检索结果-内蒙古大学图书馆

arXiv 2020年

作者： Yan, Cheng Pang, Guansong Bai, Xiao Zhou, Jun Gu, Lin School of Computer Science and Engineering Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University China School of Computer Science University of Adelaide Austria School of Information and Communication Technology Griffith University Australia National Institute of Informatics University of Tokyo Japan

Person Re-IDentification (ReID) aims at re-identifying persons from different viewpoints across multiple cameras. Capturing the fine-grained appearance differences is often the key to accurate person ReID, because many identities can be differentiated only when looking into these fine-grained differences. However, most state-of-the-art person ReID approaches, typically driven by a triplet loss, fail to effectively learn the fine-grained features as they are focused more on differentiating large appearance differences. To address this issue, we introduce a novel pairwise loss function that enables ReID models to learn the fine-grained features by adaptively enforcing an exponential penalization on the images of small differences and a bounded penalization on the images of large differences. The proposed loss is generic and can be used as a plugin to replace the triplet loss to significantly enhance different types of state-of-the-art approaches. Experimental results on four benchmark datasets show that the proposed loss substantially outperforms a number of popular loss functions by large margins;and it also enables significantly improved data efficiency. Copyright © 2020, The Authors. All rights reserved.

关键词： Large dataset

来源：评论

学校读者我要写书评

暂无评论

Increased urbanization reduced the efectiveness of school closures on seasonal infuenza epidemics in China

引用

Infectious Diseases of Poverty 2021年第5期10卷 52-61页

作者： Hao Lei Hangjin Jiang Nan Zhang Xiaoli Duan Tao Chen Lei Yang Dayan Wang Yuelong Shu School of Public Health Zhejiang UniversityHangzhouPeople’s Republic of China Center for Data Science Zhejiang UniversityHangzhouPeople’s Republic of China Key Laboratory of Green Built Environment and Energy Efcient Technology Beijing University of TechnologyBeijingPeople’s Republic of China School of Energy and Environmental Engineering University of Science and Technology BeijingBeijingChina National Institute for Viral Disease Control and Prevention Collaboration Innovation Center for Diagnosis and Treatment of Infectious DiseasesChinese Center for Disease Control and Prevention/Key Laboratory for Medical VirologyNational Health CommissionBeijing 102206People’s Republic of China School of Public Health(Shenzhen) Shenzhen Campus of Sun Yat-Sen UniversityNo.66Gongchang RoadGuangming DistrictShenzhenGuangdong 518107People’s Republic of China.

Background:School closure is a common mitigation strategy during severe infuenza epidemics and ***,the efectiveness of this strategy remains *** this study,we aimed to explore the efectiveness of school closure on seasonal infuenza epidemics in provincial-level administrative divisions(PLADs)with varying urbanization rates in ***:This study analyzed infuenza surveillance data between 2010 and 2019 provided by the Chinese National Infuenza *** into consideration the climate,this study included a region with 3 adjacent PLADs in Northern China and another region with 4 adjacent PLADs in Southern *** efect of school closure on infuenza transmission was evaluated by the reduction of the efective reproductive number of seasonal infuenza during school winter breaks compared with that before school winter *** age-structured Susceptible-Infected-Recovered-Susceptible(SIRS)model was built to model infuenza transmission in diferent levels of *** were determined using the surveillance data via robust Bayesian ***:Between 2010 and 2019,in the less urbanized provinces:Hebei,Zhejiang,Jiangsu and Anhui,during school winter breaks,the efective reproductive number of seasonal infuenza epidemics reduced 14.6%[95%confdential interval(CI):6.2–22.9%],9.6%(95%CI:2.5–16.6%),7.3%(95%CI:0.1–14.4%)and 8.2%(95%CI:1.1–15.3%)***,in the highly urbanized cities:Beijing,Tianjin and Shanghai,it reduced only 5.2%(95%CI:-0.7–11.2%),4.1%(95%CI:-0.9–9.1%)and 3.9%(95%CI:-1.6–9.4%)*** China,urbanization is associated with decreased proportion of children and increased social *** to the SIRS model,both factors could reduce the impact of school closure on seasonal infuenza epidemics,and the proportion of children in the population is thought to be the dominant infuencing ***:Efectiveness of school closure on the epidemics varies with the age structure in the pop

