检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Chatterjee, Soumick Mattern, Hendrik Dörner, Marc Sciarra, Alessandro Dubost, Florian Schnurre, Hannes Khatun, Rupali Yu, Chun-Chih Hsieh, Tsung-Lin Tsai, Yi-Shan Fang, Yi-Zeng Yang, Yung-Ching Huang, Juinn-Dar Xu, Marshall Liu, Siyu Ribeiro, Fernanda L. Bollmann, Saskia Chintalapati, Karthikesh Varma Radhakrishna, Chethan Mysuru Ram Kumar, Sri Chandana Hudukula Sutrave, Raviteja Qayyum, Abdul Mazher, Moona Razzak, Imran Rodero, Cristobal Niederer, Steven Lin, Fengming Xia, Yan Wang, Jiacheng Qiu, Riyu Wang, Liansheng Panah, Arya Yazdan Jurdi, Rosana El Fu, Guanghui Arslan, Janan Vaillant, Ghislain Valabregue, Romain Dormont, Didier Stankoff, Bruno Colliot, Olivier Vargas, Luisa Chacón, Isai Daniel Pitsiorlas, Ioannis Arbeláez, Pablo Zuluaga, Maria A. Schreiber, Stefanie Speck, Oliver Nürnberger, Andreas Faculty of Computer Science Otto von Guericke University Magdeburg Magdeburg Germany Data and Knowledge Engineering Group Otto von Guericke University Magdeburg Magdeburg Germany Genomics Research Centre Human Technopole Milan Italy Biomedical Magnetic Resonance Otto von Guericke University Magdeburg Magdeburg Germany Department of Neurology Medical Faculty University Hospital of Magdeburg Magdeburg Germany German Centre for Neurodegenerative Diseases Magdeburg Germany Centre for Behavioural Brain Sciences Magdeburg Germany University Hospital Zurich University of Zurich Zurich Switzerland Google Inc. United States Translational Radiobiology Department of Radiation Oncology Universitätsklinikum Erlangen Friedrich-Alexnder-Universität Erlangen-Nürnberg Erlangen Germany National Yang Ming Chiao Tung University Hsinchu Taiwan School of Electrical Engineering and Computer Science University of Queensland Brisbane Australia Australian eHealth Research Centre Chile National Heart and Lung Institute Faculty of Medicine Imperial College London London United Kingdom Centre for Medical Image Computing Department of Computer Science University College London London United Kingdom School of Computer Science and Engineering University of New South Wales Sydney Australia Mohamed bin Zayed University of Artificial Intelligence Abu Dhabi United Arab Emirates The Alan Turing Institute London United Kingdom School of Computing University of Leeds Leeds United Kingdom Department of Computer Science School of Informatics Xiamen University Xiamen China Manteia Technologies Co. Ltd Xiamen China Leicester International Institute Dalian University of Technology Dalian China Sorbonne Université Institut du Cerveau - Paris Brain Institute ICM CNRS Inria Inserm AP-HP Hôpital de la Pitié Salpêtrière ParisF-75013 France Centre of Formation and Research in Artificial Intelligence Universidad de Los Andes Colombia Data Science Department EURECOM Sophia Antipolis Franc

