检索结果-内蒙古大学图书馆

Multidomain postoperative recovery trajectories after lumbar and thoracolumbar spine surgery

Spine Journal 2025年

作者： Yakdan, Salim Zhang, Jingwen Benedict, Braeden Xu, Ziqi Javeed, Saad Zhang, Justin K. Steel, Benjamin A. Gupta, Vivek P. Botterbush, Kathleen Piccirillo, Jay F. Rodebaugh, Thomas L. Goodin, Burel R. Buchowski, Jacob M Neuman, Brian Hafez, Daniel Kelly, Michael Ray, Wilson Z. Lu, Chenyang Frumkin, Madelyn Greenberg, Jacob K. Department of Neurological Surgery Washington University School of Medicine St. Louis MO United States Department of Computer Science & Engineering Washington University in St. Louis St. Louis MO United States Department of Otolaryngology-Head and Neck Surgery Washington University School of Medicine St. Louis MO United States Department of Psychology and Neuroscience University of North Carolina at Chapel Hill Chapel Hill NC United States Department of Anesthesiology Washington University St. Louis MO United States Department of Orthopedic Surgery Washington University School of Medicine St. Louis MO United States Department of Orthopedic Surgery Rady Children's Hospital University of California San Diego CA United States AI for Health Institute Washington University St. Louis MO United States Center for Technology and Behavioral Health Departments of Biomedical Data Science and Psychiatry Geisel School of Medicine at Dartmouth Lebanon NH United States

BACKGROUND CONTEXT: Understanding early postoperative recovery is crucial for improving perioperative care and long-term outcomes. Traditional recovery assessments relying primarily on cross-sectional patient-reported measures may not fully capture the complexity of the recovery process. PURPOSE: Our study evaluates early postoperative recovery using mobile health assessment and explores the relationship between early recovery patterns, multidomain recovery, and long-term outcomes. STUDY DESIGN/SETTING: A prospective observational study. PATIENT SAMPLE: A total of 129 patients were included in the analysis. OUTCOME MEASURES: Patient-reported outcome measures (PROMs) of pain, physical function, and disability at 1, 6 and 12 months postoperatively. METHODS: This study recruited patients aged 21 to 85 undergoing lumbar/thoracolumbar surgery for degenerative diseases. During the first month postoperatively, patients completed daily Ecological Momentary Assessment (EMA) surveys assessing pain, depression, and physical function and were passively monitored with Fitbit. EMA and Fitbit data (steps per minute, Maximum 30-minute cadence, and activity bout number) were clustered using functional principal component analysis. RESULTS: 129 patients were included (median age: 62;56% female). Two dominant recovery patterns were identified across all domains. At 12 months, patients with more favorable recovery across all domains except for activity bout number and depression, showed better outcomes. Favorable recovery in pain intensity and steps per minute was associated with greater improvements in all outcomes, while favorable recovery in maximum 30-minute cadence was associated with greater improvement in disability and physical function. Patients with less favorable pain recovery had a significantly higher complication rate (23% vs. 7%). A greater number of favorable recovery domains was associated with better 12-month outcomes improvement and demonstrated superior prognostic v

关键词： Degenerative spine disease Disability Ecological momentary assessment Fitbit Long-term outcomes Physical function Spine surgery

来源：评论

学校读者我要写书评

暂无评论

WSSS4LUAD: Grand Challenge on Weakly-supervised Tissue Semantic Segmentation for Lung Adenocarcinoma