关键词： School closure Infuenza epidemics Urbanization Contact China

来源：评论

学校读者我要写书评

暂无评论

Single-cell landscape of innate and acquired drug resistance in acute myeloid leukemia

引用

Nature Communications 2024年第1期15卷 9402-9402页

作者： Wegmann, Rebekka Bonilla, Ximena Casanova, Ruben Chevrier, Stéphane Coelho, Ricardo Esposito, Cinzia Ficek-Pascual, Joanna Goetze, Sandra Gut, Gabriele Jacob, Francis Jacobs, Andrea Kuipers, Jack Lischetti, Ulrike Mena, Julien Milani, Emanuela S. Prummer, Michael Del Castillo, Jacobo Sarabia Singer, Franziska Sivapatham, Sujana Toussaint, Nora C. Vilinovszki, Oliver Wildschut, Mattheus H. E. Thavayogarajah, Tharshika Malani, Disha Zuend, Gregor Zoche, Martin Zimmermann, Marc Ziegler, Johanna Yu, Shuqing Wey, Norbert Wendt, Fabian Weller, Michael Weber, Walter P. Weber, Sandra Vlajnic, Tatjana Vetter, Marcus Van Drogen, Audrey Tusup, Marina Tuncel, Mustafa A. Tosevski, Vinko Tolnay, Markus Thomas, Tinu M. Tanna, Tanmay Stekhoven, Daniel J. Stark, Stefan Sreedharan, Vipin T. Sobottka, Bettina Snijder, Berend Shao, Wenguang Senti, Gabriela Schär, Tobias Schwan, Severin Schwalie, Petra C. Schlenker, Ramona Santacroce, Natascha Rätsch, Gunnar Rosano-González, María L. Rommel, Christian Ritter, Mathilde Rimmer, Natalie Prélot, Laurie Piscuoglio, Salvatore Pelkmans, Lucas Pedrioli, Patrick G. A. Lopez, Monica Nunez Nowak, Marta Nicolet, Stefan Ng, Charlotte K. Y. Murri, Riccardo Muenst, Simone Moch, Holger Miglino, Nicola Metzler, Julian M. Mehnert, Martin Markolin, Philipp Maass, Gerd Lugert, Sebastian Lombardo, Flavio C. Levesque, Mitchell Lehmann, Kjong-Van Kurzeder, Christian Kunze, Christian P. Kuebler, Werner Koelzer, Viktor H. Kahraman, Abdullah Kahles, André Jermann, Philip M. James, Alva R. Jahn, Katharina Jaeger, Tim M. Irmisch, Anja Immer, Alexander Huesser, Tamara Holtackers, Rene Herter, Sylvia Heinzelmann-Schwarz, Viola Haeuptle, Pirmin Günther, Detlef Grob, Linda Frey, Bruno Frei, Anja L. Ferreira, Pedro F. Fedier, André Eschbach, Katja Erkens, Martin Engler, Stefanie Dummer, Reinhard Drăgan, Monica-Andreea Davidson, Natalie R. Danenberg, Esther D’Costa, Maya Chicherova, Natalia Calgua, Byron Bosshard, Lars Bodenmiller, Bernd Bertolini, Anne Bernasconi, Lara Beisel, Christian Beerenwinkel, Niko Beck-Schim Institute of Molecular Systems Biology Department of Biology ETH Zurich Zurich Switzerland Department of Computer Science ETH Zurich Zurich Switzerland Department of Quantitative Biomedicine University of Zurich Zurich Switzerland Department of Biomedicine University Hospital Basel and University of Basel Basel Switzerland Institute of Molecular Life Sciences University of Zurich Zurich Switzerland Department of Health Sciences and Technology ETH Zurich Zurich Switzerland ETH PHRT Swiss Multi-Omics Center (SMOC) Lausanne Switzerland SIB Swiss Institute of Bioinformatics Lausanne Switzerland Department of Biosystems Science and Engineering ETH Zurich Zurich Switzerland NEXUS Personalized Health Technologies ETH Zurich Zurich Switzerland Swiss Data Science Center ETH Zürich Zurich Switzerland Department of Medical Oncology and Hematology University Hospital Zurich Zurich Switzerland Harvard Medical School and Dana-Farber Cancer Institute Boston United States Roche Pharmaceutical Research and Early Development Roche Innovation Center Zurich Zurich Switzerland Department of Pathology and Molecular Pathology University Hospital Zurich Zurich Switzerland University of Zurich Faculty of Medicine Zurich Switzerland Department of Dermatology University Hospital Zurich Zurich Switzerland AI Center at ETH Zurich Zurich Switzerland Institute of Medical Genetics and Pathology University Hospital Basel Basel Switzerland

