Shengyuan Liu
Scholar

Shengyuan Liu

Google Scholar ID: zP6fRqcAAAAJ
The Chinese University of Hong Kong; CASIA
Multimodal LearningGenerative modelsAI for HealthcareRadiomics
Homepage↗Google Scholar↗
Citations & Impact
All-time
Citations
116
 
H-index
5
 
i10-index
3
 
Publications
16
 
Co-authors
0
 
Contact
Emailliushengyuan@link.cuhk.edu.hkTwitterOpen ↗GitHubOpen ↗LinkedInOpen ↗
Publications
4 items
Topology-Guided Biomechanical Profiling: A White-Box Framework for Opportunistic Screening of Spinal Instability on Routine CT
2026
Cited
0
MedPruner: Training-Free Hierarchical Token Pruning for Efficient 3D Medical Image Understanding in Vision-Language Models
2026
Cited
0
MedSAM-Agent: Empowering Interactive Medical Image Segmentation with Multi-turn Agentic Reinforcement Learning
2026
Cited
0
Unveiling and Bridging the Functional Perception Gap in MLLMs: Atomic Visual Alignment and Hierarchical Evaluation via PET-Bench
arXiv.org · 2026
Cited
0
Resume (English only)
Background
  • Research interests: Generative Modeling and AI for Healthcare, especially multi-modality joint representation learning for medical vision-language Models and Agents.
Miscellany
  • Driven by curiosity and enthusiasm in research and life. A sports enthusiast and a die-hard fan of Leo Messi. Spent 4 amazing years with the XJTU football team and passionate about basketball, jogging, skiing, hiking, swimming, and tennis.
Co-authors
0 total
Co-authors: 0 (list not available)