Erjun Zhang

PhD in Biomedical Engineering
Institute of Biomedical Engineering
Polytechnique Montreal, University of Montreal
Lab: NeuroPoly (L5626, the green floor at Poly)
Room: 1.7.13, TransMedTech Institute, CHU Sainte-Justine Research Center
Email: erjunzhang [at] outlook [dot] com
Status: Available now for Postdoc (Holding 3-yr Canadian Open Work Permit)
GitHub LinkedIn Google Scholar CV

I'm a PhD-trained biomedical researcher working on diffusion MRI and quantitative biomarkers for brain development and injury. During my PhD at Polytechnique Montreal (NeuroPoly lab) and CHU Sainte-Justine, I introduced the Diffusion Bubble Model (DBM) and demonstrated its clinical utility for detecting and subtyping punctate white matter lesions (PWML) (NeuroImage, 2025). I also build reproducible neonatal MRI workflows (QC, registration, atlas-informed ROI analysis) and scalable segmentation tools to enable cohort studies linking early MRI markers to later outcomes.

I am actively seeking postdoctoral research opportunities. My research interests lie in advancing biophysical modeling, benchmarking diffusion frameworks, and translating quantitative MRI biomarkers into clinical applications for brain, spinal cord and body injury and development.

News:

Mar 11, 2026: Invited talk, FNNDSC Lecture Series, Boston Children's Hospital / Harvard Medical School (Boston, MA).
Jan 2026: Completed the development of an end-to-end mouse brain MRI preprocessing and analysis pipeline (T1w/T2w/dMRI).
Oct 23, 2025: Officially completed PhD studies and received completion letter (successfully defended on Sep 9, 2025).
June 11, 2025: First-author paper accepted in NeuroImage: Diffusion Bubble Model: A Novel MRI Approach for Detection and Subtyping of Neonatal PWML .

[expand]

[expand all]

[collapse all]

Education

PhD in Biomedical Engineering Oct 2025
Institute of Biomedical Engineering, Polytechnique Montreal
Thesis: DBM: A Spectrum-Based Diffusion MRI Framework for Neonatal Brain Segmentation, Developmental Assessment, and Injury Characterization.
Supervisors: Prof. Benjamin De Leener & Prof. Gregory A. Lodygensky.
MEng in Optical Engineering July 2018
Beijing University of Technology, China
Thesis: Numerical Simulation and Image Reconstruction for Structured Illumination Super-Resolution Fluorescence Microscopy.
Focus: Computational Imaging, Inverse Problems, Advanced Optics.
BEng in Optoelectronic Information Engineering June 2014
Shenzhen University, China
Core Coursework: Information Optics, Digital Image Processing, Computational Imaging, Image Analysis.

Research Framework & Projects

I. Methodological Innovation

Diffusion Bubble Model (DBM): Developed a spectrum-based dMRI framework to resolve sub-voxel heterogeneity. Demonstrated clinical utility in separating lesions from normal tissue (NeuroImage 2025). [Open Source Coming Soon]

II. Reproducible Workflows

Neonatal Processing Pipeline: Built atlas-informed segmentation and registration workflows robust to motion and low-resolution clinical data.

End-to-End Mouse MRI Pipeline: (New) Developed a fully automated preprocessing pipeline for mouse brain T1w/T2w/dMRI to support translational validation.

III. Clinical Translation

Neonatal Brain Injury & Development: Characterized maturation trajectories (34-40 weeks PMA) and detected Punctate White Matter Lesions (PWML) with high specificity.

Neurodegeneration (Spinal Cord): (In Progress) Applied DBM to Parkinson's disease spinal cord imaging, successfully distinguishing between different disease stages (severity stratification).

IV. Quantitative Analysis Tools

Microstructure-Informed Segmentation: Leveraged DBM and diffusion metrics to refine brain tissue segmentation (completed in PhD thesis).

V. Future Roadmap (Collaboration Opportunities)

Non-CNS MRI Expansion: Adapting diffusion MRI models for cardiac, abdominal (liver/kidney), and muscle imaging.

Broadening Disease Applications: Applying DBM to characterize complex pathologies beyond injury, such as Multiple Sclerosis (MS) and tumor.

Methodological Refinement:

Constructing population-specific parameter templates.
Refining diffusion models with rigorous mathematical constraints.

Deep Learning & Validation:

Benchmarking against histology or multimodal imaging.
Developing Physics-Informed Neural Networks for metrics estimation.

Awards & Grants

Teaching & Mentoring

Publications

Opensource

Short Bio

Dr. Erjun Zhang is a biomedical researcher specializing in diffusion MRI modeling and quantitative biomarkers for brain, spinal cord, and body injury and development. He received his PhD in Biomedical Engineering from Polytechnique Montreal (2025) and his MEng in Optical Engineering from Beijing University of Technology (2018). His doctoral work introduced the Diffusion Bubble Model (DBM) to characterize sub-voxel heterogeneity in neonatal brain injury. Dr. Zhang combines rigorous physical modeling with reproducible computational pipelines to translate advanced neuroimaging techniques into clinical applications.

Page content update time:
Statistics: visitors, views (since March 31, 2023)