Yaolei Qi

Yaolei Qi

PhD Candidate

Southeast University

Research Interests

Computer Vision
Medical Image Analysis
Foundation Models
Coronary Artery Analysis
Label-efficient Learning

About

Yaolei Qi is a Ph.D. Candidate in the School of Computer Science and Engineering at Southeast University, China. He received his B.E. degree from Southeast University in 2019 and entered a successive postgraduate-doctoral program in 2021. From March to September 2024, he was a visiting researcher at the Department of Applied Mathematics and Theoretical Physics, University of Cambridge, UK, and he is expected to graduate in September 2026. His research interests include medical image analysis, foundation models, and coronary artery analysis, with a particular focus on tubular structure analysis. He has published and co-authored papers in leading journals and conferences, including IEEE T-IP, IEEE T-CSVT, Medical Image Analysis, ICCV, and MICCAI. He also serves as a reviewer for major conferences such as NeurIPS, CVPR, ICCV, ECCV, and MICCAI, as well as journals including IEEE T-MI and Pattern Recognition.

I am always open to research collaboration and glad to connect with people who share similar interests. If you are interested in related research topics, feel free to email me.

Selected Publications

View All

Dynamic snake convolution based on topological geometric constraints for tubular structure segmentation

Yaolei Qi, Yuting He, Xiaoming Qi, Yuan Zhang, Guanyu Yang

Proceedings of the IEEE/CVF international conference on computer vision

This work presents Dynamic Snake Convolution, which adaptively models the geometry of thin and curvilinear tubular structures to improve segmentation performance.

Examinee-examiner network: Weakly supervised accurate coronary lumen segmentation using centerline constraint

Yaolei Qi, Han Xu, Yuting He, Guanyu Li, Zehang Li, Youyong Kong, Jean-Louis Coatrieux, Huazhong Shu, Guanyu Yang, Shengxian Tu

IEEE Transactions on Image Processing

This paper proposes Examinee-Examiner Network, a weakly supervised framework that leverages centerline-based topological constraints to improve the continuity and accuracy of coronary lumen segmentation in CCTA images.

News

2026-05

1 paper about Video Polyp Segmentation has been early accepted by MICCAI 2026 🎉

2025-12

1 paper about image generation from non-contrast CT has been accepted by IEEE T-CSVT

2025-07

1 paper about federated learning has been accepted by Medical Image Analysis

2024-09

1 paper about MPNs Classification has been accepted by MICCAI 2024 as oral

2023-12

1 paper about spatio temporal has been accepted by IEEE JBHI

2023-7

1 paper about network design has been accepted by ICCV 2023

2023-05

1 paper about partial annotation has been accepted by MICCAI 2023

2021-11

1 paper about coronary segmentation has been accepted by IEEE T-IP