HFG-Net: High-frequency guided multi-view graph convolution and dynamic spatio-temporal fusion for ASD diagnosis.

Autism Spectrum Disorder (ASD) is a highly heterogeneous neurodevelopmental condition characterized by significant inter-subject variability in electroencephalogram (EEG) features. Existing deep learning approaches, such as Convolutional Neural Networks (CNNs) and Graph Convolutional Networks (GCNs), often fail to fully capture intrinsic spatio-temporal dependencies or rely on static feature fusion strategies, struggling to characterize complex non-Euclidean topological abnormalities. To address these challenges, we propose HFG-Net, a High-Frequency Guided Spatio-Temporal Synergistic Network. This framework incorporates three core innovations: First, a Channel-Temporal Multi-scale Attention (CTMA) mechanism captures transient spatio-temporal coupling via explicit cross-dimensional matrix interaction. Second, a High-Frequency Guided Multi-View Graph Construction strategy uniquely leverages sparse skeletons from Beta/Gamma bands to filter all-band Pearson Correlation and Phase Locking Value matrices, constructing robust noise-resistant topologies. Third, a Dynamic Synergistic Fusion module employs a gating network to learn sample-level confidence weights for adaptive feature integration. Experiments on a clinical dataset of 120 subjects (60 ASD and 60 TD) demonstrate that HFG-Net achieves a classification accuracy of 95.79%and an F1-score of 92.89% on an independent test set, significantly outperforming state-of-the-art models.Interpretability analysis reveals that the model's focus on high-frequency abnormal connectivity aligns with neuropathological findings, while dynamic weight distribution confirms its capability to adapt to heterogeneous samples. Furthermore, the model achieves a recognition rate of 98.43% for patients with mild ASD. HFG-Net effectively addresses the challenges of synergistic spatio-temporal modeling and heterogeneity adaptation, providing an efficient, robust, and interpretable paradigm for EEG-based diagnosis.