arXiv

引用

arXiv 2022年

作者： Han, Chu Pan, Xipeng Yan, Lixu Lin, Huan Li, Bingbing Yao, Su Lv, Shanshan Shi, Zhenwei Mai, Jinhai Lin, Jiatai Zhao, Bingchao Xu, Zeyan Wang, Zhizhen Wang, Yumeng Zhang, Yuan Wang, Huihui Zhu, Chao Lin, Chunhui Mao, Lijian Wu, Min Duan, Luwen Zhu, Jingsong Hu, Dong Fang, Zijie Chen, Yang Zhang, Yongbing Li, Yi Zou, Yiwen Yu, Yiduo Li, Xiaomeng Li, Haiming Cui, Yanfen Han, Guoqiang Xu, Yan Xu, Jun Yang, Huihua Li, Chunming Liu, Zhenbing Lu, Cheng Chen, Xin Liang, Changhong Zhang, Qingling Liu, Zaiyi The Department of Radiology Guangdong Provincial People's Hospital Guangdong Academy of Medical Sciences Guangzhou510080 China Guangdong Prov. Key Laboratory of Artificial Intelligence in Medical Image Analysis and Application Guangdong Provincial People's Hospital Guangdong Academy of Medical Sciences Guangzhou510080 China The Department of Pathology Guangdong Provincial People's Hospital Guangdong Academy of Medical Sciences Guangzhou510080 China The School of Computer Science and Information Security Guilin University of Electronic Technology Guangxi541004 China The Department of Pathology Guangdong Provicial People's Hospital Ganzhou Hospital Ganzhou China The College of Artificial Intelligence Beijing University of Posts and Telecommunications Beijing100876 China The School of Computer Science and Engineering South China University of Technology Guangzhou510640 China The Institute for AI in Medicine School of Artificial Intelligence Nanjing University of Information Science and Technology Nanjing210044 China The School of Biological Science and Medical Engineering State Key Laboratory of Software Development Environment Key Laboratory of Biomechanics and Mechanobiology of Ministry of Education Research Institute of Beihang University in Shenzhen Beijing Advanced Innovation Center for Biomedical Engineering Beihang University Beijing100191 China The Department of Radiology Guangzhou First People's Hospital Hospital of South China University of Technology Guangzhou510180 China The Zhejiang Dahua Technology Co. Ltd Zhejiang310053 China The School of Biomedical Engineering University of Science and Technology of China Hefei230022 China The Tsinghua Shenzhen International Graduate School Tsinghua University Shenzhen China Electronic and Computer Engineering & Computer Science and Engineering The Hong Kong University of Science and Technology Hong Kong Hong Kong The School of Automation Science and Engineering Xi'An Jiaotong University Xian710049 Chi

Lung cancer is the leading cause of cancer death worldwide, and adenocarcinoma (LUAD) is the most common subtype. Exploiting the potential value of the histopathology images can promote precision medicine in oncology. Tissue segmentation is the basic upstream task of histopathology image analysis. Existing deep learning models have achieved superior segmentation performance but require sufficient pixel-level annotations, which is time-consuming and expensive. To enrich the label resources of LUAD and to alleviate the annotation efforts, we organize this challenge WSSS4LUAD to call for the outstanding weakly-supervised semantic segmentation (WSSS) techniques for histopathology images of LUAD. Participants have to design the algorithm to segment tumor epithelial, tumor-associated stroma and normal tissue with only patch-level labels. This challenge includes 10,091 patch-level annotations (the training set) and over 130 million labeled pixels (the validation and test sets), from 87 WSIs (67 from GDPH, 20 from TCGA). All the labels were generated by a pathologist-in-the-loop pipeline with the help of ai models and checked by the label review board. Among 532 registrations, 28 teams submitted the results in the test phase with over 1,000 submissions. Finally, the first place team achieved mIoU of 0.8413 (tumor: 0.8389, stroma: 0.7931, normal: 0.8919). According to the technical reports of the top-tier teams, CAM is still the most popular approach in WSSS. Cutmix data augmentation has been widely adopted to generate more reliable samples. With the success of this challenge, we believe that WSSS approaches with patch-level annotations can be a complement to the traditional pixel annotations while reducing the annotation efforts. The entire dataset has been released to encourage more researches on computational pathology in LUAD and more novel WSSS techniques. © 2022, CC BY.

关键词： Semantic Segmentation

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

arXiv

引用

arXiv 2022年

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 ArborMI United States Department of Internal Medicine University of Michigan Ann ArborMI United States Department of Pediatrics University of Michigan Ann ArborMI United States Department of Computational Medicine and Bioinformatics University of Michigan Ann ArborMI United States Michigan Institute for Data Science University of Michigan Ann ArborMI United States Department of Emergency Medicine University of Michigan Ann ArborMI United States Department of Computer Science and Engineering University of Michigan Ann ArborMI United States Max Harry Weil Institute for Critical Care Research and Innovation University of Michigan Ann ArborMI United States Mary H. Weiser Food Allergy Center University of Michigan Ann ArborMI United States Department of Computer Science Queens College City University of New York New YorkNY 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. Copyright © 2022, The Authors. All rights reserved.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

FedMint: Intelligent Bilateral Client Selection in Federated Learning with Newcomer IoT Devices

arXiv

引用

arXiv 2022年

作者： Wehbi, Osama Arisdakessian, Sarhad Wahab, Omar Abdel Otrok, Hadi Otoum, Safa Mourad, Azzam Guizani, Mohsen Cyber Security Systems and Applied AI Research Center Department of CSM Lebanese American University Lebanon Department of Computer and Software Engineering Polytechnique Montréal MontrealQC Canada Khalifa University Abu Dhabi United Arab Emirates College of Technological Innovation Zayed University Dubai United Arab Emirates Division of Science New York University Abu Dhabi United Arab Emirates Mohammad Bin Zayed University of Artificial Intelligence Abu Dhabi United Arab Emirates