The human brain receives nutrients and oxygen through an intricate network of blood vessels. Pathology affecting small vessels, at the mesoscopic scale, represents a critical vulnerability within the cerebral blood supply and can lead to severe conditions, such as Cerebral Small Vessel Diseases. The advent of 7 Tesla MRI systems has enabled the acquisition of higher spatial resolution images, making it possible to visualise such vessels in the brain. However, the lack of publicly available annotated datasets has impeded the development of robust, machine learning-driven segmentation algorithms. To address the complexities of mesoscopic vessel segmentation and to highlight the need for advanced techniques to manage the high noise levels and poor vessel-to-background contrast inherent in"ultra-high-resolution" data, the SMILE-UHURA challenge was organised. This challenge, held in conjunction with the ISBI 2023, in Cartagena de Indias, Colombia, aimed to provide a platform for researchers working on related topics. The SMILE-UHURA challenge addresses the gap in publicly available annotated datasets by providing an annotated dataset of Time-of-Flight angiography acquired with 7T MRI. This dataset was created through a combination of automated pre-segmentation and extensive manual refinement. In this manuscript, sixteen submitted methods and two baseline methods are compared both quantitatively and qualitatively on two different datasets: held-out test MRAs from the same dataset as the training data (with labels kept secret) and a separate 7T ToF MRA dataset where both input volumes and labels are kept secret. The results demonstrate that most of the submitted deep learning methods, trained on the provided training dataset, achieved reliable segmentation performance. Dice scores reached up to 0.838 ± 0.066 and 0.716 ± 0.125 on the respective datasets, with an average performance of up to 0.804 ± 0.15. Copyright © 2024, The Authors. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Extracting possibly representative COVID-19 Biomarkers from X-Ray images with Deep Learning approach and image data related to Pulmonary Diseases

arXiv

引用

arXiv 2020年

作者： Apostolopoulos, Ioannis D. Sokratis, Aznaouridis I. Mpesiana, Tzani A. Department of Medical Physics School of Medicine University of Patras Patras26504 Greece Department of Computer Engineering and Informatics University of Patras Patras26504 Greece Computer Technology Institute and Press "Diophantus" Patras26504 Greece

In this study, the problem of automatically classifying pulmonary diseases, including the recently emerged COVID-19, from X-Ray images, is considered. While the spread of COVID-19 is increased, new, automatic, and reliable methods for accurate detection are essential to reduce the exposure of the medical experts to the outbreak. X-Ray imaging, although limited to specific visualizations, may be helpful for the diagnosis. Deep Learning has proven to be a remarkable method to extract massive high-dimensional features from medical images. Specifically, in this paper, the state-of-the-art Convolutional Neural Network called Mobile Net is employed and trained from scratch to investigate the importance of the extracted features for the classification task. A large-scale dataset of 3905 X-Ray images, corresponding to 6 diseases is utilized for training MobileNet v2, which has been proven to achieve remarkable results in related tasks. The results suggest that training CNNs from scratch may reveal vital biomarkers related but not limited to the COVID-19 disease, while an overall classification accuracy of the seven classes reaches 87.66%. Besides, this method achieves 99.18% accuracy, 97.36% Sensitivity, and 99.42% Specificity in the detection of COVID-19. Copyright © 2020, The Authors. All rights reserved.

关键词： Biomarkers

来源：评论

学校读者我要写书评

暂无评论

Uncovering distinct public transport user profiles and the factors influencing the users’ intentions

引用

Journal of Urban Mobility 2025年 7卷

作者： Kriswardhana, Willy Ismael, Karzan Duleba, Szabolcs Esztergár-Kiss, Domokos Department of Transport Technology and Economics Faculty of Transportation Engineering and Vehicle Engineering Budapest University of Technology and Economics Műegyetem rkp. 3. Budapest H-1111 Hungary Department of Civil Engineering Faculty of Engineering Universitas Jember Jember 68121 Indonesia Department of City Planning Engineering Technical College of Engineering Sulaimani Polytechnic University Sulaimani 46001 Iraq Institute of Mathematics and Informatics University of Nyíregyháza Nyíregyháza Hungary HUN-REN Institute For Computer Science And Control (SZTAKI) Hungary

Given the escalating environmental challenges, it is clear that the approach to the usage of public transport (PT) must be reconsidered. Loyal PT users help attract other travelers through positive recommendations, this drives interest in exploring the relationships between service quality, satisfaction, and loyalty. Although these relationships in the PT context have been studied, there have been limited efforts to explore the variability of their effects. Therefore, this research explores how service quality, satisfaction, involvement, and prior knowledge influence the behavioral intentions to use PT across various traveler groups. The hypotheses are tested by using multi-group analysis in the structural equation modeling framework, and three user profiles are revealed by applying the latent class cluster analysis. The results indicate that involvement strongly influences the behavioral intentions, while satisfaction impacts the intentions solely when involvement acts as a moderator. Class-specific results indicate that younger females living outside the city center put greater emphasis on service quality improvements for their satisfaction and involvement in PT services compared to males and older individuals residing in the city center. Low-income students living in the city center are more likely to use PT if they feel confident how to use the service. The results are relevant for PT operators and related stakeholders in maintaining travelers’ loyalty. © 2025 The Author(s)

