MEDICAL IMAGE ANALYSIS

Yazhou Zhu 朱亚洲

Ph.D. in Computer Science

I received my Ph.D. in Computer Science from Nanjing University of Science and Technology in June 2026, advised by Prof. Haofeng Zhang. Previously, I earned an M.Eng. in Software Engineering from Jiangnan University and a B.Eng. in Mechanical Engineering from Changzhou University.

My research focuses on medical image analysis with limited annotation and distribution shift, spanning few-shot and cross-domain segmentation, MR image denoising, and adaptation of vision foundation models.

Earth and spacecraft seen from orbit

Research

I study how medical imaging models can make better use of scarce annotations and remain reliable across domains. My work connects three themes:

  • Few-shot and cross-domain segmentation. Prototype learning, regional representations, semantic matching, domain robustness, and uncertainty-aware prompting for medical images.
  • Medical image reconstruction and enhancement. Deep image priors, unsupervised MR denoising, and reconstruction and analysis of MR/CT data.
  • Foundation models and clinician interaction. Adaptation to medical imaging, uncertainty estimation, and learning from expert feedback on rare or unfamiliar cases.

First-author publications

Only papers with Yazhou Zhu listed as first author are shown, ordered by year. Preprints are labeled.

Education

Ph.D. in Computer Science

Feb 2022 – Jun 2026

Nanjing University of Science and Technology
School of Computer Science and Engineering

Advisor: Prof. Haofeng Zhang · Cross-domain few-shot medical image segmentation

M.Eng. in Software Engineering

Sep 2018 – Jun 2021

Jiangnan University
School of Artificial Intelligence and Computer Science

Advisor: Prof. Xiang Pan · MR image denoising and brain tumor segmentation

B.Eng. in Mechanical Engineering

Sep 2014 – Jun 2018

Changzhou University

Projects & tools

Medical Image Research Monitor

A public dashboard for journal articles and preprints across ten medical imaging research interests. Explore research topics, method combinations, and reading leads; filter by date and source, and inspect the underlying papers and abstract excerpts.

Open in English中文版Rule-based analysis · No AI model API required

Expertise & skills

Methods
Prototype networks, Transformers, U-Net/V-Net, few-shot learning, domain adaptation, and deep image priors.
Programming
Python, C/C#, and MATLAB; PyTorch and TensorFlow.
Imaging & data
SimpleITK, NiBabel, and NumPy; 2D/3D medical image preprocessing, normalization, cropping, and augmentation.
Communication
Scientific reading and writing in English and Chinese; independent research and cross-institutional collaboration.