Federated Learning (FL) is a novel distributed privacy-preserving learning paradigm, which enables the collaboration among several participants (e.g., Internet of Things devices) for the training of machine learning models. However, selecting the participants that would contribute to this collaborative training is highly challenging. Adopting a random selection strategy would entail substantial problems due to the heterogeneity in terms of data quality, and computational and communication resources across the participants. Although several approaches have been proposed in the literature to overcome the problem of random selection, most of these approaches follow a unilateral selection strategy. In fact, they base their selection strategy on only the federated server’s side, while overlooking the interests of the client devices in the process. To overcome this problem, we present in this paper FedMint, an intelligent client selection approach for federated learning on IoT devices using game theory and bootstrapping mechanism. Our solution involves the design of: (1) preference functions for the client IoT devices and federated servers to allow them to rank each other according to several factors such as accuracy and price, (2) intelligent matching algorithms that take into account the preferences of both parties in their design, and (3) bootstrapping technique that capitalizes on the collaboration of multiple federated servers in order to assign initial accuracy value for the newly connected IoT devices. Based on our simulation findings, our strategy surpasses the VanillaFL selection approach in terms of maximizing both the revenues of the client devices and accuracy of the global federated learning model. © 2022, CC BY-NC-SA.

关键词： Game theory

来源：评论

学校读者我要写书评

暂无评论

The 2022 report of the Lancet Countdown on health and climate change: health at the mercy of fossil fuels (vol 400, pg 1619, 2022)

引用

LANCET 2022年第10365期400卷 1766-1766页

作者： Romanello, M. Di Napoli, C. Drummond, P. Institute for Global Health University College London London UK School of Agriculture Policy and Development University of Reading Reading UK Institute for Sustainable Resources University College London London UK Department of Global Health Centre for Health and the Global Environment University of Washington Seattle WA USA UCL Energy Institute University College London London UK Department of Health Sciences University of York York UK Department of Meteorology University of Reading Reading UK Institute for Risk and Disaster Reduction University College London London UK School of Earth and Environment University of Leeds Leeds UK Centre on Climate Change and Planetary Health London School of Hygiene & Tropical Medicine London UK School of Population Health University of Melbourne Melbourne VIC Australia Department of Earth System Science Tsinghua University Beijing China Mercator Research Institute on Global Commons and Climate Change Berlin Germany Department of Environment Climate Change and Health World Health Organization Geneva Switzerland Institute of Environmental Sciences University of Geneva Geneva Switzerland Cardiovascular Epidemiology Unit Department of Public Health & Primary Care University of Cambridge Cambridge UK School of Government University of Birmingham Birmingham UK Economic Analysis of Climate Impacts and Policy Division Centro Euro-Mediterraneo sui Cambiamenti Climatici Venice Italy Institute for Environmental Design and Engineering University College London London UK Natural Resources Institute University of Greenwich London UK Department of Environmental Health Sciences and Yale Center on Climate Change and Health Yale University New Haven CT USA Department of Civil and Environmental Engineering Northeastern University Boston MA USA Department of Fish and Wildlife Conservation Virginia Polytechnic Institute and State University Blacksburg VA USA Department of Geography University College London London UK NU

With advancements in the science of detection and attribution studies, the influence of climate change over many events has now been quantified. Because of the rapidly increasing temperatures, vulnerable populations (adults older than 65 years, and children younger than one year of age) were exposed to 3·7 billion more heatwave days in 2021 than annually in 1986–2005 (indicator 1.1.2), and heat-related deaths increased by 68% between 2000–04 and 2017–21 (indicator 1.1.5), a death toll that was significantly exacerbated by the confluence of the COVID-19 pandemic. Despite the local cooling and overall health benefits of urban greenspaces, only 277 (27%) of 1038 global urban centres were at least moderately green in 2021 (indicator 2.2.3), and the number of households with air conditioning increased by 66% from 2000 to 2020, a maladaptive response that worsens the energy crisis and further increases urban heat, air pollution, and greenhouse gas emissions. [...]millions of people do not have access to the energy needed to keep their homes at healthy temperatures, preserve food and medication, and meet the seventh Sustainable Development Goal (to ensure access to affordable, reliable, sustainable, and modern energy for all). Aggravating this situation even further, governments continue to incentivise fossil fuel production and consumption: 69 (80%) of 86 countries reviewed had net-negative carbon prices (ie, provided a net subsidy to fossil fuels) for a net total of US$400 billion

