Zijie Fang

Zijie Fang

Auto-generated profile, based on publicly available information

Zijie Fang

Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen, China

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
26 June 2023 in Proceedings of the AAAI Conference on Artificial Intelligence
Verified
Journal Article
Open access
PDF file
ZFZijie Fang
Zijie Fang
YCYang Chen
Yang Chen
YWYifeng Wang
Yifeng Wang
ZWZhi Wang
Zhi Wang
XJXiangyang Ji
Xiangyang Ji
Yongbing ZhangYongbing Zhang
Yongbing Zhang
Tissue segmentation is a critical task in computational pathology due to its desirable ability to indicate the prognosis of cancer patients. Currently, numerous studies attempt to use image-level labels to achieve pixel-level segmentation to reduce the need for fine annotations. However, most of these methods are based on class activation map, which suffers from inaccurate segmentation boundaries. To address this problem, we propose a novel weakly-supervised tissue segmentation framework named PistoSeg, which is implemented under a fully-supervised manner by transferring tissue category labels to pixel-level masks. Firstly, a dataset synthesis method is proposed based on Mosaic transformation to generate synthesized images with pixel-level masks. Next, considering the difference between synthesized and real images, this paper devises an attention-based feature consistency, which directs the training process of a proposed pseudo-mask refining module. Finally, the refined pseudo-masks are used to train a precise segmentation model for testing. Experiments based on WSSS4LUAD and BCSS-WSSS validate that PistoSeg outperforms the state-of-the-art methods. The code is released at https://github.com/Vison307/PistoSeg.

13 June 2023 in IEEE Journal of Biomedical and Health Informatics
Verified
Journal Article
Closed access
YCYang Chen
Yang Chen
ZSZhuchen Shao
Zhuchen Shao
HBHao Bian
Hao Bian
ZFZijie Fang
Zijie Fang
YWYifeng Wang
Yifeng Wang
Yongbing ZhangYongbing Zhang
Yongbing Zhang
Automated classification of lymph node metastasis (LNM) plays an important role in the diagnosis and prognosis. However, it is very challenging to achieve satisfactory performance in LNM classification, because both the morphology and spatial distribution of tumor regions should be taken into account. To address this problem, this paper proposes a two-stage dMIL-Transformer framework, which integrates both the morphological and spatial information of the tumor regions based on the theory of multiple instance learning (MIL). In the first stage, a double Max-Min MIL (dMIL) strategy is devised to select the suspected top-K positive instances from each input histopathology image, which contains tens of thousands of patches (primarily negative). The dMIL strategy enables a better decision boundary for selecting the critical instances compared with other methods. In the second stage, a Transformer-based MIL aggregator is designed to integrate all the morphological and spatial information of the selected instances from the first stage. The self-attention mechanism is further employed to characterize the correlation between different instances and learn the bag-level representation for predicting the LNM category. The proposed dMIL-Transformer can effectively deal with the thorny classification in LNM with great visualization and interpretability. We conduct various experiments over three LNM datasets, and achieve 1.79%-7.50% performance improvement compared with other state-of-the-art methods.

19 December 2024 in IEEE Transactions on Medical Imaging
Verified
Journal Article
Closed access
ZFZijie Fang
Zijie Fang
YWYifeng Wang
Yifeng Wang
PXPeizhang Xie
Peizhang Xie
ZWZhi Wang
Zhi Wang
Yongbing ZhangYongbing Zhang
Yongbing Zhang
Tissue semantic segmentation is one of the key tasks in computational pathology. To avoid the expensive and laborious acquisition of pixel-level annotations, a wide range of studies attempt to adopt the class activation map (CAM), a weakly-supervised learning scheme, to achieve pixel-level tissue segmentation. However, CAM-based methods are prone to suffer from under-activation and over-activation issues, leading to poor segmentation performance. To address this problem, we propose a novel weakly-supervised semantic segmentation framework for histopathological images based on image-mixing synthesis and consistency regularization, dubbed HisynSeg. Specifically, synthesized histopathological images with pixel-level masks are generated for fully-supervised model training, where two synthesis strategies are proposed based on Mosaic transformation and Bézier mask generation. Besides, an image filtering module is developed to guarantee the authenticity of the synthesized images. In order to further avoid the model overfitting to the occasional synthesis artifacts, we additionally propose a novel self-supervised consistency regularization, which enables the real images without segmentation masks to supervise the training of the segmentation model. By integrating the proposed techniques, the HisynSeg framework successfully transforms the weakly-supervised semantic segmentation problem into a fully-supervised one, greatly improving the segmentation accuracy. Experimental results on three datasets prove that the proposed method achieves a state-of-the-art performance. Code is available at https://github.com/Vison307/HisynSeg.

