NeuBCIs the abbreviation of Neural computation and Brain-computer interaction.

Our group belongs to the interdisciplinary field of neuroscience, pattern recognition and biomedical engineering. The research team is positioned at the forefront of basic research. It will combine the experimental research methods of cognitive neuroscience and modern artificial intelligence technology to reveal the information coding mechanism in the process of brain information processing and motion control, study the neural signal decoding technology, develop a new high-speed brain computer interface platform, and realize the real "what you think is what you get".

Main research direction

1. Brain computer interface
Research and develop new theories and methods for multi-mode neural signal acquisition in the brain computer interface, and analyze the neural information in the brain computer interface, including real-time artifact suppression of EEG, extraction and analysis of rhythmic waves, fMRI based localization of brain functional areas and analysis of temporal and spatial patterns of EEG activities, synchronous acquisition and joint analysis of eye movement and EEG data, and online implementation of the above algorithms, On this basis, the BCI experimental platform based on virtual reality, the prototype system of multimodal brain computer interface and the corresponding software platform are realized.

2. NeuroImage computing -- Research on brain like mechanism of deep learning based on visual information encoding and decoding
In order to explore the theoretical basis of deep learning, we will use deep neural network to encode and decode visual information, hoping to find the brain like biological basis of deep learning and reveal the mechanism of brain visual computing. Focus on breaking through the problems of multimodal information fusion and brain information decoding for complex natural scene stimulation.
The research is divided into two aspects. (1) Research on visual information encoding and decoding: study the brain's processing mechanism of visual stimulation, analyze the neural processing process of image information and semantic information, and reconstruct the viewed images. (2) Research on non-invasive brain computer interface: Research on fast neural decoding algorithm and develop an online brain computer interface system for motor intention decoding.

3. Method and application of medical image big data processing based on multimodal image
Using pattern recognition and image processing technology, it is applied to medical images, analyzed in combination with clinical information, and combined with multimodal image data to develop a medical image big data processing method based on deep learning and brain like intelligence. It is hoped that the imaging features related to specific diseases can be found to assist doctors in early diagnosis, clinical treatment and prognosis evaluation. The specific research contents include medical image segmentation, registration, visualization, shape analysis, fMRI function analysis, brain network analysis and so on. At present, the diseases of interest include amblyopia, glaucoma and so on.