Search

Your search keyword '"Li, Lingrui"' showing total 127 results
Start Over Author "Li, Lingrui"
127 results on '"Li, Lingrui"'

1. XNet v2: Fewer Limitations, Better Results and Greater Universality

Author

Zhou, Yanfeng, Li, Lingrui, Wang, Zichen, Liu, Guole, Liu, Ziwen, and Yang, Ge

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Artificial Intelligence
Abstract

XNet introduces a wavelet-based X-shaped unified architecture for fully- and semi-supervised biomedical segmentation. So far, however, XNet still faces the limitations, including performance degradation when images lack high-frequency (HF) information, underutilization of raw images and insufficient fusion. To address these issues, we propose XNet v2, a low- and high-frequency complementary model. XNet v2 performs wavelet-based image-level complementary fusion, using fusion results along with raw images inputs three different sub-networks to construct consistency loss. Furthermore, we introduce a feature-level fusion module to enhance the transfer of low-frequency (LF) information and HF information. XNet v2 achieves state-of-the-art in semi-supervised segmentation while maintaining competitve results in fully-supervised learning. More importantly, XNet v2 excels in scenarios where XNet fails. Compared to XNet, XNet v2 exhibits fewer limitations, better results and greater universality. Extensive experiments on three 2D and two 3D datasets demonstrate the effectiveness of XNet v2. Code is available at https://github.com/Yanfeng-Zhou/XNetv2 .

Published
2024

3. Robust Source-Free Domain Adaptation for Fundus Image Segmentation

Author

Li, Lingrui, Zhou, Yanfeng, and Yang, Ge

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Unsupervised Domain Adaptation (UDA) is a learning technique that transfers knowledge learned in the source domain from labelled training data to the target domain with only unlabelled data. It is of significant importance to medical image segmentation because of the usual lack of labelled training data. Although extensive efforts have been made to optimize UDA techniques to improve the accuracy of segmentation models in the target domain, few studies have addressed the robustness of these models under UDA. In this study, we propose a two-stage training strategy for robust domain adaptation. In the source training stage, we utilize adversarial sample augmentation to enhance the robustness and generalization capability of the source model. And in the target training stage, we propose a novel robust pseudo-label and pseudo-boundary (PLPB) method, which effectively utilizes unlabeled target data to generate pseudo labels and pseudo boundaries that enable model self-adaptation without requiring source data. Extensive experimental results on cross-domain fundus image segmentation confirm the effectiveness and versatility of our method. Source code of this study is openly accessible at https://github.com/LinGrayy/PLPB., Comment: 10 pages, WACV2024

Published
2023

5. Application of fused graphical lasso to statistical inference for multiple sparse precision matrices

Author

Zhang, Qiuyan, Bai, Zhidong, Li, Lingrui, and Yang, Hu

Subjects
Mathematics - Statistics Theory
Abstract

In this paper, the fused graphical lasso (FGL) method is used to estimate multiple precision matrices from multiple populations simultaneously. The lasso penalty in the FGL model is a restraint on sparsity of precision matrices, and a moderate penalty on the two precision matrices from distinct groups restrains the similar structure across multiple groups. In high-dimensional settings, an oracle inequality is provided for FGL estimators, which is necessary to establish the central limit law. We not only focus on point estimation of a precision matrix, but also work on hypothesis testing for a linear combination of the entries of multiple precision matrices. Inspired by Jankova a and van de Geer [confidence intervals for high-dimensional inverse covariance estimation, Electron. J. Stat. 9(1) (2015) 1205-1229.], who investigated a de-biasing technology to obtain a new consistent estimator with known distribution for implementing the statistical inference, we extend the statistical inference problem to multiple populations, and propose the de-biasing FGL estimators. The corresponding asymptotic property of de-biasing FGL estimators is provided. A simulation study shows that the proposed test works well in high-dimensional situations.

