Zhe Xu
Scholar

Zhe Xu

Google Scholar ID: K-EoeRgAAAAJ
CUHK | MGH/Harvard | THU
Medical Image AnalysisHealthcare AIAgentic AIMultimodalFoundation Models
Citations & Impact
All-time
Citations
1,169
 
H-index
16
 
i10-index
19
 
Publications
20
 
Co-authors
19
list available
Resume (English only)
Academic Achievements
  • Logit-based Log-sum-exp Aggregator for MIL-based WSI analysis accepted by BIBM 2025.
  • MaskMatch for semi-supervised segmentation accepted by Pattern Recognition.
  • MAPN for Unified Multi-anatomy MRI Reconstruction accepted by IEEE JBHI (IF=6.8).
  • GM-ABS for generalist-driven label efficiency accepted by IEEE TMI (IF=9.8).
  • MF-ConS for active barely-supervised 3D segmentation accepted by Medical Image Analysis (IF=11.8).
  • Two papers on continual learning and trustworthy report generation accepted by MICCAI 2025.
  • MMedAgent highlighted in Annual Progress Report on Al4Healthcare @ VALSE 2025!
  • Label-efficient abnormality detection and report generation accepted by IEEE TMI.
  • Scalable Multi-modal Medical Agent (MMedAgent) accepted by EMNLP 2024.
  • Survey on continual learning in MIA (PaperList) accepted by Computers in Biology and Medicine (IF=7).
  • D-Persona on multi-rater segmentation accepted by CVPR 2024 with Highlight (2.8%).
  • Trustworthy radiology report generation accepted by Neurocomputing (IF=6.0).
  • SCL for MS-SSL accepted by Medical Image Analysis (IF=10.9).
Background
  • Research Interests: Advancing medical image analysis with AI, focusing on learning under various imperfect data scenarios, multi-modality representation learning and decision making, human-centric personalized and structured analysis powered by LLM-based AI agents, and reliable multimodal health insights and prediction models for the general population.
Miscellany
  • Personal interests not mentioned