关键词：

来源：评论

学校读者我要写书评

暂无评论

Correlation Analysis with Experimental Procedures and Peak Structures in Xps Spectral Round-Robin Tests on Mno Powder Sample

SSRN

引用

SSRN 2022年

作者： Murakami, Ryo Harada, Yoshitomo Sonobayashi, Yutaka Oji, Hiroshi Makino, Hisao Tanaka, Hiromi Taguchi, Hideyuki Sakamoto, Takanori Morita, Haruka Wakamori, Akihiko Kibe, Naoko Nishida, Shinsuke Nagata, Kenji Shinotsuka, Hiroshi Shouno, Hayaru Yoshikawa, Hideki The University of Electro-Communications Graduate School of Informatics and Engineering Chofu182-8585 Japan National Institute for Materials Science Research and Services Division of Materials Data and Integrated System Ibaraki Tsukuba305-0044 Japan Kyoto University Department of Materials Science and Engineering Kyoto606-8501 Japan Nagoya University Synchrotron radiation Research center Nagoya464-8603 Japan Kochi University of Technology School of Systems Engineering Japan National Institute of Technology Yonago College Department of Electrical and Computer Engineering Japan Nihon Parkerizing Co. Ltd Central Research Laboratories Kanagawa 254-0012 Japan JX Nippon Mining & Metals Corporation Technology Development Center Japan Melco Semiconductor Engineering Corporation Chemical Analysis & Evaluation Engineering Dept. Fukuoka Japan Kobelco Welding TechnoSolutions Co. Technical Solution Dept. Ltd Kanagawa 251-8551 Japan Furukawa Electric Co. Ltd Analysis Technology Center Sustainable Technology Laboratorie Kanagawa 220-0073 Japan

Spectral measurements provide valuable information on the electronic states and crystal structures of materials. In particular, X-ray photoelectron spectroscopy (XPS) can facilitate analysis of the chemical bond state of material surfaces. In the analysis of XPS spectra, the spectrum to be measured is typically compared to a reference spectrum measured on a known single-phase sample. Therefore, establishing a reference spectral database is imperative. In this study, round-robin tests of XPS spectra are performed with the aim of obtaining recommended XPS reference spectra and experimental procedures, including preprocessing. Determining the variations in peak structures between instrument- and sample-derived spectral shapes, and quantitatively discussing them in round-robin tests of XPS spectra are challenging. Therefore, an analysis method that automatically determines the above mentioned variations and extracts the samplederived common spectral structures is developed. This study presents a new approach to the quantitative analysis of round-robin tests of XPS spectra. The required common © 2022, The Authors. All rights reserved.

关键词： X ray photoelectron spectroscopy

来源：评论

学校读者我要写书评

暂无评论

Final Report of the NTCIR-14 FinNum Task: Challenges and Current Status of Fine-Grained Numeral Understanding in Financial Social Media data 14th

Final Report of the NTCIR-14 FinNum Task: Challenges and Cur...

引用

14th International Conference on NII Testbeds and Community for Information Access research, NTCIR 2019

作者： Chen, Chung-Chi Huang, Hen-Hsen Takamura, Hiroya Chen, Hsin-Hsi Department of Computer Science and Information Engineering National Taiwan University Taipei Taiwan Department of Computer Science National Chengchi University Taipei Taiwan Artificial Intelligence Research Center National Institute of Advanced Industrial Science and Technology Tokyo Japan MOST Joint Research Center for AI Technology and All Vista Healthcare Taipei Taiwan

ISBN: (纸本)9783030368043

NTCIR-14 FinNum task aims to disambiguate the meanings of the numerals in financial social media data from both coarse-grained taxonomy and fine-grained taxonomy. This task attracted 13 teams from 15 institutions in 6 different countries to register, received 16 submissions from 9 participants, and finally accepted 6 papers from participants. In this paper, we provide the analysis of the proposed dataset in depth, and discuss the challenges of fine-grained numeral understanding in financial social media data. © 2019, Springer Nature Switzerland AG.

