检索结果-内蒙古大学图书馆

An Explainable Biomedical Foundation Model via Large-Scale Concept-Enhanced Vision-Language Pre-training

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Nie, Yuxiang He, Sunan Bie, Yequan Wang, Yihui Chen, Zhixuan Yang, Shu Cai, Zhiyuan Wang, Hongmei Wang, Xi Luo, Luyang Wu, Mingxiang Wu, Xian Kin Chan, Ronald Cheong Lau, Yuk Ming Zheng, Yefeng Rajpurkar, Pranav Chen, Hao Department of Computer Science and Engineering The Hong Kong University of Science and Technology Hong Kong Department of Biomedical Informatics Harvard University Boston United States Department of Radiology Shenzhen People’s Hospital Shenzhen China Jarvis Research Center Tencent YouTu Lab Shenzhen China Department of Anatomical and Cellular Pathology The Chinese University of Hong Kong Hong Kong State Key Laboratory of Translational Oncology The Chinese University of Hong Kong Hong Kong Division of Dermatology Department of Medicine Queen Mary Hospital Hong Kong Medical Artificial Intelligence Laboratory Westlake University Hangzhou China Department of Chemical and Biological Engineering The Hong Kong University of Science and Technology Hong Kong Division of Life Science The Hong Kong University of Science and Technology Hong Kong State Key Laboratory of Molecular Neuroscience The Hong Kong University of Science and Technology Hong Kong Shenzhen-Hong Kong Collaborative Innovation Research Institute The Hong Kong University of Science and Technology Shenzhen China

The clinical adoption of artificial intelligence (AI) in medical imaging requires models that are both diagnostically accurate and interpretable to clinicians. While current multimodal biomedical foundation models prioritize performance, their black-box nature hinders explaining the decision-making process in clinically meaningful concepts. Here, we present ConceptCLIP, the first explainable biomedical foundation model that achieves state-of-the-art diagnostic accuracy while delivering human-interpretable explanations across diverse imaging modalities. We curate MedConcept-23M, the largest pre-training dataset comprising 23 million image-text-concept triplets across diverse medical modalities, where clinical concepts are derived from the Unified Medical Language System. Leveraging this dataset, we develop ConceptCLIP through a novel dual-alignment approach that simultaneously learns global image-text representations and fine-grained region-concept associations for precise and interpretable medical image analysis. We curate the most extensive evaluation benchmark for multimodal biomedical foundation models, covering 52 clinical tasks spanning 10 imaging modalities. Extensive experiments demonstrate that ConceptCLIP outperforms existing state-of-the-art multimodal biomedical foundation models. Importantly, ConceptCLIP demonstrates superior diagnostic performance while providing human-understandable explanations validated by clinical experts. As the first precise and interpretable biomedical foundation model, ConceptCLIP represents a critical milestone toward the widespread clinical adoption of AI, thereby advancing trustworthy AI in medicine. Copyright © 2025, The Authors. All rights reserved.

关键词： Visual languages

Performance Analysis of InSb/InGaAs TFETs for Improved Electrical Characteristics with Black Phosphorus Channel for Analog/RF Applications

学校读者我要写书评

暂无评论

ECS Journal of Solid State science and Technology 2025年第6期14卷 063004-063004页

作者： N. M. Mary Sindhuja M. Navaneetha Velammal Hashim Elshafie Azath Mubarakali Department of ECE Kamaraj College of Engineering and Technology S.P.G Chidambara Nadar C Nagammal Campus Virudhunagar Tamil Nadu—625701 India Department of CSE Saveetha Engineering College Saveetha Nagar Sriperumbadur Chennai Tamil Nadu—602105 India Department of computer Engineering College of Computer Science King Khalid University Abha - 61421 Saudi Arabia Department of Informatics and Computer Systems College of Computer Science King Khalid University Abha - 61421 Saudi Arabia

This work investigates the impact of dielectric halo doping on the radio-frequency (RF) and analog characteristics of InSb/InGaAs tunnel field-effect transistors (TFETs) with a black phosphorus (BP) channel. Source-side dielectric halo doping significantly reduces ambipolar conduction, especially when the negative gate bias is strong. The study evaluates how TiO₂-high-k and SiC-low-k dielectric halo materials affect important analog and radio frequency parameters, including gate-to-source capacitance, transconductance, output resistance, gain-bandwidth product, and cut-off frequency, using comprehensive TCAD 3D simulations with Silvaco. The results demonstrate that while high-k halo doping effectively lessens ambipolar effects, low-k materials enhance analog and radio frequency performance. Furthermore, a comparison of transient behaviour shows that the BP-channel InSb/InGaAs TFET performs better than conventional TFET designs, making it a viable option for high-frequency and energy-efficient circuit applications.