关键词： Heterogeneity Intention Involvement Public transport Satisfaction Service quality

来源：评论

学校读者我要写书评

暂无评论

Potential to identify neutrino mass ordering with reactor antineutrinos at JUNO

引用

Chinese Physics C 2025年第3期49卷 37-63页

作者： Angel Abusleme Thomas Adam Shakeel Ahmad Rizwan Ahmed Sebastiano Aiello Muhammad Akram Abid Aleem Fengpeng An Qi An Giuseppe Andronico Nikolay Anfimov Vito Antonelli Tatiana Antoshkina Burin Asavapibhop João Pedro Athayde Marcondes de André Didier Auguste Weidong Bai Nikita Balashov Wander Baldini Andrea Barresi Davide Basilico Eric Baussan Marco Bellato Marco Beretta Antonio Bergnoli Daniel Bick Lukas Bieger Svetlana Biktemerova Thilo Birkenfeld Iwan Morton-Blake David Blum Simon Blyth Anastasia Bolshakova Mathieu Bongrand Clément Bordereau Dominique Breton Augusto Brigatti Riccardo Brugnera Riccardo Bruno Antonio Budano Jose Busto Anatael Cabrera Barbara Caccianiga Hao Cai Xiao Cai Yanke Cai Zhiyan Cai Stéphane Callier Antonio Cammi Agustin Campeny Chuanya Cao Guofu Cao Jun Cao Rossella Caruso Cédric Cerna Vanessa Cerrone Chi Chan Jinfan Chang Yun Chang Auttakit Chatrabhuti Chao Chen Guoming Chen Pingping Chen Shaomin Chen Yixue Chen Yu Chen Zhangming Chen Zhiyuan Chen Zikang Chen Jie Cheng Yaping Cheng Yu Chin Cheng Alexander Chepurnov Alexey Chetverikov Davide Chiesa Pietro Chimenti Yen-Ting Chin Ziliang Chu Artem Chukanov Gérard Claverie Catia Clementi Barbara Clerbaux Marta Colomer Molla Selma Conforti DiLorenzo Alberto Coppi Daniele Corti Simon Csakli Flavio Dal Corso Olivia Dalager Jaydeep Datta Christophe De La Taille Zhi Deng Ziyan Deng Xiaoyu Ding Xuefeng Ding Yayun Ding Bayu Dirgantara Carsten Dittrich Sergey Dmitrievsky Tadeas Dohnal Dmitry Dolzhikov Georgy Donchenko Jianmeng Dong Evgeny Doroshkevich Wei Dou Marcos Dracos Frédéric Druillole Ran Du Shuxian Du Katherine Dugas Stefano Dusini Hongyue Duyang Jessica Eck Timo Enqvist Andrea Fabbri Ulrike Fahrendholz Lei Fan Jian Fang Wenxing Fang Marco Fargetta Dmitry Fedoseev Zhengyong Fei Li-Cheng Feng Qichun Feng Federico Ferraro Amélie Fournier Haonan Gan Feng Gao Alberto Garfagnini Arsenii Gavrikov Marco Giammarchi Nunzio Giudice Maxim Gonchar Guanghua Gong Hui Gong Yuri Gornushkin Alexandre Göttela Marco Grassi Maxim Gromov Vasily Gromov Minghao Gu X Yerevan Physics Institute YerevanArmenia Université Libre de Bruxelles BrusselsBelgium Universidade Estadual de Londrina LondrinaBrazil Pontificia Universidade Catolica do Rio de Janeiro Rio de JaneiroBrazil Millennium Institute for SubAtomic Physics at the High-energy Frontier(SAPHIR) ANIDChile Pontificia Universidad Católica de Chile SantiagoChile Beijing Institute of Spacecraft Environment Engineering BeijingChina Beijing Normal University BeijingChina China Institute of Atomic Energy BeijingChina Institute of High Energy Physics BeijingChina North China Electric Power University BeijingChina School of Physics Peking UniversityBeijingChina Tsinghua University BeijingChina University of Chinese Academy of Sciences BeijingChina Jilin University ChangchunChina College of Electronic Science and Engineering National University of Defense TechnologyChangshaChina Chongqing University ChongqingChina Dongguan University of Technology DongguanChina Jinan University GuangzhouChina Sun Yat-Sen University GuangzhouChina Harbin Institute of Technology HarbinChina University of Science and Technology of China HefeiChina The Radiochemistry and Nuclear Chemistry Group in University of South China HengyangChina Wuyi University JiangmenChina Shandong University JinanChinaand Key Laboratory of Particle Physics and Particle Irradiation of Ministry of EducationShandong UniversityQingdaoChina Nanjing University NanjingChina Guangxi University NanningChina East China University of Science and Technology ShanghaiChina School of Physics and Astronomy Shanghai Jiao Tong UniversityShanghaiChina Tsung-Dao Lee Institute Shanghai Jiao Tong UniversityShanghaiChina Institute of Hydrogeology and Environmental Geology Chinese Academy of Geological SciencesShijiazhuangChina Nankai University TianjinChina Wuhan University WuhanChina Xi’an Jiaotong University Xi’anChina Xiamen University XiamenChina School of Physics and Microelectronics Zhengzhou UniversityZhengzhouChina Institute of Physics National

