祝歆韵 🎓
祝歆韵 Zhù Xīnyùn

PhD Student

Zhejiang University

About Me

Hi there! I’m a Ph.D. student at the College of Computer Science and Technology, Zhejiang University, where I’m fortunate to be advised by Prof. Yueming Wang. My research centers on invasive motor brain-computer interfaces (BCIs), with a focus on dynamic encoding and decoding of movement intention.

Before that, I received my Bachelor’s degree in Software Engineering from Northwestern Polytechnical University. I’ve been involved in building clinically oriented BCI systems, including China’s first closed-loop invasive BCI for cursor control and the world’s first BCI-powered robotic arm that writes Chinese characters.

Beyond research, I’m a big believer in living life to the fullest. I love trying out all kinds of new things—road trips, skiing, aerial photography, diving, baking, crafting, Chinese calligraphy and painting, even DIY home design. There’s always something new to explore!

Let’s connect and make cool things happen!

Download CV
Interests
  • Brain-Computer Interface
  • Computational Neuroscience
  • Artificial Intelligence
Education

  • PhD Computer Science

    Zhejiang University

  • BEng Software Engineering

    Northwestern Polytechnical University

📚 My Research

I am a researcher in the field of brain-computer interfaces (BCIs), with a focus on invasive neural encoding and decoding for motor restoration. My work aims to understand and leverage the neural mechanisms underlying movement and language to develop next-generation neuroprosthetics.

I lead a translational BCI project involving invasive electrophysiological recordings. My research integrates neuroscience, biomedical engineering, and machine learning to advance both scientific discovery and clinical application.

I’m always open to collaboration in the areas of neural signal processing, BCI systems, and neurotechnology. Feel free to reach out! 😃

Featured Publications
Human motor cortex encodes complex handwriting through a sequence of stable neural states
Revealing unexpected complex encoding but simple decoding mechanisms in motor cortex via separating behaviorally relevant neural signals
Tracking functional changes in nonstationary signals with evolutionary ensemble bayesian model for robust neural decoding
A brain-controlled mahjong game with artificial intelligence augmentation
Dynamic ensemble bayesian filter for robust control of a human brain-machine interface
Recent Publications
Yu Qi, 祝歆韵, Xinzhu Xiong, Xiaomeng Yang, Nai Ding, Hemmings Wu, Kedi Xu, Junming Zhu, Jianmin Zhang, Yueming Wang (2025). Human motor cortex encodes complex handwriting through a sequence of stable neural states. NHB.
PDF Cite Code Poster Video Paper
Yu Qi, Yueming Wang, 祝歆韵, Kedi Xu, Gang Pan (2024). Adaptive brain-computer interface decoding method based on multi-model dynamic integration. PCT.
Cite
Yangang Li, 祝歆韵, Yu Qi, Yueming Wang (2024). Revealing unexpected complex encoding but simple decoding mechanisms in motor cortex via separating behaviorally relevant neural signals. eLife.
PDF Cite Code Poster Paper
Jiyu Wei, Dazhong Rong, 祝歆韵, Qinming He, Yueming Wang (2024). Speed-enhanced Subdomain Alignment for Long-term Stable Neural Decoding in Brain-computer Interfaces. BIBM.
PDF Cite Paper
祝歆韵, Yu Qi, Gang Pan, Yueming Wang (2022). Tracking functional changes in nonstationary signals with evolutionary ensemble bayesian model for robust neural decoding. NeurIPS.
PDF Cite Poster Slides Video Paper
See all publications
Awards
SPOTLIGHT POSTER AWARD
BCI AWARD NOMINEE
BEST POSTER AWARD

BEST POSTER AWARD

One critical problem in BCI systems is how to achieve robust control. The challenge lies in the variability of neural signals, that can be induced by: noises and variability of neural encoding associates with brain states. When neural signal changes, there can be decoding errors given a static neural decoder, exhibiting unstable BCI control. Ideally, a neural decoder should dynamically change along with changes in neural signals to achieve robust control. But how to achieve this? We propose a DyEnsemble model, which builds a pool of models to encode various neural functions, assigns the model weights dynamically according to the incoming neural signals, and assembles them in time with Bayesian model averaging rules. In this way, our decoder adjusts itself along with changes in neural signals.