关键词： ambipolar suppression Analog/RF performance black phosphorus channel dielectric halo doping InSb/InGaAs TFETs

Vision Transformer for Contrastive Clustering

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Ling, Hua-Bao Zhu, Bowen Huang, Dong Chen, Ding-Hua Wang, Chang-Dong Lai, Jian-Huang The College of Mathematics and Informatics South China Agricultural University Guangzhou China The School of Computer Science and Engineering Sun Yat-sen University Guangzhou China Guangdong Key Laboratory of Information Security Technology Guangzhou China Key Laboratory of Machine Intelligence and Advanced Computing Ministry of Education China

Vision Transformer (ViT) has shown its advantages over the convolutional neural network (CNN) with its ability to capture global long-range dependencies for visual representation learning. Besides ViT, contrastive learning is another popular research topic recently. While previous contrastive learning works are mostly based on CNNs, some recent studies have attempted to combine ViT and contrastive learning for enhanced self-supervised learning. Despite the considerable progress, these combinations of ViT and contrastive learning mostly focus on the instance-level contrastiveness, which often overlook the global contrastiveness and also lack the ability to directly learn the clustering result (e.g., for images). In view of this, this paper presents a novel deep clustering approach termed Vision Transformer for Contrastive Clustering (VTCC), which for the first time, to our knowledge, unifies the Transformer and the contrastive learning for the image clustering task. Specifically, with two random augmentations performed on each image, we utilize a ViT encoder with two weight-sharing views as the backbone. To remedy the potential instability of the ViT, we incorporate a convolutional stem to split each augmented sample into a sequence of patches, which uses multiple stacked small convolutions instead of a big convolution in the patch projection layer. By learning the feature representations for the sequences of patches via the backbone, an instance projector and a cluster projector are further utilized to perform the instance-level contrastive learning and the global clustering structure learning, respectively. Experiments on eight image datasets demonstrate the stability (during the training-from-scratch) and the superiority (in clustering performance) of our VTCC approach over the state-of-the-art. © 2022, CC BY.

关键词： Convolution

Ambiguity learning and data correlation in multi-cross modal cyber-physical systems for detecting fake information

学校读者我要写书评

暂无评论

Cyber-Physical Systems 2025年

作者： Byeon, Haewon AlMahadin, Ghayth Bsoul, Qusay Quraishi, Aadam Soni, Mukesh Khan, Shahi Raza Shabaz, Mohammad Department of AI and Software Inje University Gimhae Korea Republic of Data Science Department College of Information Technology Mutah University Mutah Jordan Department of Cybersecurity College of Computer Sciences and Informatics Amman Arab University Amman Jordan M.D. Research Intervention Treatment Institute HoustonTX United States Dr. D. Y. Patil School of Science & Technology Dr. D. Y. Patil Vidyapeeth Pune Pune India MJMC Centre for Media and Mass Communication Jamia Hamdard Delhi India Model Institute of Engineering and Technology Jammu India

This paper proposes an improved cyber-physical false information detection model based on cross-modal ambiguous learning. This work introduced the IC2 LFD model with an emphasis on ‘text-image’ ambiguity, focusing on the main goal of intelligent multi-cross modal data fake information identification. The experiment revealed that the significance of text and images differs in intelligent multi-cross modal data fake information detection. The effectiveness of the model is verified on the twitter dataset, showing a 6% accuracy improvement compared to the baseline model and a 1.6% performance improvement over detection methods without dynamic weight allocation. © 2025 Informa UK Limited, trading as Taylor & Francis Group.