Deep single-cell multi-omic profiling offers a promising approach to understand and overcome drug resistance in relapsed or refractory (rr) acute myeloid leukemia (AML). Here, we combine single-cell ex vivo drug profiling (pharmacoscopy) with single-cell and bulk DNA, RNA, and protein analyses, alongside clinical data from 21 rrAML patients. Unsupervised data integration reveals reduced ex vivo response to the Bcl-2 inhibitor venetoclax (VEN) in patients treated with both a hypomethylating agent (HMA) and VEN, compared to those pre-exposed to chemotherapy or HMA alone. Integrative analysis identifies both known and unreported mechanisms of innate and treatment-related VEN resistance and suggests alternative treatments, like targeting increased proliferation with the PLK inhibitor volasertib. Additionally, high CD36 expression in VEN-resistant blasts associates with sensitivity to CD36-targeted antibody treatment ex vivo. This study demonstrates how single-cell multi-omic profiling can uncover drug resistance mechanisms and treatment vulnerabilities, providing a valuable resource for future AML research. © 2024. The Author(s).

关键词：

来源：评论

学校读者我要写书评

暂无评论

Improving active learning by data balance to reduce annotation efforts

Improving active learning by data balance to reduce annotati...

引用

7th International Symposium on Test Automation and Instrumentation, ISTAI 2018

作者： Lei, Han Wang, Shuai Zheng, Dezhi Qu, Xiaolei Fan, Shangchun Cui, Chongyang School of Instrumentation Science and Opto-electronics Engineering Beihang University Beijing100191 China Beijing Advanced Innovation Center for Big Data-based Precision Medicine Beihang University Beijing100191 China School of Computer Science and Engineering Beihang University Beijing100191 China

Image classification is a fundamental task in image analysis. Recent advances in deep learning have achieved promising results on many image classification benchmarks. However, in some particular tasks, especially in biomedical image analysis, preparing a large number of labelled images for the model's training is costly and unpractical. In this study, the authors aim to address the following questions: With limited effort (e.g. time, cost and manpower) for labelling, what instances should be chosen to annotate and how to train to model using limited annotated data. For that, they present an active learning algorithm combining with data balancing, making the model (e.g. convolutional neural network) fine-tuned continuously and incrementally to reduce the effort of labelling and making model's training process more robust and efficient in both binary and multi-class classification with high-performance. They have evaluated the authors’ method of both binary natural dataset and three classes biomedical dataset, demonstrating that active learning with data balancing could help models’ training more robust and broaden active learning's field to multi classification and more application scenarios. More significantly, their experiments showed that at least a half of effort in labelling could be saved for satisfied performance by their method. © 2019 Institution of engineering and Technology. All rights reserved.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

Evaluation of TRMM 3B43 V7 precipitation data in varied Moroccan climatic and topographic zones

