Study on the Topological Properties of Language Brain Region Networks in Children With Autism Based on Diffusion Tensor Imaging (DTI).

objectiveTo compare the topological properties of language-related structural brain networks between children with Autism Spectrum Disorder (ASD) and Global Developmental Delay (GDD) and examine their associations with language function.

methodsSixty-five children with ASD and fifty-one with GDD underwent diffusion tensor imaging (DTI). Graph theoretical analysis was used to evaluate nodal metrics in language-related structural networks. Hierarchical regression models analyzed whether diagnostic group moderated the relationship between network metrics and language performance.

resultsBehaviorally, children with ASD exhibited significantly lower auditory and visual comprehension scores than those with GDD, while expressive language scores were comparable. At the network level, the ASD group showed significantly higher degree centrality in the right middle temporal gyrus (MTG-R). In the ASD group specifically, degree centrality of the MTG-R, and both degree centrality and nodal efficiency of the right angular gyrus (ANG-R), were negatively correlated with language performance. Regression analyses confirmed that diagnostic group significantly moderated these structure-function relationships.

conclusionChildren with ASD and GDD exhibit differences in the local topological properties of language networks and distinct patterns of association with language ability. Specifically, increased nodal connectivity in the MTG-R and ANG-R was associated with poorer language outcomes in the ASD group, suggesting disorder-specific differences in structural network organization. This study provides network-level evidence for divergent brain organization underlying language impairment in ASD versus GDD.