检索结果-内蒙古大学图书馆

Towards High Generalization Performance on Electrocardiogram Classification

学校读者我要写书评

暂无评论

Towards High Generalization Performance on Electrocardiogram...

Computers in Cardiology (CinC)

作者： Hyeongrok Han Seongjae Park Seonwoo Min Hyun-Soo Choi Eunji Kim Hyunki Kim Sangha Park Jinkook Kim Junsang Park Junho An Kwanglo Lee Wonsun Jeong Sangil Chon Kwonwoo Ha Myungkyu Han Sungroh Yoon Seoul National University Seoul South Korea HUINNO Co. Ltd. Seoul South Korea LG AI Research Seoul South Korea Kangwon National University Chuncheon South Korea Interdisciplinary Program in Bioinformatics Interdisciplinary Program in Artificial Intelligence ASRI INMC and Institute of Engineering Research Seoul National University Seoul South Korea

ISBN: (纸本)9781665467216

Recently, many electrocardiogram (ECG) classification algorithms using deep learning have been proposed. The characteristics of ECG vary from dataset to dataset for various reasons (i.e., hospital, race, etc.). Therefore, it is important for a model to have high generalization performance consistently over all datasets. In this paper, as part of the PhysioNet / Computing in Cardiology Challenge 2021, we present a model developed to classify cardiac abnormalities from 12 lead and reduced-lead ECGs. In particular, to upgrade our previous model for improving generalization performance, we newly adopt constant-weighted cross-entropy loss, additional features, Mixup augmentation, and squeeze/excitation block, OneCycle learning rate scheduler, which are selected via evaluation of generalization performance using leave-one-dataset-out cross-validation setting. With the present model, our DSAIL_SNU team has received challenge scores of 0.55, 0.58, 0.58, 0.57 and 0.57 (ranked 2nd, 1st, 1st, 2nd, 2nd out of 39 teams) for the 12-lead, 6-lead, 4-lead, 3-lead, and 2-lead versions of the hidden test set, respectively. The present model achieves higher generalization performance over all versions of the hidden test set than the model submitted last year.

关键词： Deep learning Hospitals Computational modeling Electrocardiography Lead Classification algorithms Cardiology

TargetNet: Functional microRNA target prediction with deep neural networks

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Min, Seonwoo Lee, Byunghan Yoon, Sungroh Department of Electrical and Computer Engineering Seoul National University Seoul08826 Korea Republic of LG AI Research Seoul07796 Korea Republic of Department of Electronic and IT Media Engineering Seoul National University of Science and Technology Seoul01811 Korea Republic of Interdisciplinary Program in Artificial Intelligence and Interdisciplinary Program in Bioinformatics Seoul National University Seoul08826 Korea Republic of

Motivation: MicroRNAs (miRNAs) play pivotal roles in gene expression regulation by binding to target sites of messenger RNAs (mRNAs). While identifying functional targets of miRNAs is of utmost importance, their prediction remains a great challenge. Previous computational algorithms have major limitations. They use conservative candidate target site (CTS) selection criteria mainly focusing on canonical site types, rely on laborious and time-consuming manual feature extraction, and do not fully capitalize on the information underlying miRNA-CTS interactions. Results: In this paper, we introduce TargetNet, a novel deep learning-based algorithm for functional miRNA target prediction. To address the limitations of previous approaches, TargetNet has three key components: (1) relaxed CTS selection criteria accommodating irregularities in the seed region, (2) a novel miRNA-CTS sequence encoding scheme incorporating extended seed region alignments, and (3) a deep residual network-based prediction model. The proposed model was trained with miRNA-CTS pair datasets and evaluated with miRNA-mRNA pair datasets. TargetNet advances the previous state-of-theart algorithms used in functional miRNA target classification. Furthermore, it demonstrates great potential for distinguishing high-functional miRNA targets. © 2021, CC BY-NC-ND.

关键词： RNA

Characterizing the mechanical response of metallic glasses to uniaxial tension using a spring network model

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Nawano, Aya Schroers, Jan Shattuck, Mark D. O’Hern, Corey S. Department of Mechanical Engineering and Materials Science Yale University New HavenCT06520 United States Benjamin Levich Institute and Physics Department The City College of New York New YorkNY10031 United States Department of Physics Yale University New HavenCT06520 United States Department of Applied Physics Yale University New HavenCT06520 United States Graduate Program in Computational Biology and Bioinformatics Yale University New HavenCT06520 United States