The Jiangmen Underground Neutrino Observatory(JUNO)is a multi-purpose neutrino experiment under construction in South *** paper presents an updated estimate of JUNO’s sensitivity to neutrino mass ordering using the reactor antineutrinos emitted from eight nuclear reactor cores in the Taishan and Yangjiang nuclear power *** measurement is planned by studying the fine interference pattern caused by quasi-vacuum oscillations in the oscillated antineutrino spectrum at a baseline of 52.5 km and is completely independent of the CP violating phase and neutrino mixing angleθ_(23).The sensitivity is obtained through a joint analysis of JUNO and Taishan Antineutrino Observatory(TAO)detectors utilizing the best available knowledge to date about the location and overburden of the JUNO experimental site,local and global nuclear reactors,JUNO and TAO detector responses,expected event rates and spectra of signals and backgrounds,and systematic uncertainties of analysis *** find that a 3σmedian sensitivity to reject the wrong mass ordering hypothesis can be reached with an exposure of about 6.5 years×26.6 GW thermal power.

关键词： JUNO neutrino neutrino mass ordering

来源：评论

学校读者我要写书评

暂无评论

SmartCHANGE: AI-based long-term health risk evaluation for driving behaviour change strategies in children and youth

SmartCHANGE: AI-based long-term health risk evaluation for d...

引用

International Conference on Applied Mathematics & computer Science (ICAMCS)

作者： Nina Reščič Janna Alberts Teatske M. Altenburg Mai J. M. Chinapaw Antonio De Nigro Dario Fenoglio Martin Gjoreski Anton Gradišek Gregor Jurak Athanasios Kiourtis Dimosthenis Kyriazis Marc Langheinrich Elena Mancuso Argyro Mavrogiorgou Mykola Pechenizkiy Roberto Pratola José Ribeiro Maroje Sorić Fawad Taj Tuija H. Tammelin Martijn Vastenburg Anna Vilanova Tanja G. M. Vrijkotte Mitja Luštrek Department of Intelligent Systems Jožef Stefan Institute Ljubljana Slovenia Research & Development ConnectedCare Nijmegen The Netherlands Department of Public and Occupational Health Amsterdam UMC Amsterdam Public Health Research Institute Vrije Universiteit Amsterdam Amsterdam The Netherlands Engineering Ingegneria Informatica S.p.A. Rome Italy Faculty of Informatics Università Della Svizzera Italiana Lugano Switzerland Faculty of Sports University of Ljubljana Ljubljana Slovenia Department of Digital Systems University of Piraeus Piraeus Greece Department of Mathematics and Computer Science Eindhoven University of Technology Eindhoven Netherlands Faculty of Sport University of Porto Porto Portugal Likes School of Health and Social Studies Jamk University of Applied Sciences Jyväskylä Finland

