Research
Research
AI-enabled closed-loop neural interfaces for personalized brain modulation.
Full-Cycle BCI/BMI Research Platform
Neural Interface Laboratory aims to build an AI-enabled closed-loop minimally invasive brain-computer interface (BCI) platform. Our major vision is to move beyond conventional neural recording or stimulation systems and develop an integrated platform that can record, decode, and modulate brain activity in real time. By linking neural sensing, AI-based interpretation, and adaptive stimulation within a closed-loop architecture, we seek to overcome the limitations of open-loop BMI systems and realize precise, personalized neuromodulation technologies.
The closed-loop BCI platform developed in our laboratory is expected to expand toward applications for intractable neurological disorders. We aim to develop closed-loop BCI systems that continuously monitor brain states and adaptively regulate stimulation to maximize functional efficacy.
To accomplish these goals, we are developing a full-cycle BCI/BMI research platform composed of four core research areas: Digital Twin Modeling, Neural Interfaces for 3D Modulation and Recording, BCI and Neural Prosthesis, and Preclinical Validation.
Digital Twin Modeling
The first research area is digital twin modeling for personalized brain interfaces. Because each individual has distinct anatomical structures, neural connectivity, disease states, and responses to stimulation, effective BMI and neuromodulation systems require models that reflect subject-specific brain characteristics. Our laboratory develops multiscale computational models that can predict electric fields, magnetic fields, optical stimulation effects, and neural circuit responses. These models are used to optimize stimulation sites, intensity, frequency, waveform, electrode configuration, and optical stimulation patterns.
Neural Interfaces for 3D Modulation and Recording
The second research area is the development of neural interfaces for selective three-dimensional recording and modulation of the brain and nervous system. Many conventional neural interfaces focus on recording or stimulating a limited surface area or a specific depth. However, brain function and neurological disorders emerge from distributed three-dimensional neural networks. Therefore, our goal is to develop neural interface technologies capable of selectively recording and modulating multiple brain regions and depths.
BCI and Neural Prosthesis
The third research area is the development of BCI and neural prosthesis systems. We aim to build bidirectional BCI platforms that decode neural signals to control external devices or artificial sensory systems and provide feedback from those systems back to the nervous system. The key concept is not one-way decoding, but a closed-loop BCI architecture in which neural recording, AI-based interpretation, and stimulation feedback are continuously connected.
Preclinical Validation
The fourth research area is preclinical validation. For neural interface and closed-loop BMI technologies to become clinically meaningful platforms, their safety, stability, and functional efficacy must be validated in biological systems. Our laboratory uses animal models to evaluate whether newly developed interfaces and stimulation strategies operate as intended in vivo.
By integrating digital twin modeling, 3D neural interface technology, BCI and neural prosthesis development, and preclinical validation, Neural Interface Laboratory aims to establish a full-cycle research platform for personalized closed-loop brain modulation and next-generation human-machine interfaces.
