AI-Driven Clinical Neuroscience
Leveraging cutting edge machine learning tools to decode the neuropsychiatric disorders. My work focuses on creating AI pipelines for earlier diagnosis, aim to bridge the gap between clinical neuroscience and machine learning.
Whole-Brain Connectome Dynamic models
Integrating multimodal neuroimaging data, from molecular transcriptomics to macroscopic PET/MRI and two-photon imaging—into a unified topological framework. By employing hyperbolic manifold learning and self-supevised learning, decode the underlying structural and functional rules that govern brain function.
Brain-Inspired Computing Architecture
Reverse-engineering biological intelligence to inform next-generation AI architectures. By simulating biological cognition state and behavior, I extract fundamental cognitive algorithms to design more efficient and computational models. My research seeks to overcome the limitations of traditional deep learning by embedding neurobiological principles into bio-inspired AI systems.