopic location) and homophily (affinity) simultaneously. By leveraging the resting-state fMRI data from the Human Connectome Project (HCP) and the Chinese HCP (CHCP), we mapped the 700ms-2mm high spatiotemporal resolution HFA and evaluated its test-retest reliability with linear mixed models, exhibiting generally fair-to-substantial reliable measurements of individual differences in HFA. The lowest HFA observed in the temporo-parietal junction (TPJ) inspired to perform an edge-detection algorithm on its surface render and derived three clearly differentiable and adjacent TPJ subregions: the anterior TPJ (TPJa), the central TPJ (TPJc), and the posterior TPJ (TPJp). We further validated the HFA for the three TPJ regions through a set of comprehensive analyses, including the delineation of their functional connectivity fingerprints, the meta-analysis of their cognitive functions, and their task-activation correlation. Finally, we linked the cortical HFA map to those multimodal brain maps of gene expression, evolution, myelination, functional hierarchy, and cognitive association. The systematic subregion analysis revealed the complex hemispheric specialization of TPJ in attention, social cognition, and language functions. In general, functional specialization of the TPJ areas was stronger in the left hemisphere. The findings from the task activation correlation were highly consistent with those of the meta-analysis. Notably, there were significant differences in social cognition relevant to the three TPJ areas between HCP and CHCP datasets. Furthermore, the correlation analysis of multimodal brain maps illustrated a close relationship between the HFA map and multimodal brain maps. The consistency of maps derived in distinct analyses demonstrated the feasibility of HFA in further understanding psychological and behavioral mechanisms on neural lateralization from the perspective of hemispheric functional integration and specialization. In summary, the proposed HFA framework provides a reliable and valid functional brain mapping tool, with broad applicability in population neuroscience.