关键词： Taxonomies

来源：评论

学校读者我要写书评

暂无评论

JARVIS-Leaderboard: A Large Scale Benchmark of Materials Design Methods

arXiv

引用

arXiv 2023年

作者： Choudhary, Kamal Wines, Daniel Garrity, Kevin F. Williams, Maureen Tavazza, Francesca Li, Kangming Hattrick-Simpers, Jason Gupta, Vishu Romero, Aldo H. Krogel, Jaron T. Saritas, Kayahan Fuhr, Addis Ganesh, Panchapakesan Kent, Paul R.C. Yan, Keqiang Lin, Yuchao Ji, Shuiwang Blaiszik, Ben Reiser, Patrick Friederich, Pascal Agrawal, Ankit Tiwary, Pratyush Beyerle, Eric Minch, Peter Rhone, Trevor David Takeuchi, Ichiro Wexler, Robert B. Mannodi-Kanakkithodi, Arun Ertekin, Elif Mishra, Avanish Mathew, Nithin Wood, Mitchell Rohskopf, Andrew Dale Wang, Shih-Han Achenie, Luke E.K. Xin, Hongliang Biacchi, Adam J. Material Measurement Laboratory National Institute of Standards and Technology MD20899 United States Department of Materials Science and Engineering University of Toronto 27 King’s College Cir TorontoON Canada Department of Electrical and Computer Engineering Northwestern University EvanstonIL60208 United States Lewis-Sigler Institute for Integrative Genomics Princeton University PrincetonNJ08544 United States Ludwig Institute for Cancer Research Princeton University PrincetonNJ08544 United States Department of Physics and Astronomy West Virginia University MorgantownWV26506 United States Materials Science and Technology Division Oak Ridge National Laboratory Oak RidgeTN37831 United States Center for Nanophase Materials Science Oak Ridge National Laboratory Oak RidgeTN37831 United States Computational Sciences and Engineering Division Oak Ridge National Laboratory Oak RidgeTN37831 United States Department of Computer Science and Engineering Texas A&M University College StationTX77843 United States Globus University of Chicago IL60637 United States Data Science Learning Division National Lab ArgonneIL60439 United States Institute of Nanotechnology Karlsruhe Institute of Technology Kaiserstraße 12 Karlsruhe76131 Germany Institute of Theoretical Informatics Karlsruhe Institute of Technology Kaiserstraße 12 Karlsruhe76131 Germany Department of Chemistry and Biochemistry Institute for Physical Science and Technology University of Maryland College ParkMD20742 United States Department of Physics Applied Physics and Astronomy Rensselaer Polytechnic Institute TroyNY12180 United States Department of Materials Science and Engineering University of Maryland College ParkMD20742 United States Department of Chemistry Institute of Materials Science and Engineering Washington University in St. Louis St. LouisMO63130 United States School of Materials Engineering Purdue University West LafayetteIN47907 United States Department of Mechan

Lack of rigorous reproducibility and validation are significant hurdles for scientific development across many fields. Materials science, in particular, encompasses a variety of experimental and theoretical approaches that require careful benchmarking. Leaderboard efforts have been developed previously to mitigate these issues. However, a comprehensive comparison and benchmarking on an integrated platform with multiple data modalities with perfect and defect materials data is still lacking. This work introduces JARVIS-Leaderboard, an open-source and community-driven platform that facilitates benchmarking and enhances reproducibility. The platform allows users to set up benchmarks with custom tasks and enables contributions in the form of dataset, code, and meta-data submissions. We cover the following materials design categories: Artificial Intelligence (ai), Electronic Structure (ES), Force-fields (FF), Quantum Computation (QC) and Experiments (EXP). For ai, we cover several types of input data, including atomic structures, atomistic images, spectra, and text. For ES, we consider multiple ES approaches, software packages, pseudopotentials, materials, and properties, comparing results to experiment. For FF, we compare multiple approaches for material property predictions. For QC, we benchmark Hamiltonian simulations using various quantum algorithms and circuits. Finally, for experiments, we use the inter-laboratory approach to establish benchmarks. There are 1281 contributions to 274 benchmarks using 152 methods with more than 8 million data-points, and the leaderboard is continuously expanding. The JARVIS-Leaderboard is available at the website: https://***/jarvis_leaderboard/ © 2023, CC BY.