关键词： Adversarial machine learning

Sustainable Wireless Networks via Reconfigurable Intelligent Surfaces (RISs): Overview of the ETSI ISG RIS

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Liu, Ruiqi Zheng, Shuang Wu, Qingqing Jiang, Yifan Zhang, Nan Liu, Yuanwei Di Renzo, Marco Alexandropoulos, George C. The State Key Laboratory of Mobile Network and Mobile Multimedia Technology Shenzhen China Wireless and Computing Research Institute ZTE Corporation Shenzhen China The Department of Electronic Engineering Shanghai Jiao Tong University Shanghai200240 China The School of Electronic Engineering and Computer Science Queen Mary University of London United Kingdom Université Paris-Saclay CNRS CentraleSupélec Laboratoire des Signaux et Systèmes 3 Rue Joliot-Curie Gif-sur-Yvette91192 France The Department of Informatics and Telecommunications National and Kapodistrian University of Athens Panepistimiopolis Ilissia Athens15784 Greece

Reconfigurable Intelligent Surfaces (RISs) are a novel form of ultra-low power devices that are capable to increase the communication data rates as well as the cell coverage in a cost- and energy-efficient way. This is attributed to their programmable operation that enables them to dynamically manipulate the wireless propagation environment, a feature that has lately inspired numerous research investigations and applications. To pave the way to the formal standardization of RISs, the European Telecommunications Standards Institute (ETSI) launched the Industry Specification Group (ISG) on the RIS technology in September 2021. This article provides a comprehensive overview of the status of the work conducted by the ETSI ISG RIS, covering typical deployment scenarios of reconfigurable metasurfaces, use cases and operating applications, requirements, emerging hardware architectures and operating modes, as well as the latest insights regarding future directions of RISs and the resulting smart wireless environments. © 2024, CC BY-NC-ND.

关键词： Standardization

Predicting cognitive scores with graph neural networks through sample selection learning

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Hanik, Martin Demirtaş, Mehmet Arif Gharsallaoui, Mohammed Amine Rekik, Islem Zuse Institute Berlin Berlin Germany BASIRA Lab Faculty of Computer and Informatics Istanbul Technical University Istanbul Turkey Tunis Tunisia BASIRA Lab Faculty of Computer and Informatics Istanbul Technical University Istanbul Turkey Computing School of Science and Engineering University of Dundee Dundee United Kingdom

Analyzing the relation between intelligence and neural activity is of the utmost importance in understanding the working principles of the human brain in health and disease. In existing literature, functional brain connectomes have been used successfully to predict cognitive measures such as intelligence quotient (IQ) scores in both healthy and disordered cohorts using machine learning models. However, existing methods resort to flattening the brain connectome (i.e., graph) through vectorization which overlooks its topological properties. To address this limitation and inspired from the emerging graph neural networks (GNNs), we design a novel regression GNN model (namely RegGNN) for predicting IQ scores from brain connectivity. On top of that, we introduce a novel, fully modular sample selection method to select the best samples to learn from for our target prediction task. However, since such deep learning architectures are computationally expensive to train, we further propose a learning-based sample selection method that learns how to choose the training samples with the highest expected predictive power on unseen samples. For this, we capitalize on the fact that connectomes (i.e., their adjacency matrices) lie in the symmetric positive definite (SPD) matrix cone. Our results on full-scale and verbal IQ prediction outperforms comparison methods in autism spectrum disorder cohorts and achieves a competitive performance for neurotypical subjects using 3-fold cross-validation. Furthermore, we show that our sample selection approach generalizes to other learning-based methods, which shows its usefulness beyond our GNN architecture. © 2021, CC BY.

关键词： Forecasting

Enabling cross-sectorial, circular economy transition in SME via digital platform integrated operational services

学校读者我要写书评

暂无评论

Procedia Manufacturing 2021年 54卷 70-75页

作者： Gianfranco Pedone Richárd Beregi Krisztián Balázs Kis Marcello Colledani Centre of Excellence in Production Informatics and Control Institute for Computer Science and Control Kende u. 13-17 1111 Budapest Hungary Politecnico di Milano Department of Mechanical Engineering Via la Masa 1 20156 Milan Italy

Circular Economy (CE) is an emerging paradigm aiming at establishing a new sustainable development path by decoupling economic growth and resource consumption. The main limitation in current production models is that manufacturing and de-and remanufacturing operations are carried out independently and without sharing information and economic benefits. This can be overcome by rethinking the current European industrial system into new collaborative and sustainable value networks. New business cases can be generated also for individual Small and Medium Enterprises (SME) by taking advantage of the connection with the Industry 4.0 opportunities and by building on cross-sectorial markets and the potential of digital solutions. The operational services (DOS) presented in this paper are the key part of a novel digital platform called DigiPrime, whose aim is to provide SME with a federated service infrastructure that help overcome the current information asymmetry and explore the possibility to access and transit towards a CE cross-sectorial value-chain. The manuscript presents a selection of DOS considered fundamental for an SME in such a process, through a relevant exploitation use-case scenario. The combination of such DOS has been validated within the project for the case of an electric car repair-shop network and the preliminary outcome of their integration process is also outlined.

