K-MaT: Knowledge-Anchored Manifold Transport for Cross-Modal Prompt Learning in Medical Imaging

Publication
arXiv preprint arXiv:2603.06340
Jiajun Zeng
Jiajun Zeng
PhD Candidate

Jiajun Zeng is a PhD candidate at the University of Bonn, supervised by Prof. Shadi Albarqouni. He holds an MSc in Biomedical Engineering from Shenzhen University. During his master’s studies, he focused on AI for ultrasound image computing under the supervision of Assoc. Prof. Ruobing Huang and Prof. Dong Ni. His research interests span AI for healthcare, particularly in enhancing radiologists’ workflows through anatomical structure segmentation, disease diagnosis, and treatment support — with a focus on scenarios involving limited annotations and scarce computational resources. Currently, he is exploring efficient vision-language models and multimodal large language models for applications in the healthcare domain.

Shadi Albarqouni
Shadi Albarqouni
Professor of Computational Medical Imaging Research at University of Bonn