Yiming Qian¹ (), K. Suzanne Scherf; ¹Department of Psychology, the Pennsylvania State University

Most neuroimaging studies assessing brain-behavior relations tend to evaluate pairwise associations between a single neural metric (e.g., fMRI activation or microstructural properties) and a behavioral metric (e.g., accuracy). However, this approach does not characterize how brain structure and function interact holistically to subserve behavior. To address this gap, we integrated fMRI, sMRI, DTI, and functional connectivity measures to investigate how relational patterns among these metrics predict face recognition (FR) behavior over time. The data presented here represent the baseline comparison between autistic (N = 29) and non-autistic (N = 22) adolescents (11-17 years). We chose to study the FR system due to the prominent difficulty in face recognition behavior that exists in autism and the whole-brain nature of the FR neural system. For the fMRI and sMRI metrics, we individually defined six ROIs—the bilateral fusiform face area (FFA) and amygdala as the core regions in the FR system, and bilateral early visual cortex (EVC). Structural connectivity was defined as the radial diffusivity of the white matter tracts seeding from each ROI. Functional connectivity was defined across all ROIs using GIMME and quantified using overall node strength. Partial Least Squares, a multivariate method, was then utilized to understand how the relation among these metrics predicted FR behavior. The results showed robust brain-behavior associations. Specifically, for both groups, higher functional activation within regions (right FFA and left amygdala), functional connectivity (left EVC), and radial diffusivity in most ROIs within the network together positively predicted higher FR scores (i.e., CFMT). The relation between the brain metrics and FR behavior was stronger in the autistic than the non-autistic group. This work may inspire new ways of thinking about how neural networks dynamically organize through development to support behavioral change and, more specifically, why face recognition is a challenging social behavior for autistic individuals.

Acknowledgements: The research was supported by Pennsylvania Department of Health SAP grant 4100047862 (M.B., K.S.S., N.M.), NICHD/NIDCD P01/U19 HD35469-07 (M.B., PI Nancy Minshew).