Metallic glasses are frequently used as structural materials. Therefore, it is important to develop methods to predict their mechanical response as a function of the microstructure prior to loading. We develop a novel coarse-grained spring network model, which describes the mechanical response of metallic glasses using an equivalent series network of springs, which can break and re-form to mimic atomic rearrangements during deformation. To validate the spring network model, we perform numerical simulations of quasistatic, uniaxial tensile deformation of Lennard-Jones and embedded atom method (EAM) potentials for Cu50Zr50 metallic glasses in the absence of large-scale shear band formation. We consider samples prepared using a wide range of cooling rates and with different amounts of crystalline order. We show that both the Lennard-Jones and EAM models possess qualitatively similar stress σ versus strain γ curves. By specifying five parameters in the spring network model (ultimate strength, strain at ultimate strength, slopes of σ(γ) at γ = 0 and at large strain, and strain at fracture where σ = 0), we can accurately describe the form of the stress-strain curves during uniaxial tension for the computational studies of Cu50Zr50, as well as recent experimental studies of several Zr-based metallic glasses. For the computational studies of Cu50Zr50, we find that the yield strain distribution is shifted to larger strains for slowly cooled glasses compared to rapidly cooled glasses. In addition, the average number of new springs and their rate of formation decreases with decreasing cooling rate. These effects offset each other at large strains, causing the stress-strain curve to become independent of the sample preparation protocol in this regime. In future studies, we will extract the parameters that define the spring network model directly from atomic rearrangements that occur during uniaxial deformation. © 2023, CC BY-NC-ND.

关键词： Metallic glass

MorphoITH: A Framework for Deconvolving Intra-Tumor Heterogeneity Using Tissue Morphology

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Nielsen, Aleksandra Weronika Manoochehri, Hafez Eslami Zhong, Hua Panwar, Vandana Jarmale, Vipul Jasti, Jay Nourani, Mehrdad Rakheja, Dinesh Brugarolas, James Kapur, Payal Rajaram, Satwik Lyda Hill Department of Bioinformatics University of Texas Southwestern Medical Center DallasTX United States Department of Electrical and Computer Engineering The University of Texas at Dallas RichardsonTX United States Department of Pathology University of Texas Southwestern Medical Center DallasTX United States Kidney Cancer Program Simmons Comprehensive Cancer Center University of Texas Southwestern Medical Center DallasTX United States University of Texas Southwestern Medical Center DallasTX United States Department of Urology University of Texas Southwestern Medical Center at Dallas DallasTX United States

The ability of tumors to evolve and adapt by developing subclones in different genetic and epigenetic states is a major challenge in oncology. Traditional tools like multi-regional sequencing used to study tumor evolution and the resultant intra-tumor heterogeneity (ITH) are often impractical because of their resource-intensiveness and limited scalability. Here, we present MorphoITH, a novel framework that leverages histopathology slides to deconvolve molecular ITH through tissue morphology. MorphoITH integrates a self-supervised deep learning similarity measure to capture phenotypic variation across multiple dimensions (cytology, architecture, and microenvironment) with rigorous methods to eliminate spurious sources of variation. Using a prototype of ITH, clear cell renal cell carcinoma (ccRCC), we show that MorphoITH captures clinically-significant biological features, such as vascular architecture and nuclear grades. Furthermore, we find that MorphoITH recognizes differential biological states corresponding to subclonal changes in key driver genes (BAP1/PBRM1/SETD2). Finally, by applying MorphoITH to a multi-regional sequencing experiment, we postulate evolutionary trajectories that largely recapitulate genetic evolution. In summary, MorphoITH provides a scalable phenotypic lens that bridges the gap between histopathology and genomics, advancing precision oncology. © 2025, CC BY-NC-ND.

关键词： Tumors

DIFFUSION CURVATURE FOR ESTIMATING LOCAL CURVATURE IN HIGH DIMENSIONAL DATA

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Bhaskar, Dhananjay MacDonald, Kincaid Fasina, Oluwadamilola Thomas, Dawson Rieck, Bastian Adelstein, Ian Krishnaswamy, Smita Department of Genetics Yale School of Medicine New HavenCT United States Department of Mathematics Yale University New HavenCT United States Program for Applied Mathematics Yale University New HavenCT United States Department of Physics Yale University New HavenCT United States Institute of AI for Health Helmholtz Pioneer Campus Munich Germany Department of Computer Science Yale University New HavenCT United States Program for Computational Biology and Bioinformatics Yale University New HavenCT United States

We introduce a new intrinsic measure of local curvature on point-cloud data called diffusion curvature. Our measure uses the framework of diffusion maps, including the data diffusion operator, to structure point cloud data and define local curvature based on the laziness of a random walk starting at a point or region of the data. We show that this laziness directly relates to volume comparison results from Riemannian geometry. We then extend this scalar curvature notion to an entire quadratic form using neural network estimations based on the diffusion map of point-cloud data. We show applications of both estimations on toy data, single-cell data and on estimating local Hessian matrices of neural network loss landscapes. © 2022, CC BY.

