Publications
A collection of my research work.
Knowledge-Guided Adaptation of Pathology Foundation Models Effectively Improves Cross-domain Generalization and Demographic Fairness
Yanyan Huang, Weiqin Zhao, Zhengyu Zhang, Yihang Chen, Yu Fu, Feng Wu, Yuming Jiang, Li Liang, Shujun Wang, Lequan Yu†
Nature Communications
2025This work presents a knowledge-guided adaptation framework that leverages task-specific information bottlenecks to disentangle robust pathological features from site-specific and demographic artifacts, achieving superior cross-domain generalization and fairness in computational pathology.
Bridging Radiological Images and Factors with Vision-Language Model for Accurate Diagnosis of Proliferative Hepatocellular Carcinoma
Yanyan Huang, Wanli Zhang, Peixiang Huang, Yu Fu, Ruimeng Yang, Lequan Yu†
International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI)
2025This work introduces a vision-language foundation model-based approach that bridges radiological images and clinical factors through aligned embedding spaces, enabling effective multimodal fusion for accurate diagnosis of proliferative hepatocellular carcinoma.
HyperPath: Knowledge-Guided Hyperbolic Semantic Hierarchy Modeling for WSI Analysis
Peixiang Huang, Yanyan Huang, Weiqin Zhao, Junjun He, Lequan Yu
International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI)
2025This work introduces a hyperbolic geometry-based approach that leverages textual knowledge from vision-language models to model semantic hierarchies in whole slide images, employing geodesic distance-based classification to better capture hierarchical relationships and improve diagnostic accuracy.
Towards Multi-scenario Generalization: Text-Guided Unified Framework for Low-Dose CT and Total-Body PET Reconstruction
Weitao Wang, Yanyan Huang, Shunjie Dong, Le Xue, Kuangyu Shi, Yu Fu
International Conference on Medical Image Computing and Computer-Assisted Intervention
2025This work presents a text-guided unified framework inspired by cold diffusion paradigms that employs mean-preserving degradation operators and dual-domain fusion networks to achieve high-quality reconstruction across low-dose CT and total-body PET imaging with enhanced cross-scenario generalization.
Free lunch in pathology foundation model: Task-specific model adaptation with concept-guided feature enhancement
Yanyan Huang, Weiqin Zhao, Yihang Chen, Yu Fu, Lequan Yu†
Advances in Neural Information Processing Systems (NeurIPS)
2024This work proposes a concept-guided feature enhancement paradigm that dynamically calibrates pathology foundation models through task-specific concept anchors, improving feature expressivity and discriminativeness for downstream clinical tasks while maintaining strong generalizability.
Unleash the Power of State Space Model for Whole Slide Image With Local Aware Scanning and Importance Resampling
Yanyan Huang, Weiqin Zhao, Yu Fu, Lingting Zhu, Lequan Yu†
IEEE Transactions on Medical Imaging (TMI)
2024This work presents a state space model-based framework for whole slide image analysis that incorporates local-aware hierarchical scanning and test-time importance resampling to efficiently process gigapixel pathology images while maintaining high accuracy and robustness.
MPGAN: Multi Pareto Generative Adversarial Network for the denoising and quantitative analysis of low-dose PET images of human brain
Yu Fu, Shunjie Dong, Yanyan Huang, Meng Niu, Chao Ni, Lequan Yu, Kuangyu Shi, Zhijun Yao, Cheng Zhuo
Medical Image Analysis (MedIA)
2024This work introduces a multi-Pareto generative adversarial network that integrates diffused multi-round cascade generation with dynamic Pareto-efficient discrimination to achieve clinically-viable 3D end-to-end denoising of low-dose PET brain images while preserving quantitative accuracy.
Sex-dependent nonlinear Granger connectivity patterns of brain aging in healthy population
Yu Fu, Le Xue, Meng Niu, Yuanhang Gao, Yanyan Huang, Hong Zhang, Mei Tian, Cheng Zhuo
Progress in Neuro-Psychopharmacology and Biological Psychiatry (PNP)
2024This work employs neural Granger causality techniques to investigate sex-dependent nonlinear connectivity patterns during brain aging, revealing that females exhibit greater heterogeneity and reduced stability in causal connectivity, particularly in visual network interactions, with implications for understanding sex differences in neurological disorders.