The SmartCHANGE project, a Horizon-Europe Research & Innovation project that has been ongoing since May 2023 and is scheduled to conclude in April 2027, aims to develop AI-driven decision-support tools to identify and reduce the long-term risk of non-communicable chronic diseases (NCDs) in children and youth. NCDs have become a significant public health burden in developed countries, with common risk factors including obesity, low physical fitness, and unhealthy lifestyle habits. Childhood and adolescence are critical periods for establishing healthy habits, and early intervention can prevent or delay the onset of NCDs later in life. However, current tools for identifying high-risk individuals mainly focus on adults, leading to missed opportunities for early detection. The SmartCHANGE project aims to address this gap by creating trustworthy AI tools that accurately assess risk factors in children and youth and promote optimised strategies for risk reduction. The paper outlines the plan for the project from the project’s proposed methodologies, including dataset utilisation, data cleaning, risk prediction models (basic and complex), and the implementation of federated learning. Additionally, it discusses user experience aspects such as participatory design, suggested applications, and the evaluation of project outcomes.

关键词：

来源：评论

学校读者我要写书评

暂无评论

The status-based optimization: Algorithm and comprehensive performance analysis

引用

Neurocomputing 2025年 647卷

作者： Jian Wang Yi Chen Chenglang Lu Ali Asghar Heidari Zongda Wu Huiling Chen Key Laboratory of Intelligent Informatics for Safety & Emergency of Zhejiang Province Wenzhou University Wenzhou 325035 China College of modern information technology Zhejiang Institute of Mechanical and Electrical Engineering Hangzhou Zhejiang 310051 China School of Surveying and Geospatial Engineering College of Engineering University of Tehran Tehran Iran Department of Computer Science and Engineering Shaoxing University Shaoxing 312000 China

In recent years, real-world optimization problems have grown increasingly complex, challenging the effectiveness of traditional deterministic approaches. This paper introduces the Status-based Optimization (SBO), an efficient algorithm inspired by human desire for status advancement. By emulating how individuals seek proximity to and learn from or acquire resources from high-status figures, SBO translates these social patterns into a robust computational method for challenging optimization tasks. Comprehensive comparative analyses against various meta-heuristic algorithms, such as those based on human behaviors, classical methods, and state-of-the-art techniques, validate its strength across problem dimensions of 10, 30, 50, and 100 as defined by the IEEE CEC 2017 test suite. Moreover, SBO excels in feature selection by identifying fewer but more relevant features than 8 established binary optimizers across 9 high-dimensional datasets. The algorithm's adaptability is further demonstrated in a multi-threshold image segmentation application at 3 threshold levels. SBO was benchmarked against 7 advanced meta-heuristic methods to segment 9 breast cancer histology images, successfully identifying optimal threshold values to enhance image clarity and segmentation accuracy. These promising results underscore the potential of SBO for future research and practical applications in a wide array of optimization scenarios. The SBO source code is available at https://***/*** for interested researchers.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Medical imaging with deep learning for COVID-19 diagnosis: A comprehensive review

arXiv

引用

arXiv 2021年

作者： Bharati, Subrato Podder, Prajoy Mondal, M. Rubaiyat Hossain Prasath, V.B. Surya Institute of Information and Communication Technology Bangladesh University of Engineering and Technology Dhaka1205 Bangladesh Division of Biomedical Informatics Cincinnati Children’s Hospital Medical Center Cincinnati45229 United States Department of Pediatrics University of Cincinnati College of Medicine CincinnatiOH45257 United States Department of Biomedical Informatics College of Medicine University of Cincinnati CincinnatiOH45267 United States Department of Electrical Engineering and Computer Science University of Cincinnati OH45221 United States

