About me

My name is Haoyu Dong, a forth-year Ph.D. student in Electrical & Computer Engineering at Duke University, advised by Prof. Maciej A. Mazurowski. I previously earned my M.S. in Computer Science at Duke, working with Prof. Guillermo Sapiro, and my B.S. in Mathematics–Computer Science and Cognitive Science at UC San Diego, working with Prof. Zhuowen Tu. I also spent a summer as a Research Intern at Siemens Healthineers, working on 3D in-context learning for medical image segmentation.

My research focuses on foundation models and data-centric AI for medical imaging. I build large-scale vision and vision–language foundation models, develop in-context learning approaches for segmentation, and design multimodal agents that integrate imaging with language and tool use. I am also interested in anomaly detection, test-time adaptation, and clinical decision support, with the broader goal of enabling robust, universal, and clinically meaningful medical AI systems.

Experience

Research Intern @ Siemens Healthineers, Princeton, NJ (Jun. – Aug. 2025)
- Advised by Sasa Grbic and Han Liu
- Developed SNAIL, a 3D in-context learning framework for CT segmentation
- Designed multi-stage cross-attention and masked foreground pooling to generalize to unseen anatomical classes from a handful of labeled support volumes
- Trained on ~15k CT volumes; applied GPT-4o for large-scale label harmonization across heterogeneous datasets

Haoyu Dong

Experience

Recent News