Conslide: Asynchronous hierarchical interaction transformer with breakup-reorganize rehearsal for continual whole slide image analysis
Yanyan Huang, Weiqin Zhao, Shujun Wang, Yu Fu, Yuming Jiang, Lequan Yu†
Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)
2023This work introduces a continual learning framework for whole slide image analysis that employs hierarchical interaction transformers and breakup-reorganize rehearsal strategies to mitigate catastrophic forgetting while adapting to sequential datasets from evolving imaging technologies.
OTFPF: Optimal transport based feature pyramid fusion network for brain age estimation
Yu Fu#, Yanyan Huang#, Zhe Zhang, Shunjie Dong, Le Xue, Meng Niu, Yunxin Li, Zhiguo Shi, Yalin Wang, Hong Zhang, Mei Tian, Cheng Zhuo
Information Fusion
2023This work proposes an optimal transport-based feature pyramid fusion network that effectively integrates semi-multimodal and multi-level features from T1-weighted MRI scans, achieving highly accurate brain age estimation with enhanced interpretability for understanding brain development and aging.
HDNet: Hierarchical dynamic network for gait recognition using millimeter-wave radar
Yanyan Huang, Yong Wang, Kun Shi, Chaojie Gu, Yu Fu, Cheng Zhuo, Zhiguo Shi
ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
2023This work presents a hierarchical dynamic network for millimeter-wave radar-based gait recognition that introduces point flow as a novel descriptor and employs dynamic frame sampling to efficiently capture temporal dynamics while maintaining robust performance in challenging environments.
SFCNeXt: a simple fully convolutional network for effective brain age estimation with small sample size
Yu Fu, Yanyan Huang, Shunjie Dong, Yalin Wang, Tianbai Yu, Meng Niu, Cheng Zhuo
2023 IEEE 20th International Symposium on Biomedical Imaging (ISBI)
2023This work proposes a lightweight fully convolutional network that combines single pathway encoded ConvNeXt architecture with hybrid ranking loss to enable effective brain age estimation in small-sized cohorts with biased age distributions, reducing computational requirements while maintaining high accuracy.
AIGAN: Attention--encoding Integrated Generative Adversarial Network for the reconstruction of low-dose CT and low-dose PET images
Yu Fu, Shunjie Dong, Meng Niu, Le Xue, Hanning Guo, Yanyan Huang, Yuanfan Xu, Tianbai Yu, Kuangyu Shi, Qianqian Yang, Yiyu Shi, Hong Zhang, Mei Tian, Cheng Zhuo
Medical Image Analysis
2023This work introduces an attention-encoding integrated generative adversarial network that employs cascade generation with dual-scale discrimination and multi-scale spatial fusion to achieve universal and clinically-viable reconstruction of low-dose CT and PET images while reducing radiation exposure.
Altered nonlinear Granger causality interactions in the large-scale brain networks of patients with schizophrenia
Yu Fu, Meng Niu, Yuanhang Gao, Shunjie Dong, Yanyan Huang, Zhe Zhang, Cheng Zhuo
Journal of Neural Engineering
2022This work applies neural Granger causality techniques to reveal extensive alterations in nonlinear causal interactions across large-scale brain networks in schizophrenia patients, providing novel insights into the functional dysconnectivity patterns underlying the pathophysiology of this disorder through both static and dynamic network analyses.
Active index: an integrated index to reveal disrupted brain network organizations of major depressive disorder patients
Yu Fu, Yanyan Huang, Meng Niu, Le Xue, Shunjie Dong, Shunlin Guo, Junqiang Lei, Cheng Zhuo
2022 IEEE 19th international symposium on biomedical imaging (ISBI)
2022This work introduces a novel active index metric that integrates rich club and diverse club analyses to characterize brain network organization, revealing complementary insights into functional integration and segregation patterns that effectively distinguish major depressive disorder patients from healthy controls.