The outbreak of novel coronavirus disease (COVID-19) has claimed millions of lives and has affected all aspects of human life. This paper focuses on the application of deep learning (DL) models to medical imaging and drug discovery for managing COVID-19 disease. In this article, we detail various medical imaging-based studies such as X-rays and computed tomography (CT) images along with DL methods for classifying COVID-19 affected versus pneumonia. The applications of DL techniques to medical images are further described in terms of image localization, segmentation, registration, and classification leading to COVID-19 detection. The reviews of recent papers indicate that the highest classification accuracy of 99.80% is obtained when InstaCovNet-19 DL method is applied to an X-ray dataset of 361 COVID-19 patients, 362 pneumonia patients and 365 normal people. Furthermore, it can be seen that the best classification accuracy of 99.054% can be achieved when EDL_COVID DL method is applied to a CT image dataset of 7500 samples where COVID-19 patients, lung tumor patients and normal people are equal in number. Moreover, we illustrate the potential DL techniques in drug or vaccine discovery in combating the coronavirus. Finally, we address a number of problems, concerns and future research directions relevant to DL applications for COVID-19. © 2021, CC BY.

关键词： Coronavirus

来源：评论

学校读者我要写书评

暂无评论

Factors considered for the success of the Massive Open Online Course in the era of Smart Education: Systematic Literature Review 10

Factors considered for the success of the Massive Open Onlin...

引用

10th International Conference on ICT for Smart Society, ICISS 2019

作者： Suwita, Jaka Kosala, Raymond Ranti, Benny Supangkat, Suhono Harso Bina Nusantara University Department of Computer Science Jakarta11480 Indonesia Bina Nusantara University Computer Science Department Jakarta11480 Indonesia Bandung Institute of Technology Electrical Engineering and Informatics School Bandung Indonesia

ISBN: (纸本)9781728148809

The Internet of Things has a big influence on the appearance of new technologies in learning. Smart education is a learning concept that is developed by intelligent technology development. Therefore smart education must be able to provide an update of learning methods. Smart education must make it easier for users to interact and collaborate to make, organize, and share knowledge from anywhere and at any time accessible. In this paper, we conduct a Systematic Literature Review (SLR) research method to find out the features that exist in MOOCs in the era of intelligent education. This paper presents a comprehensive view of the development of MOOC features in the era of smart education in higher education. Our contribution is comprehensive information about MOOC features in the era of intelligent education in higher education, to improve the quality of learning, and possible use in practice. © 2019 IEEE.

关键词： Learning systems

来源：评论

学校读者我要写书评

暂无评论

Solving multi-armed bandit problems using a chaotic microresonator comb

arXiv

引用

arXiv 2023年

作者： Cuevas, Jonathan Iwami, Ryugo Uchida, Atsushi Minoshima, Kaoru Kuse, Naoya Graduate School of Sciences and Technology for Innovation Tokushima University 2-1 Minami-Josanjima Tokushima Tokushima770-8506 Japan Department of Information and Computer Sciences Saitama University 255 Shimo-okubo Sakura-ku Saitama Saitama City338-8570 Japan Graduate School of Informatics and Engineering The University of Electro-Communications 1-5-1 Chofugaoka Chofu Tokyo182-8585 Japan Institute of Post-LED Photonics Tokushima University 2-1 Minami-Josanjima Tokushima Tokushima770-8506 Japan

—The Multi-Armed Bandit (MAB) problem, foundational to reinforcement learning-based decision-making, addresses the challenge of maximizing rewards amidst multiple uncertain choices. While algorithmic solutions are effective, their computational efficiency diminishes with increasing problem complexity. Photonic accelerators, leveraging temporal and spatial-temporal chaos, have emerged as promising alternatives. However, despite these advancements, current approaches either compromise computation speed or amplify system complexity. In this paper, we introduce a chaotic microresonator frequency comb (chaos comb) to tackle the MAB problem, where each comb mode is assigned to a slot machine. Through a proof-of-concept experiment, we employ 44 comb modes to address an MAB with 44 slot machines, demonstrating performance competitive with both conventional software algorithms and other photonic methods. Further, the scalability of decision making is explored with up to 512 slot machines using experimentally obtained temporal chaos in different time slots. Power-law scalability is achieved with an exponent of 0.96, outperforming conventional software-based algorithms. Moreover, we find that a numerically calculated chaos comb accurately reproduces experimental results, paving the way for discussions on strategies to increase the number of slot machines. © 2023, CC BY.