引用

Mediterranean Geoscience Reviews 2024年第2期6卷 159-175页

作者： Aqnouy, Mourad Ommane, Younes Ouallali, Abdessalam Gourfi, Abdelali Ayele, Gebiaw T. El Yousfi, Yassine Bouizrou, Ismail Stitou El Messari, Jamal Eddine Zettam, Amin Melesse, Assefa M. Ndehedehe, Christopher E. Abioui, Mohamed Briak, Hamza Applied Geology Research Laboratory AGRSRT Department of Geosciences Faculty of Sciences and Techniques Moulay Ismail University of Meknes Errachidia Morocco Modelling Simulation and Data Analysis Mohammed VI Polytechnic University (UM6P) Ben Guerir Morocco GPE Laboratory Faculty of Sciences and Techniques Hassan II University of Mohammedia Mohammedia Morocco ESO UMR 6590 UNIV Angers CNRS SFR Confluences UFR LLSH Angers France Australian Rivers Institute and School of Engineering and Built Environment Griffith University Nathan Australia Research Team: Water and Environment Management Laboratory of Applied Sciences (LSA) National School of Applied Sciences Al Hoceima Abdelmalek Essaadi University Tétouan Morocco Functional Ecology and Environmental Engineering Laboratory Faculty of Sciences and Techniques Sidi Mohamed Ben Abdellah University Fez Morocco Department of Agriculture Food Environment and Forestry (DAGRI) University of Florence Florence Italy Department of Geology Faculty of Sciences Abdelmalek Essaadi University Tétouan Morocco LECGEN SNVSTU Abou Bekr Belkaid University (UABB) Tlemcen Algeria Department of Earth and Environment Florida International University Miami FL United States School of Environment and Science Griffith University Nathan QLD Australia Geosciences Environment and Geomatics Laboratory (GEG) Department of Earth Sciences Faculty of Sciences Ibnou Zohr University Agadir Morocco MARE-Marine and Environmental Sciences Centre-Sedimentary Geology Group Department of Earth Sciences Faculty of Sciences and Technology University of Coimbra Coimbra Portugal Laboratory for Sustainable Innovation and Applied Research Universiapolis—International University of Agadir Agadir Morocco Center of Excellence for Soil and Fertilizer Research in Africa (CESFRA) AgroBioSciences (AgBS) Mohammed VI Polytechnic University (UM6P) Ben Guerir Morocco

Hydrologists in developing countries face a major challenge due to the lack of ground data. This study evaluates the performance of the monthly Tropical Rainfall Measurement Mission (TRMM) algorithm 3B43 V7 using 172 ground rainfall measurement stations (GRMS) over Morocco (1998–2018). The performance of the TRMM 3B43 V7 was assessed using four well-known metrics (KGE, R², PBIAS, and RSR). The evaluation was undertaken based on (1) a point-to-pixel scale, (2) sub-regions of Morocco, (3) climatic zones, and (4) different ranges of altitudes. Results indicate that the TRMM 3B43 V7 provides satisfactory attribution in most regions of Morocco, with an average R² higher than 0.54, except in the south areas, which have the lowest R² between 0.29 and 0.41. The best R² of 0.95 was observed in the Tangier-Tetouan-Al Hoceima region. The other performance criteria (KGE: 0.93, RSR: 0.21, PBIAS: − 5.5) indicate that TRMM 3B43 V7 is consistent with in situ data but slightly underestimated. At the point-to-pixel scale, KGE and R² show that 61% and 52% of stations present satisfactory to good results, respectively. The PBIAS shows a large overestimation over most parts of Morocco, particularly the southern part. TRMM 3B43V7 performs better in the Mediterranean climate than in other climatic zones. TRMM 3B43 V7 harmonises GRMS slightly in low elevation zones, while average PBIAS show a global overestimation, increasing in high and low elevations. The results of this study confirm that TRMM 3B43 V7 provides a vital baseline tool to assist Moroccan Hydraulic Basin agencies. Graphical abstract: (Figure presented.) © The Author(s), under exclusive licence to Springer Nature Switzerland AG 2024.

