Contributed Talk Sessions | Poster Sessions | All Posters | Search Papers

Poster Session A: Tuesday, August 12, 1:30 – 4:30 pm, de Brug & E‑Hall

Graph-Attention-Based Integration of Brain Structure and Function for Trait Anxiety Prediction: Preliminary Results

Jungyoun Janice Min¹, Jiong Chen, Jingxuan Bao², Shu Yang², Yize Zhao³, Li Shen⁴, Duy Duong-Tran⁵; ¹Seoul National University, ²University of Pennsylvania, University of Pennsylvania, ³Yale University, ⁴University of Pennsylvania, ⁵United States Naval Academy

Presenter: Jungyoun Janice Min

Trait anxiety is a stable personality trait linked to increased vulnerability for internalizing disorders. Although altered intrinsic activity in individuals with trait anxiety has been frequently reported in resting-state fMRI studies, its relationship to structural connectivity, which is axonal pathways of large-scale brain dynamics, remains underexplored. Leveraging the LEMON dataset (N = 132), we trained a graph-attention network integrating temporally structured functional signals at rest with subject-specific structural constraints. Our model outperformed a traditional structure–function coupling baseline, achieving statistically significant prediction (r = 0.194, p = 0.026). Attention based interpretation highlighted importance of frontal–parietal and occipital pathways, suggesting that the attentional and sensory networks may contribute to trait anxiety.

Topic Area: Brain Networks & Neural Dynamics

Extended Abstract: Full Text PDF