Our research group focuses on developing biologically plausible, interpretable, and causally grounded machine/deep learning algorithms, bridging the gap between advanced AI methodologies and impactful applications in biomedicine and scientific discovery.

Our key research pillars include:

• Brain Foundation Models & BCIs: Developing cross-subject, multi-paradigm foundation models for EEG/fMRI; advancing brain decoding and neural signal alignment (EEG-to-fMRI) using modern generative architectures (LLMs, Diffusion).
• Causal Learning & Trustworthy AI: Integrating causal discovery and inference with deep learning to capture non-linear, cause-and-effect dynamics in high-dimensional biomedical signals and vision data.
• AI in Healthcare & Vision: Translating heterogeneous clinical data (EHR, multimodal medical/computer vision, knowledge graphs) into actionable insights for personalized medicine, automated radiology reporting, and precision diagnosis/prognosis of neurological disorders (AD, MCI, ASD, MDD).
• AI for Science: Constructing next-generation computational frameworks to model complex molecular and biological systems, accelerating therapeutic discovery and novel drug development.