关键词： GRMS Morocco Rainfall Topography TRMM 3B43 V7

来源：评论

学校读者我要写书评

暂无评论

Comparison of immune profiles between hepatocellular carcinoma subtypes

引用

Biophysics Reports 2020年第1期6卷 19-32页

作者： Xuemin Pan Ping Lin Fangyoumin Feng Jia Li Yuan-Yuan Li Wentao Dai Bo Hu Xin-Rong Yang Jia Fan Hong Li Yixue Li School of Life Sciences and Biotechnology Shanghai Jiao Tong University Bio-Med Big Data Center Key Laboratory of Computational Biology CAS-MPG Partner Institute for Computational Biology Shanghai Institute of Nutrition and Health Shanghai Institutes for Biological Sciences University of Chinese Academy of Sciences Chinese Academy of Sciences University of the Chinese Academy of Sciences School of Life Science and Techonology Shanghai Tech University Shanghai Center for Bioinformation Technology Shanghai Academy of Science & Technology Shanghai Engineering Research Center of Pharmaceutical Translation Department of Liver Surgery Liver Cancer Institute Zhongshan Hospital Fudan University Key Laboratory of Carcinogenesis and Cancer Invasion Ministry of Education Collaborative Innovation Center of Genetics and Development Fudan University

Immunotherapy, especially immune checkpoint inhibitors, is becoming a promising treatment for hepatocellular carcinoma(HCC). However, the response rate remains limited due to the heterogeneity of HCC samples. Molecular subtypes of HCC vary in genomic background, clinical features, and prognosis. This study aims to compare the immune profiles between HCC subtypes and find subtype-specific immune characteristics that might contribute to the prognosis and potential of immunotherapy. The immune profiles consist of immune-related genes, cytolytic activity, immune pathways, and tumorinfiltrating lymphocytes. HCC-c1 samples showed an overall higher activation level of immune genes and pathways, and this pattern was consistent in validation sets. We associated the difference in immune profiles with the activation level of cancer hallmarks and genomic mutations. There was a negative correlation between most of the metabolism pathway and immune-related pathways in HCC samples. CTNNB1/WNT signaling pathway mutation, one of the common mutations in HCC, appears to be associated with the expression of immune genes as well. These results reveal the difference of immune profiles between HCC subtypes and possible reasons and influence, which may also deepen our understanding of the carcinogenesis process.

关键词： Hepatocellular carcinoma Immune Tumor-infiltrating lymphocytes Metabolism CTNNB1 mutation

来源：评论

学校读者我要写书评

暂无评论

Prediction of oral food challenge outcomes via ensemble learning

引用

Informatics in Medicine Unlocked 2023年 36卷

作者： Zhang, Justin Lee, Deborah Jungles, Kylie Shaltis, Diane Najarian, Kayvan Ravikumar, Rajan Sanders, Georgiana Gryak, Jonathan Department of Electrical and Computer Engineering University of Michigan Ann Arbor MI United States Department of Internal Medicine University of Michigan Ann Arbor MI United States Department of Pediatrics University of Michigan Ann Arbor MI United States Department of Computational Medicine and Bioinformatics University of Michigan Ann Arbor MI United States Michigan Institute for Data Science University of Michigan Ann Arbor MI United States Department of Emergency Medicine University of Michigan Ann Arbor MI United States Department of Computer Science and Engineering University of Michigan Ann Arbor MI United States Max Harry Weil Institute for Critical Care Research and Innovation University of Michigan Ann Arbor MI United States Mary H. Weiser Food Allergy Center University of Michigan Ann Arbor MI United States Department of Computer Science Queens College City University of New York New York NY United States

