Wei Zhang

Wei Zhang(张维)

Research Fellow
Centre for Brain-Computing Research (CBCR), Nanyang Technological University

I am a research fellow at NTU, working at the intersection of neuroscience, AI, and brain-computer interfaces. My research focuses on understanding the spatiotemporal neural dynamics of human cognition using multimodal neuroimaging (EEG, fMRI, MEG, PET), and translating these insights into non-invasive speech decoding BCIs.

I received my BS in Engineering Physics from Tsinghua University, and my PhD in medical imaging from the University of Chinese Academy of Sciences (IHEP), mentored by Prof. Baoci Shan. I currently work with Prof. Cuntai Guan, Prof. Victoria Leong, and Prof. Balazs Gulyas.

Multimodal Neuroimaging Covert Speech Decoding Brain-Computer Interfaces EEG-fMRI Integration AI for Neuroscience Infant Brain Development

Contact

Email
weizhang2022@outlook.com
Google Scholar
Wei Zhang
GitHub
Wilson-WeiZhang

Education & Experience

Selected Publications

Adult-to-infant unidirectional neural coupling mediates selective social learning in infants from the United Kingdom and Singapore
W. Zhang, K. Clackson, S. Georgieva, L. Santamaria, V. Reindl, V. Noreika, N. Darby, V. Valsdóttir, P. Santhanakrishnan, V. Leong*
Nature Communications, 2026 · 17(1):7558
Unidirectional adult-to-infant neural coupling, modulated by eye contact, mediates selective learning of artificial language in infants across UK and Singapore — a better predictor than infants’ own neural activity.
PDF DOI
Decoding Covert Speech From EEG by Functional Areas Spatio-Temporal Transformer
M. Jiang, W. Zhang, Y. Ding, K.A.C. Teo, L.G. Fong, S. Zhang, Z. Guo, C. Liu, R. Bhuvanakantham, W.K.J. Sim, ..., B. Gulyás, C. Guan
IEEE Journal of Biomedical and Health Informatics, 2026 · 30(6):4974–4984
A functional-area-informed spatio-temporal transformer decodes covert speech from EEG, targeting the low signal-to-noise ratio that limits non-invasive speech BCIs.
PDF DOI
Revealing the spatiotemporal brain dynamics of covert speech compared with overt speech: A simultaneous EEG-fMRI study
W. Zhang, M. Jiang, K.A.C. Teo, R. Bhuvanakantham, L.G. Fong, W.K.J. Sim, ...
NeuroImage, 2024 · 293:120629
Simultaneous EEG-fMRI reveals covert speech activates left putamen earlier than overt speech, with weaker connectivity to speech regions — offering mechanistic insights for BCI speech applications.
PDF DOI
Contrastive Graph Pooling for Explainable Classification of Brain Networks
J. Xu, Q. Bian, X. Li, A. Zhang, Y. Ke, M. Qiao, W. Zhang, W.K.J. Sim, B. Gulyás
IEEE Transactions on Medical Imaging, 2024 · 43(9):3292–3305
ContrastPool combines contrastive dual-attention with differentiable graph pooling for fMRI-based brain network classification, with extracted patterns aligning with established neuroscience knowledge.
PDF DOI
Investigating sea-level brain predictors for acute mountain sickness: A multimodal MRI study
W. Zhang, J. Feng, W. Liu, S. Zhang, X. Yu, J. Liu, B. Shan, L. Ma
American Journal of Neuroradiology, 2024 · 45(6):809–818
Sea-level resting-state fMRI of somatomotor network function predicts acute mountain sickness at 3,650m with 86.4% accuracy, enabling pre-exposure screening.
PDF DOI
Deep learning with 18F-FDG-PET gives valid diagnoses for the uncertain cases in memory impairment of Alzheimer's disease
W. Zhang, T. Zhang, T. Pan, S. Zhao, B. Nie, H. Liu, B. Shan, ...
Frontiers in Aging Neuroscience, 2021 · 13:764272
Deep learning on FDG-PET achieves 95.65% accuracy on definitive cases and resolves uncertain diagnoses by identifying neurodegeneration vs. depression-related metabolic signatures.
PDF DOI

Full list: Google Scholar · 20 publications · 279 citations · h-index 10 · i10-index 10

Research

Covert Speech and Speech BCI

EEG-fMRIEEG-to-SpeechBCI

Neural dynamics and decoding models for imagined/covert speech, spanning simultaneous EEG-fMRI, source-informed EEG, and transformer-based speech BCI datasets.

NeuroImage 2024 · IEEE JBHI 2026

Infant Social Learning and Neural Coupling

Dual-Brain EEGInfantsSocial Gaze

Adult-to-infant unidirectional neural coupling and gaze-mediated learning across UK and Singapore cohorts, linking social attention to selective statistical learning.

Nature Communications · 2026

STARS-LLM and Social AI Modeling

LLMDyadic InteractionDevelopment

LLM-based modeling of mother-infant interaction and neural coupling, extending dual-brain EEG work toward developmental risk prediction and interpretable social AI.

In progress

Lower-Limb Motor Imagery and Gait Decoding

EEGMotor ImageryRehabilitation

EEG decoding of lower-limb motor imagery, imagined walking, and gait-related neural dynamics for non-invasive rehabilitation and BCI applications.

In progress

Brain Graph Neural Networks and Connectomics

fMRIGNNExplainability

Graph learning methods for brain-network classification, including contrastive graph pooling, multi-atlas fusion, and clinical connectomics surveys.

IEEE TMI · 2024

Clinical Neuroimaging Biomarkers

PETMRI/fMRIAlzheimer's Disease

PET/MRI and multimodal imaging biomarkers for Alzheimer's disease, acute mountain sickness, obesity-related brain morphology, and scanner harmonization.

MRI/PET studies · 2019-2024

Professional Services

Teaching

  • Guest Lecturer, HP4021 Advanced Neuroscience Seminar, NTU
  • AI Tutorial Instructor, GP8000 AI Literacy, Graduate College PGR ICC, NTU

Membership & Talks

  • American Society of Neuroradiology (ASNR)
  • Young Speaker, AP-CCN 2026, Singapore

Journal Reviewer

  • NeuroImage
  • Imaging Neuroscience
  • IEEE Journal of Biomedical and Health Informatics (JBHI)
  • Developmental Cognitive Neuroscience
  • Frontiers in Human Neuroscience
  • Signal, Image and Video Processing
  • Scientific Reports
  • Computers, Materials & Continua (CMC)