Objective. Motor imagery EEG (MI-EEG) decoding remains challenging due to low signal-to-noise ratios and pronounced inter-subject variability. Although end-to-end deep models reduce reliance on manual feature engineering, many existing architectures may introduce temporal leakage through non-causal operations and often rely on fixed spatial topologies that cannot accommodate subject- and trial-specific connectivity patterns. Approach. We propose MAGCANet, which integrates five core components: (i) a Multiscale Causal Convolution Module (MCCM) for hierarchical temporal encoding under explicit causal constraints, (ii) a Temporal Convolution Module (TCM) to capture complex temporal dynamics, (iii) an Adaptive Graph Convolution Module (AGCM) for sample-specific topology learning in latent space, (iv) a Multi-Head Self-Attention Module (MHSAM) for global feature aggregation, and (v) a Classification Block for final decision making. Together, these components enforce temporal causality, adapt spatial interactions to individual dynamics, and produce discriminative representations robust to inter-subject variability. Results. On the BCI Competition IV-2a and IV-2b datasets, MAGCANet achieves strong single-subject accuracies of 88.58\% and 91.13\%, respectively. Under Leave-One-Subject-Out (LOSO) evaluation, the model maintains accuracies of 70.49\% and 79.49\%, demonstrating competitive and stable cross-subject generalization. MAGCANet is highly lightweight, with only 0.0194M parameters, and achieves low inference latency (2.23 ms). Qualitative analyses, including feature clustering and channel occlusion, further highlight the model’s interpretability and its ability to capture relevant EEG patterns. Significance. MAGCANet provides a robust and interpretable solution for MI-EEG decoding, balancing high precision with computational efficiency, and offering a reliable method for real-time BCI applications.

Developmental dyslexia (DD) is a common neurodevelopmental disorder, and abnormalities in the functional brain network structure are suggested to be closely related to dysregulation of connections between brain regions. However, the applicability of different brain atlas segmentation strategies in DD brain network modeling has not been systematically evaluated. To this end, this study constructed brain functional networks using three brain atlases (AAL, Craddock, and Power) for comparative analysis and built a Graph Convolutional Network (GCN) classification framework that integrates traditional graph-theoretic features and higher-order global topological features extracted using Persistent Homology. The results of the experiment show that the AAL template performs best in DD recognition, with a classification accuracy rate of 83.31%. Additionally, by converging cross-threshold topological feature stability analysis with Grad-CAM model attribution, we identified a set of brain regions that exhibited both robust intergroup differences across multiple sparsity levels and high importance in the GCN’s decision-making process, thereby jointly pinpointing the core neural substrates of DD. Our results provide a theoretical basis for brain atlas selection and functional connectivity modeling, and offer a new path for optimizing DD-assisted recognition methods.