关键词： Scalability

来源：评论

学校读者我要写书评

暂无评论

Two-Layer Retrieval-Augmented Generation Framework for Low-Resource Medical Question Answering Using Reddit Data: Proof-of-Concept Study

arXiv

引用

arXiv 2024年

作者： Das, Sudeshna Ge, Yao Guo, Yuting Rajwal, Swati Hairston, JaMor Powell, Jeanne Walker, Drew Peddireddy, Snigdha Lakamana, Sahithi Bozkurt, Selen Reyna, Matthew Sameni, Reza Xiao, Yunyu Kim, Sangmi Chandler, Rasheeta Hernandez, Natalie Mowery, Danielle Wightman, Rachel Love, Jennifer Spadaro, Anthony Perrone, Jeanmarie Sarker, Abeed Department of Biomedical Informatics School of Medicine Emory University AtlantaGA United States Department of Computer Science and Informatics Emory University AtlantaGA United States Department of Behavioral Social & Health Education Sciences Rollins School of Public Health Emory University AtlantaGA United States Department of Biomedical Engineering Georgia Institute of Technology Emory University AtlantaGA United States Department of Population Health Sciences Weill Cornell Medicine New YorkNY United States Nell Hodgson Woodruff School of Nursing Emory University AtlantaGA United States Center for Maternal Health Equity Morehouse School of Medicine AtlantaGA United States Department of Biostatistics Epidemiology and Informatics Perelman School of Medicine University of Pennsylvania PhiladelphiaPA United States Department of Emergency Medicine Warren Alpert Medical School of Brown University ProvidenceRI United States Department of Emergency Medicine Icahn School of Medicine at Mount Sinai New YorkNY United States Department of Emergency Medicine Rutgers New Jersey Medical School NewarkNJ United States Department of Emergency Medicine Perelman School of Medicine the University of Pennsylvania PhiladelphiaPA United States

Background: The increasing use of social media to share lived and living experiences of substance use presents a unique opportunity to obtain information on side effects, use patterns, and opinions on novel psychoactive substances. However, due to the large volume of data, obtaining useful insights through natural language processing technologies such as large language models is challenging. Objective: This paper aims to develop a retrieval-augmented generation (RAG) architecture for medical question answering pertaining to clinicians’ queries on emerging issues associated with health-related topics, using user-generated medical information on social media. Methods: We proposed a two-layer RAG framework for query-focused answer generation and evaluated a proof of concept for the framework in the context of query-focused summary generation from social media forums, focusing on emerging drug-related information. Our modular framework generates individual summaries followed by an aggregated summary to answer medical queries from large amounts of user-generated social media data in an efficient manner. We compared the performance of a quantized large language model (Nous-Hermes-2-7B-DPO), deployable in low-resource settings, with GPT-4. For this proof-of-concept study, we used user-generated data from Reddit to answer clinicians’ questions on the use of xylazine and ketamine. Results: Our framework achieves comparable median scores in terms of relevance, length, hallucination, coverage, and coherence when evaluated using GPT-4 and Nous-Hermes-2-7B-DPO, evaluated for 20 queries with 76 samples. There was no statistically significant difference between GPT-4 and Nous-Hermes-2-7B-DPO for coverage (Mann-Whitney U=733.0;n1=37;n2=39;P=.89 two-tailed), coherence (U=670.0;n1=37;n2=39;P=.49 two-tailed), relevance (U=662.0;n1=37;n2=39;P=.15 two-tailed), length (U=672.0;n1=37;n2=39;P=.55 two-tailed), and hallucination (U=859.0;n1=37;n2=39;P=.01 two-tailed). A statistically signifi

关键词： Natural language processing systems

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：