Selected Research

Click a figure to enlarge
SmartStu framework
MICCAI 2026Early Accept · First author

SmartStu

Harnessing Adversarial Distillation to Customise Debiased, Disease-Specific Pathology Foundation Models for Breast Cancer
30× compression; surpasses the teacher (AUC 0.994)
arXivCode
G-Interp framework
MICCAI 2026 WSOral · Second author

G-Interp

Cellular-Communication-Level Interpretability for Pathology Foundation Models via Graph Distillation on Microenvironment
Cell-level interpretability with 20× compression
arXiv
CNRL framework
IJCNN 2025Third author

CNRL

Cross-Network Relationship Learning via Optimal Transport for Link Prediction
Optimal transport for cross-domain link prediction
IEEE