The Brain-Computer Interface and Integrated Intelligence Lab aims to systematically develop invasive brain-computer interfaces, elucidate the neural mechanisms of brain function, and advance brain-inspired computing. These technologies will be applied to fundamental neuroscience research and personalized diagnosis and treatment of brain diseases, driving the advancement of brain-computer intelligence.

Main research directions:

1. Minimally invasive brain-computer interface system

Invasive brain-computer interfaces (BCIs) represent a frontier in brain-computer integrated intelligence. By establishing high-speed bidirectional information exchange between the brain and external devices, they aim to restore motor/sensory functions in tetraplegic patients, rehabilitate damaged vision and hearing, and treat neuropsychiatric disorders. Our lab aims to develop minimally invasive BCI technologies and integrated systems. This includes individualized non-invasive precision neuro-navigation, minimally invasive cranial access techniques, and automated high-precision implantation of flexible electrodes. These innovations will enable automated, craniotomy-free multi-region implantation with vascular avoidance capabilities, achieving precise and efficient surgeries.

2. Brain science research

Behavior and cognition emerge through neuronal interaction, which requires integration between local and distant areas orchestrated by densely connected networks. Our lab explores information processing through neural circuit interactions, and context-dependent dynamic regulation of these networks for functional optimization across behavioral states. These include neural signal encoding and decoding, brain-machine dual-learning fusion control, brain network structure-function relationships, and identification of individualized functional brain regions etc. These advances will apply to brain-computer control systems, bidirectional neural interfaces, and biomarker identification for brain disorders.

3. Brain-inspired computing

We apply principles of neural information processing from biological networks to guide novel artificial neural network designs. These include leveraging self-organized critical states to enhance artificial neural network computational capabilities, and developing new learning algorithms with biomimetic architectures to enable continuous, context-aware learning in deep neural networks.