关键词： Circular Economy Industry 4.0 Service-oriented Platform Operational Services as a Service Small Mid-size Enterprise

Advanced Deep Learning Human Herpes Virus 6 (HHV-6) Molecular Detection in Understanding Human Infertility

学校读者我要写书评

暂无评论

Computational Intelligence and Neuroscience 2022年 2022卷 1422963页

作者： Alazzam, Malik Bader Al-Radaideh, Ahmad Tawfig Binsaif, Nasser Alghamdi, Ahmed S. Rahman, Md Adnan Faculty of Computer Science and Informatics Amman Arab University Amman Jordan Department of MIS College of Business Jadara University Irbid Jordan Department of E-Commerce College of Administrative and Financial Sciences Saudi Electronic University Riyadh Saudi Arabia Department of Computer Engineering Collage of Computers and Information Technology Taif University P.O. Box 11099 Taif21944 Saudi Arabia Dhaka Bangladesh

To see if HHV-6 may be a cause of infertility, researchers looked at 18 men and 10 women who had unexplained critical fertility and had at least one prior pregnancy. HHV-6 DNA was discovered in both infertile and fertile peripheral blood mononuclear cells (PBMC) (12 and 14%, respectively);endometrial epithelial cells from 4/10 (40%) infertile women were positive for HHV-6 DNA;this viral DNA was not found in the endometrium of fertile women. When endometrial epithelial cells were cultivated, they produced viral early and late proteins, suggesting the existence of an infectious virus. Endometrial HHV-6 infection creates an aberrant NK cell and cytokine profile, resulting in a uterine domain that is not favorable to conception, according to the findings. To corroborate the findings, studies of extra fertile and barren women should be done. Semen samples were taken from 18 guys who visited the Government General Hospital Guntur's infertility department because they were having reproductive issues with their partners. Herpes virus DNA has been discovered in the sperm of symptomatic fertile and infertile male patients on rare instances. Furthermore, researchers must investigate the role of viral diseases in male infertility. © 2022 Malik Bader Alazzam et al.

关键词： Viruses

Transformers in Medical Image Analysis: A Review

学校读者我要写书评

暂无评论

arXiv 2022年

作者： He, Kelei Gan, Chen Li, Zhuoyuan Rekik, Islem Yin, Zihao Ji, Wen Gao, Yang Wang, Qian Zhang, Junfeng Shen, Dinggang Medical School of Nanjing University Nanjing China National Institute of Healthcare Data Science Nanjing University Nanjing China State Key Laboratory for Novel Software Technology Nanjing University Nanjing China BASIRA Lab Faculty of Computer and Informatics Engineering Istanbul Technical University Istanbul Turkey School of Science and Engineering Computing University of Dundee United Kingdom School of Biomedical Engineering ShanghaiTech University Shanghai China Department of Research and Development Shanghai United Imaging Intelligence Co. Ltd. Shanghai China

Transformers have dominated the field of natural language processing, and recently impacted the computer vision area. In the field of medical image analysis, Transformers have also been successfully applied to full-stack clinical applications, including image synthesis/reconstruction, registration, segmentation, detection, and diagnosis. Our paper aims to promote awareness and application of Transformers in the field of medical image analysis. Specifically, we first overview the core concepts of the attention mechanism built into Transformers and other basic components. Second, we review various Transformer architectures tailored for medical image applications and discuss their limitations. Within this review, we investigate key challenges revolving around the use of Transformers in different learning paradigms, improving the model efficiency, and their coupling with other techniques. We hope this review can give a comprehensive picture of Transformers to the readers in the field of medical image analysis. © 2022, CC BY-SA.

关键词： Diagnosis