Xuzhe Zhang 张旭哲
A lifelong learner
I am interested in studying AI-empowered medical imaging acquisition, processing, and interpretation to transform medical research and clinical practice. My research is multidisciplinary, building on techniques in deep learning, computer vision, and medical imaging. I am interested in developing robust vision models for medical image analysis and computing. My research focused on generative models, self-/semi-supervised learning, and unsupervised domain adaptation. Recently, I am exploring in-context learning in (large) vision model.
On the application side, I focused on deploying open-source frameworks in medical / clinical research communities, with the goal of assisting large-scale and cross-modality clinical studies to discover quantitative image-based markers and thereby improve healthcare.
TL;DR: Computer Vision & AI in Medical Imaging = Intelligent Vision in Healthcare
news
Feb 27, 2024 | MAPSeg has been accepted at CVPR 2024! See you in Seattle! |
---|---|
Nov 28, 2023 | Our recent work on self-supervised learning and unsupervised domain adaptation for heterogeneous medical image segmentation is now available on arXiv: MAPSeg: Unified Unsupervised Domain Adaptation for Heterogeneous Medical Image Segmentation Based on 3D Masked Autoencoding and Pseudo-Labeling. MAPSeg is the first unified UDA framework that works for centralized, federated, and test-time UDA. |
May 16, 2023 | I started my 2023 summer internship at GE Healthcare as an AI/ML PhD Intern. I will continue my research in advancing medical vision! |
May 13, 2022 | Our paper PTNet3D: A 3D High-Resolution Longitudinal Infant Brain MRI Synthesizer Based on Transformers has been accepted for publication in IEEE Transactions on Medical Imaging (IEEE-TMI). link |
Mar 16, 2022 | I passed my qualifying exam and am officially a Ph.D. candidate! |
selected publications
- arXivMAPSeg: Unified Unsupervised Domain Adaptation for Heterogeneous Medical Image Segmentation Based on 3D Masked Autoencoding and Pseudo-Labeling2023
- JournalPTNet3D: A 3D High-Resolution Longitudinal Infant Brain MRI Synthesizer Based on TransformersIEEE Transactions on Medical Imaging 2022