Skip to content
View haonguyenthai's full-sized avatar

Block or report haonguyenthai

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
haonguyenthai/README.md

👋 Hi, I’m Hao Nguyen Thai

PhD Candidate @ The University of Queensland (UQ) · Computational Pathology · Spatial Omics · Machine Learning

LinkedIn Email ORCID Blog

Research theme: Learning molecular signals and clinical biomarkers directly from routine histopathology slides by aligning pathology foundation models with spatial transcriptomics/proteomics.


🔄 Career pivot

Chemical Engineering → Digital Transformation → Computational Biology → Computational Pathology. I started in Chemical Engineering (Honours), led digital transformation projects in industry (Unilever and Nestle), pivoted to structual biology (Tokyo Tech), and found my home in AI for pathology & spatial omics—building scalable, reproducible pipelines to bridge images and molecules.


🔭 What I’m working on

  • Slide → Gene Expression
  • Multimodal Fusion: Combine WSIs + spatial omics + clinical metadata for subtype and risk stratification in breast cancer.
  • Efficiency @ Scale: Memory-safe WSI+ST pipelines on HPC (Slurm, Apptainer), using Dask/Zarr where helpful.

🧠 Interests

Computational pathology, spatial omics, representation learning, contrastive/knowledge distillation, uncertainty & interpretability.

🧰 Tech & Tools

Python PyTorch NumPy Pandas Scanpy AnnData OpenSlide CUDA Slurm Apptainer Git DVC


🗂️ Featured projects


🤝 Collaboration

I’m happy to collaborate on WSI × Spatial Omics, foundation model adaptation, and benchmarks. If you have relevant datasets or clinical questions, let’s talk!

Contact: [email protected]


⚡ Quick facts

  • ENTJ · enjoys building structured learning trackers and clean pipelines.
  • Friendly reminder: science is a team sport — please open an issue if any repo is missing steps or you hit env problems.
  • StrengthsFinder (Top 5): Competition · Command · Maximizer · Significance · Ideation

Pinned Loading

  1. graph-neural-networks-roadmap graph-neural-networks-roadmap Public

    Forked from guillaumejaume/graph-neural-networks-roadmap

  2. Kaggle-PANDA-1st-place-solution Kaggle-PANDA-1st-place-solution Public

    Forked from kentaroy47/Kaggle-PANDA-1st-place-solution

    1st place solution for the Kaggle PANDA Challenge

    Jupyter Notebook

  3. LSTM-based-language-model LSTM-based-language-model Public

    Forked from guillaumejaume/LSTM-based-language-model

    Python

  4. TransPath TransPath Public

    Forked from Xiyue-Wang/TransPath

    Python