EEG and Clinical AI System

EEG overview

We work on bridging EEG signals and machine learning to develop models that understand how brain activity reflects human physiological states and health. Our research focuses on EEG representation learning, brain dynamics modeling, disease prediction, and clinical decision support. Our methods are grounded in the intrinsic properties of EEG data, such as frequency structures and transient waveforms. We design learning strategies that explicitly integrate data characteristics with prior knowledge from neuroscience and clinical practice (e.g., domain specifications), enabling the development of reliable and trustworthy clinical AI systems.

Selected publications – EEG representation Learning

  • Kotoge, R., Chen, Z., Kimura, T., Matsubara, …, H., Sakurai, Y. “EvoBrain: Dynamic Multi-Channel EEG Graph Modeling for Time-Evolving Brain Networks”. NeurIPS (Spotlight) (2025).
  • Pradeepkumar, J., Piao, X., Chen, Z., Sun, J. “Single-Channel EEG Tokenization Through Time-Frequency Modeling”. ICLR (2025).
  • Piao, X., Chen, Z., Murayama, T., Matsubara, Y., Sakurai, Y. “Fredformer: Frequency Debiased Transformer for Time Series Forecasting.” KDD (2024).

Selected publications – Brain dynamic modeling

  • Jia, H., Chen, Z., Z, L., K, R., P, J., Matsubara, Y., Sun, J., Sakurai, Y., Matsubara, T. “ODEBrain: Continuous-Time EEG Graph for Modeling Dynamic Brain Networks”. ICLR (2026).
  • Jia, H., Chen, Z., Zhu, L., Cao, X., Matsubara, Y., Matsubara, T., Sakurai, Y. “RepSPD: Enhancing SPD Manifold Representation in EEGs via Dynamic Graphs.” ArXiv (2026).

Selected publications – Clinical support system

  • Chen, Z., Matsubara, Y., Sakurai, Y., Sun, J. “Long-Term EEG Partitioning for Seizure Onset Detection.” AAAI (2025).
  • Pradeepkumar, J., Chen, Z., Sun, J. “Neural Signals Generate Clinical Notes in the Wild.” arXiv:2601.22197 (2026).

Multi-omics and Drug discovery

Multiomics overview

We develop machine learning methods for integrating multi-omics data to support biomarker discovery and patient stratification. A central theme of our work is building computational frameworks that connect molecular entities, pathways, and phenotypes through knowledge graphs and structured representations, enabling interpretable and function-aware explanations of model predictions, including molecular signicance and clinical relevance. In parallel, we develop agentic AI workflows that orchestrate bioinformatics tools, statistical models, and visualization components into semi-automated analysis pipelines. We also work on open-source benchmarks that construct AI-ready datasets for diverse omics modalities and evaluate machine learning models across a range of biological tasks. We hope these systems enable hypothesis generation and testing—from data ingestion and quality control to downstream modeling for scientific discovery.

Selected publications – multi-omics & computational oncology

  • Yang, Z., Kotoge, R., Chen, Z., et al. “GeSubNet: Graph Subnetwork Discovery for Neural Biomarkers.” ICLR Oral (2025).
  • Kotoge, R., Yang, Z., Chen, Z., Dong, Y., Matsubara, Y., Sun, J., Sakurai, Y. “ExPath: Towards Explaining Targeted Pathways for Biological Knowledge Bases.” AAAI (2025).
  • Gao, P., Chen, Z., Liu, X., Chen, P., Matsubara, Y., Sakurai, Y. “Antimicrobial Resistance Recommendations via Electronic Health Records with Graph Representation and Patient Population Modeling.” Computer Methods and Programs in Biomedicine (2025).
  • Chen, Z., Zhu, L., Yang, Z., Matsubara, T. “Automated Cancer Subtyping via Vector Quantization Mutual Information Maximization.” ECML-PKDD, (2023).
  • Chen, Z., Yang, Z., Zhu, L., Gao, P., Matsubara, T., Kanaya, S., Altaf-Ul-Amin, M. “Learning Vector-Quantized Representation for Cancer Subtype Identification.” Computer Methods and Programs in Biomedicine, vol (2023).

Selected publications – Benchmark amd Biomedicine AI Agents

  • Wang, Z., Danek, B., Yang, Z., Chen, Z., Sun, J. “Can Large Language Models Replace Data Scientists in Biomedical Research?” Nature Biomedical Engineering (2025).
  • Yang, Z., Kotoge, R., Piao, X., Chen, Z., Zhu, L., Gao, P., Matsubara, Y., Sakurai, Y., Sun, J. “MLOmics: Cancer Multi-Omics Database for Machine Learning.” Scientific Data (2025).