Published
2023

10. The effect of high-Z dopant on the ablation of carbon–hydrogen polymer target.

Author

Xiong, Wei, Yang, Xiaohu, Zhang, Guobo, Chen, Zehao, Cui, Ye, Zeng, Bo, Li, Ze, Li, Lingrui, Xu, Bihao, Dong, Yulong, and Ma, Yanyun

Subjects
DOPING agents (Chemistry), INERTIAL confinement fusion, POLYMERS, ELECTRON-ion collisions, HOT carriers, SILICON polymers, BROMINE
Abstract

High-Z dopants such as chlorine, bromine and silicon in carbon–hydrogen polymer (CH) targets play a crucial role during the ablation of inertial confinement fusion (ICF). These dopants can serve as diagnostic tools in experiments and mitigate hot electron preheating, but they also influence the laser ablation. In this paper, the process of high-power laser ablating doped CH targets has been studied through radiation hydrodynamic simulations. Our findings reveal that the laser absorption rate in the doped targets increase as a result of the increasing electron-ion collision frequency. This leads to the increase of the electron, ion and radiation temperatures. Furthermore, high-Z dopants contribute to a decrease in the ablation pressure, which tends to a constant. Moreover, the saturation phenomenon of the mass ablation rate has been found. For the targets with low doping ratios (e.g. 6.25%–12.5%), the mass ablation rate increases until reaching the saturation at a doping ratio of 18.75%, after which it decreases. This indicates that an appropriate doping ratio can increase the laser absorption and ablation. The results are helpful to comprehensively understand the effects of high-Z dopant on all stages of ICF. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

15. Application of fused graphical lasso to statistical inference for multiple sparse precision matrices.

Author

Zhang, Qiuyan, Li, Lingrui, and Yang, Hu

Subjects
*SPARSE matrices, *DIFFUSE large B-cell lymphomas
Abstract

In this paper, the fused graphical lasso (FGL) method is used to estimate multiple precision matrices from multiple populations simultaneously. The lasso penalty in the FGL model is a restraint on sparsity of precision matrices, and a moderate penalty on the two precision matrices from distinct groups restrains the similar structure across multiple groups. In high-dimensional settings, an oracle inequality is provided for FGL estimators, which is necessary to establish the central limit law. We not only focus on point estimation of a precision matrix, but also work on hypothesis testing for a linear combination of the entries of multiple precision matrices. We apply a de-biasing technology, which is used to obtain a new consistent estimator with known distribution for implementing the statistical inference, and extend the statistical inference problem to multiple populations. The corresponding de-biasing FGL estimator and its asymptotic theory are provided. A simulation study and an application of the diffuse large B-cell lymphoma data show that the proposed test works well in high-dimensional situation. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

19. The Mechanism of Hepatic Encephalopathy Induced by Thioacetamide Based on Metabolomics and Proteomics: A Preliminary Study.

Author

Guo, Honghui, Wang, Guang, Huang, Wei, Li, Lingrui, Bai, Yang, Wang, Haifeng, and Gao, Lina

Subjects
HEPATIC encephalopathy, PROTEOMICS, STAPHYLOCOCCUS aureus infections, METABOLOMICS, THIOACETAMIDE
Abstract

Hepatic encephalopathy (HE) is a central nervous system dysfunction syndrome caused by acute and chronic liver failure or various portal systemic shunt disorders. HE arises from metabolic disorder and excludes other known types of encephalopathy. HE is a major cause of death in people with liver disease. Early diagnosis and timely treatment are key to improving HE prognosis. Herein, we established a model of HE and performed metabolomics to identify 50 significantly differential metabolites between the HE group and control group. The main metabolic pathways associated with these differential metabolites were the purine metabolism, pyrimidine metabolism, aminoacyl tRNA biosynthesis, and glucose metabolism. Through proteomics analysis, we identified 226 significantly differential proteins (52 up-regulated and 174 down-regulated). The main (Kyoto Encyclopedia of Genes and Genomes) enrichment pathways were the Staphylococcus aureus infection, vitamin digestion and absorption, and complement and coagulation cascades. Through the conjoint analysis of proteomics and metabolomics, the differentially present proteins and metabolites were found to be involved in vitamin digestion and absorption, and ferroptosis pathways. In HE, malondialdehyde was significantly elevated, but glutathione was significantly diminished, and the redox balance was destroyed, thus leading to changes in proteins' levels associated with the ferroptosis pathway. In conclusion, this study preliminarily explored the molecular and metabolic mechanisms underlying HE. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results