关键词： Electronic structure

来源：评论

学校读者我要写书评

暂无评论

µVulDeePecker: A deep learning-based system for multiclass vulnerability detection

arXiv

引用

arXiv 2020年

作者： Zou, Deqing Wang, Sujuan Xu, Shouhuai Li, Zhen Jin, Hai National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab Big Data Security Engineering Research Center School of Cyber Science and Engineering Huazhong University of Science and Technology Wuhan430074 China Shenzhen Huazhong University of Science and Technology Research Institute Shenzhen518057 China National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab Big Data Security Engineering Research Center School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China Department of Computer Science University of Texas at San Antonio San AntonioTX78249 United States

Fine-grained software vulnerability detection is an important and challenging problem. Ideally, a detection system (or detector) not only should be able to detect whether or not a program contains vulnerabilities, but also should be able to pinpoint the type of a vulnerability in question. Existing vulnerability detection methods based on deep learning can detect the presence of vulnerabilities (i.e., addressing the binary classification or detection problem), but cannot pinpoint types of vulnerabilities (i.e., incapable of addressing multiclass classification). In this paper, we propose the first deep learning-based system for multiclass vulnerability detection, dubbed µVulDeePecker. The key insight underlying µVulDeePecker is the concept of code attention, which can capture information that can help pinpoint types of vulnerabilities, even when the samples are small. For this purpose, we create a dataset from scratch and use it to evaluate the effectiveness of µVulDeePecker. Experimental results show that µVulDeePecker is effective for multiclass vulnerability detection and that accommodating control-dependence (other than data-dependence) can lead to higher detection capabilities. Copyright © 2020, 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

MedAlign: A Clinician-Generated dataset for Instruction Following with Electronic Medical Records

arXiv

引用

arXiv 2023年

作者： Fleming, Scott L. Lozano, Alejandro Haberkorn, William J. Jindal, Jenelle A. Reis, Eduardo Thapa, Rahul Blankemeier, Louis Genkins, Julian Z. Steinberg, Ethan Nayak, Ashwin Patel, Birju Chiang, Chia-Chun Callahan, Alison Huo, Zepeng Gatidis, Sergios Adams, Scott Fayanju, Oluseyi Shah, Shreya J. Savage, Thomas Goh, Ethan Chaudhari, Akshay S. Aghaeepour, Nima Sharp, Christopher Pfeffer, Michael A. Liang, Percy Chen, Jonathan H. Morse, Keith E. Brunskill, Emma P. Fries, Jason A. Shah, Nigam H. Department of Biomedical Data Science Stanford School of Medicine StanfordCA United States Department of Computer Science Stanford School of Engineering StanfordCA United States Department of Anesthesiology Peri-operative and Pain Medicine Stanford School of Medicine StanfordCA United States Department of Pediatrics Stanford School of Medicine StanfordCA United States Stanford Center for Biomedical Informatics Research Stanford University StanfordCA United States Department of Radiology Stanford School of Medicine StanfordCA United States Stanford University StanfordCA United States Hospital Israelita Albert Einstein SP Sao Paulo Brazil Technology and Digital Solutions Stanford Health Care Palo AltoCA United States Department of Electrical Enginering Stanford School of Engineering StanfordCA United States Department of Medicine Vanderbilt University School of Medicine NashvilleTN United States Department of Biomedical Informatics Vanderbilt University Medical Center NashvilleTN United States Department of Medicine Stanford School of Medicine StanfordCA United States Department of Neurology Mayo Clinic RochesterMN United States Human-Centered Artificial Intelligence Institute Stanford University StanfordCA United States Division of Hospital Medicine Stanford University StanfordCA United States Clinical Excellence Research Center Stanford School of Medicine StanfordCA United States

The ability of large language models (LLMs) to follow natural language instructions with human-level fluency suggests many opportunities in healthcare to reduce administrative burden and improve quality of care. However, evaluating LLMs on realistic text generation tasks for healthcare remains challenging. Existing question answering datasets for electronic health record (EHR) data fail to capture the complexity of information needs and documentation burdens experienced by clinicians. To address these challenges, we introduce MedAlign, a benchmark dataset of 983 natural language instructions for EHR data. MedAlign is curated by 15 clinicians (7 specialities), includes clinician-written reference responses for 303 instructions, and provides 276 longitudinal EHRs for grounding instruction-response pairs. We used MedAlign to evaluate 6 general domain LLMs, having clinicians rank the accuracy and quality of each LLM response. We found high error rates, ranging from 35% (GPT-4) to 68% (MPT-7B-Instruct), and 8.3% drop in accuracy moving from 32k to 2k context lengths for GPT-4. Finally, we report correlations between clinician rankings and automated natural language generation metrics as a way to rank LLMs without human review. MedAlign is provided under a research data use agreement1 to enable LLM evaluations on tasks aligned with clinician needs and preferences. © 2023, CC BY.

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：