13 December 2024 in IEEE Transactions on Circuits and Systems for Video Technology
Verified
Journal Article
Closed access
YZYe Zhang
Ye Zhang
YWYifeng Wang
Yifeng Wang
ZFZijie Fang
Zijie Fang
HBHao Bian
Hao Bian
LCLinghan Cai
Linghan Cai
Yongbing ZhangYongbing Zhang
Yongbing Zhang
Weakly supervised segmentation methods have garnered considerable attention due to their potential to alleviate the need for labor-intensive pixel-level annotations during model training. Traditional weakly supervised nuclei segmentation approaches typically involve a two-stage process: pseudo-label generation followed by network training. The performance of these methods is highly dependent on the quality of the generated pseudo-labels, which can limit their effectiveness. In this paper, we propose a novel domain-adaptive weakly supervised nuclei segmentation framework that addresses the challenge of pseudolabel generation through cross-task interaction strategies. Specifically, our approach leverages weakly annotated data to train an auxiliary detection task, which facilitates domain adaptation of the segmentation network. To improve the efficiency of domain adaptation, we introduce a consistent feature constraint module that integrates prior knowledge from the source domain. Additionally, we develop methods for pseudo-label optimization and interactive training to enhance domain transfer capabilities. We validate the effectiveness of our proposed method through extensive comparative and ablation experiments conducted on six datasets. The results demonstrate that our approach outperforms existing weakly supervised methods and achieves performance comparable to or exceeding that of fully supervised methods. Our code is available at https://github.com/zhangye-zoe/DAWN.

16 September 2024 in IEEE Transactions on Medical Imaging
Verified
Journal Article
Closed access
YLYiyang Lin
Yiyang Lin
YWYifeng Wang
Yifeng Wang
ZFZijie Fang
Zijie Fang
ZLZexin Li
Zexin Li
XGXianchao Guan
Xianchao Guan
Yongbing ZhangYongbing Zhang
Yongbing Zhang
In clinical practice, frozen section (FS) images can be utilized to obtain the immediate pathological results of the patients in operation due to their fast production speed. However, compared with the formalin-fixed and paraffin-embedded (FFPE) images, the FS images greatly suffer from poor quality. Thus, it is of great significance to transfer the FS image to the FFPE one, which enables pathologists to observe high-quality images in operation. However, obtaining the paired FS and FFPE images is quite hard, so it is difficult to obtain accurate results using supervised methods. Apart from this, the FS to FFPE stain transfer faces many challenges. Firstly, the number and position of nuclei scattered throughout the image are hard to maintain during the transfer process. Secondly, transferring the blurry FS images to the clear FFPE ones is quite challenging. Thirdly, compared with the center regions of each patch, the edge regions are harder to transfer. To overcome these problems, a multi-perspective self-supervised GAN, incorporating three auxiliary tasks, is proposed to improve the performance of FS to FFPE stain transfer. Concretely, a nucleus consistency constraint is designed to enable the high-fidelity of nuclei, an FFPE guided image deblurring is proposed for improving the clarity, and a multi-field-of-view consistency constraint is designed to better generate the edge regions. Objective indicators and pathologists’ evaluation for experiments on the five datasets across different countries have demonstrated the effectiveness of our method. In addition, the validation in the downstream task of microsatellite instability prediction has also proved the performance improvement by transferring the FS images to FFPE ones. Our code link is https://github.com/linyiyang98/Self-Supervised-FS2FFPE.git.

4 August 2021 in Wireless Networks
Verified
Journal Article
Closed access
PDF file
ZFZijie Fang
Zijie Fang
XXXiaolong Xu
Xiaolong Xu
Muhammad BilalMuhammad Bilal
Muhammad Bilal
AJAlireza Jolfaei
Alireza Jolfaei