Oral Food Challenges (OFCs) are essential to accurately diagnosing food allergy due to the limitations of existing clinical testing. However, some patients are hesitant to undergo OFCs, while those willing suffer from limited access to allergists in rural/community healthcare settings. Despite its success in predicting patient outcomes in other clinical settings, few applications of machine learning to food allergy have been developed. Thus, in this study, we seek to leverage machine learning methodologies for OFC outcome prediction. Retrospective data was gathered from 1,112 patients who collectively underwent a total of 1,284 OFCs, and consisted of clinical factors including serum-specific Immunoglobulin E (IgE), total IgE, skin prick tests (SPTs), comorbidities, sex, and age. Using these features, multiple machine learning models were constructed to predict OFC outcomes for three common allergens: peanut, egg, and milk. The best performing model for each allergen was an ensemble of random forest (egg) or Learning Using Concave and Convex Kernels (LUCCK) (peanut, milk) models, which achieved an Area under the Curve (AUC) of 0.91, 0.96, and 0.94, in predicting OFC outcomes for peanut, egg, and milk, respectively. Moreover, all such models had sensitivity and specificity values ≥89%. Model interpretation via SHapley Additive exPlanations (SHAP) indicates that specific IgE, along with wheal and flare values from SPTs, are highly predictive of OFC outcomes. The results of this analysis suggest that ensemble learning has the potential to predict OFC outcomes and reveal relevant clinical factors for further study. © 2022 The Author(s)

关键词： Ensemble learning Learning Using Concave and Convex Kernels Machine learning Oral food challenges SHAP

来源：评论

学校读者我要写书评

暂无评论

Genome assembly and transcriptome analysis provide insights into the antischistosome mechanism of Microtus fortis

引用

Journal of Genetics and Genomics 2020年第12期47卷 743-755页

作者： Hong Li Zhen Wang Shumei Chai Xiong Bai Guohui Ding Yuanyuan Li Junyi Li Qingyu Xiao Benpeng Miao Weili Lin Jie Feng Mingyue Huang Cheng Gao Bin Li Wei Hu Jiaojiao Lin Zhiqiang Fu Jianyun Xie Yixue Li Bio-Med Big Data Center CAS Key Laboratory of Computational BiologyShanghai Institute of Nutrition and HealthUniversity of Chinese Academy of SciencesChinese Academy of SciencesShanghai 200031China Shanghai Veterinary Research Institute Chinese Academy of Agricultural SciencesKey Laboratory of Animal ParasitologyMinistry of AgricultureShanghai 200241China Shanghai Laboratory Animal Research Center Shanghai 201203China Institute for Digital Health International Human Phenome Institutes(Shanghai)Shanghai 200433China Shanghai Center for Bioinformation Technology Shanghai Academy of Science and TechnologyShanghai 201203China School of Computer Science and Technology Harbin Institute of Technology(Shenzhen)ShenzhenGuangdong 518055China Shanghai Institute of Immunology Department of Immunology and MicrobiologyShanghai JiaoTong University School of MedicineShanghai 200025China State Key Laboratory of Genetic Engineering Ministry of Education Key Laboratory of Contemporary AnthropologyCollaborative Innovation Center for Genetics and DevelopmentSchool of Life SciencesFudan UniversityShanghai 200438China Hangzhou Institute for Advanced Study University of Chinese Academy of SciencesHangzhouZhejiang 330106China

Microtus fortis is the only mammalian host that exhibits intrinsic resistance against Schistosoma japonicum ***,the underlying molecular mechanisms of this resistance are not yet ***,we perform the first de novo genome assembly of ***,comprehensive gene annotation analysis,and evolution ***,we compare the recovery rate of schistosomes,pathological changes,and liver transcriptomes between *** and mice at different time points after *** observe that the time and type of immune response in *** are different from those in *** activates immune and inflammatory responses on the 10th day post infection,such as leukocyte extravasation,antibody activation,Fc-gamma receptor-mediated phagocytosis,and the interferon signaling cascade,which play important roles in preventing the development of *** contrast,an intense immune response occurrs in mice at the late stages of infection and could not eliminate *** mice suffer severe pathological injury and continuous decreases in cell cycle,lipid metabolism,and other *** findings offer new insights into the intrinsic resistance mechanism of *** against schistosome *** genome sequence also provides the basis for future studies of other important traits in ***.

关键词： Genome assembly Microtus fortis Schistosome Immune Transcriptome

来源：评论

学校读者我要写书评

暂无评论

A user authentication enabled piezoelectric force touch system for the internet of things

TechRxiv

引用

TechRxiv 2020年

作者： Huang, Anbiao Gao, Shuo Nathan, Arokia School of Instrumentation Science and Optoelectronic Engineering Beihang University Beijing100191 China Beijing Advanced Innovation Center for Big Data-Based Precision Medicine Interdisciplinary Innovation Institute of Medicine and Engineering Beihang University Beijing100191 China Cambridge Touch Technologies Inc. 154 Cambridge Science Park Milton Rd Milton CambridgeCB4 0GN United Kingdom

In Internet of Things (IoT) applications, secure access to smart systems, e.g., smartphones, is important for protecting private information. Among various authentication techniques, keystroke authentication methods based on touch behavior of the user have received increasing attention. This is due to the unique benefits, such as no additional hardware component and the ease of use in most smart systems. In this paper, we present a technique for obtaining high user authentication accuracy by utilizing a user’s touch time and force information, which are obtained from a piezoelectric touch panel. After combining artificial neural networks with the user’s touch features, an equal error rate (EER) of 1.09% is achieved, validating the feasibility of the proposed technique for achieving highly secure user authentication, hence advancing the development of security techniques potentially deployable in the field of IoT. © 2020, CC BY.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

TASL-Net: Tri-attention selective learning network for intelligent diagnosis of bimodal ultrasound video

引用

Expert Systems with Applications 2025年 290卷

作者： Zhao, Chengqian Yao, Zhao Hu, Zhaoyu Xie, Yuanxin Zhang, Yafang Wang, Yuanyuan Li, Shuo Zhou, Jianhua Zhou, Jianqiao Wang, Yin Yu, Jinhua College of Biomedical Engineering Fudan University Shanghai200438 China Department of Ultrasound Sun Yat-Sen University Cancer Center State Key Laboratory of Oncology in South China Collaborative Innovation Center for Cancer Medicine Guangzhou510060 China Department of Biomedical Engineering and the Department of Computer and Data Science Case Western Reserve University ClevelandOH44106 United States Department of Ultrasound Ruijin Hospital Shanghai Jiaotong University School of Medicine Shanghai200025 China Department of Ultrasound Shanghai Pulmonary Hospital Tongji University School of Medicine Shanghai200433 China

In the intelligent diagnosis of bimodal (grayscale and contrast-enhanced) ultrasound videos, embedding medical domain knowledge, such as temporal, spatial, and bimodal diagnostic attention of sonographers, enhances both performance and clinical reliability of the deep learning networks. However, automatically focusing on patient-specific features in videos, while efficiently encoding bimodal information, presents intractable challenges. This paper introduces the tri-attention selective learning network (TASL-Net) to address these challenges. TASL-Net incorporates the tri-attention mechanism into a mutual transformer framework for the intelligent diagnosis of bimodal ultrasound videos. It uses a time-intensity-curve based video selector to embed temporal attention and remove redundant information, an earliest-enhanced position detector to apply spatial attention by capturing perfusion differences inside and outside tumors, and a mutual encoding strategy combining convolution and transformer to integrate structural features from grayscale videos with perfusion variations from contrast-enhanced videos. Comprehensive ablation and comparison experiments on three datasets (lung, breast, and liver) comprising 791 cases demonstrate that TASL-Net achieves superior performance, with AUCs of 0.86, 0.86, and 0.97, and an average diagnostic accuracy improvement of 6.43 % over state-of-the-art methods. These results confirm the effectiveness of combining tri-attention mechanism with mutual spatiotemporal feature encoding in improving diagnostic performance of bimodal ultrasound videos. © 2025 Elsevier Ltd

关键词： Network coding

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：