关键词： Geometry

Multiomic Mapping of Copy Number and Structural Variation on Chromosome 1 (Chr1) Highlights Multiple Recurrent Disease Drivers

学校读者我要写书评

暂无评论

BLOOD 2021年第Sup1期138卷 721-721页

作者： Blaney, Patrick Boyle, Eileen M. Wang, Yubao Ghamlouch, Hussein Choi, Jinyoung Williams, Louis James, Stoeckle Siegel, Ariel Razzo, Beatrice Braunstein, Marc Kaminetzky, David Arbini, Arnaldo A. Bruno, Benedetto Corre, Jill Montes, Lydia Auclair, Daniel Davies, Faith E. Tsirigos, Aristotelis Rustad, Even H. Maura, Francesco Landgren, Ola Bauer, Michael A. Walker, Brian Morgan, Gareth Perlmutter Cancer Center NEW YORK UNIVERISITY New York NY Perlmutter Cancer Center New York University Langone Health New York NY Perlmutter Cancer Center Myeloma Research Program New York NY NYU Langone Health Myeloma Research Program Perlmutter Cancer Center NYU Grossman School of Medicine New York NY Myeloma Research Program Laura and Isaar Perlmutter Cancer Center NEW YORK UNIVERSITY New York NY Laura & Isaac Perlmutter Cancer Center NEW YORK UNIVERSITY SCHOOL OF MEDICINE New York NY Myeloma Research Program NYU Permutter's cancer center New-York NY Myeloma Research Program NYU Langone's Perlmutter Cancer Center New York NY Myeloma Research Program Perlmutter Cancer Center NYU Langone Health New York NY Blood and Marrow Transplantation and Cellular Therapy New York University Langone Health New York NY Unité de Génomique du Myélome IUCT- Oncopole Toulouse France Hematology department CHU Amiens Amiens France The MMRF Norwalk CT Perlmutter Cancer Center NYU Langone Health New York NY Applied Bioinformatics Laboratories NYU Langone New York NY Department of Clinical and Molecular Medicine Norwegian University of Science and Technology Trondheim Norway Myeloma Program Sylvester Comprehensive Cancer Center Miami Department of Biomedical Informatics University of Arkansas for Medical Sciences Little Rock AR Division of Hematology Oncology Melvin and Bren Simon Comprehensive Cancer Center Indianapolis IN NYU Langone Health New York University School of Medicine New York NY

Introduction Copy number abnormalities (CNA) and structural variants (SV) are crucial to driving cancer progression and in multiple myeloma (MM). Chr1 CNA are seen in up to 40% of cases and associate with poor prognosis. Variants include deletions, gains, translocations and complex SV events such as chromothripsis (CT), chromoplexy (CP) and templated insertions (TI) which result in aberrant transcriptional patterns. Abnormal expression of genes on chr1 lead to the adverse clinical outcome and studies focussed on 1p12, 1p32.3 and 1q12-21 identified potential causal genes including TENT5C, CDKN2C, CKS1B, PDZK1, BCL9, ANP32E, ILF2, ADAR, MDM2 and MCL1 but none fully explain the clinical behavior. To address this deficiency and to relate chromatin structure to gene deregulation we present a multiomic bioinformatic analysis of SV, CNA, mutation and expression changes in relation to the chromatin structure of *** We analysed data derived from 1,154 CoMMpass trial patients. We analyzed 972 NDMM patients with whole exome for mutations, and 752 whole genomes for copy number, translocations, complex rearrangements such as CP, CT and TI as previously described. Using GISTIC 2.0, we identified hotspots of CNA. This information was then analyzed in conjunction to the RNA-seq data derived from 643 patients to determine the aberrant transcriptional landscape of chr1. Using HiC data derived from U266 MM cell line, we associated these changes with TAD structures, A/B compartments, and histone marks along chr1, to gene expression changes, and recurrent SV. Using the cell line dependency map for CRISPR knockdown of the gene set on chr1 derived from 20 MM cell lines we related cell viability to chr1 copy number *** •We identified 7 hotspots of deletion, 9 of gain, 3 of CT and 2 of templated-insertion across chr1. We mapped these regions to epigenetic plots and show that gained regions are hypomethylated compared to the rest of chr1 (Wilcoxon, p=0.0002). Overall

关键词：

Full-privacy secured search engine empowered by efficient genome-mapping algorithms

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Chang, Yuan-Yu Wong, Sheng-Tang Salawu, Emmanuel O. Wang, Yu-Xuan Hung, Jui-Hung Yang, Lee-Wei Institute of Bioinformatics and Structural Biology National Tsing Hua University Hsinchu300044 Taiwan Praexisio Taiwan Inc. New Taipei221425 Taiwan Bioinformatics Program Institute of Information Sciences Academia Sinica Taipei115201 Taiwan HerndonVA United States Department of Computer Science College of Computer Science National Yang Ming Chiao Tung University Hsinchu300044 Taiwan Physics Division National Center for Theoretical Sciences Taipei106216 Taiwan PhD Program in Biomedical Artificial Intelligence National Tsing Hua University Hsinchu300044 Taiwan

Search engine privacy has not been an option for web exploration in the past. Search history returns to users as personalized merchandise promotion from search engine companies. Here we demonstrate a privacy-guaranteed internet search method, provided with the entire article as a query, can be correctly carried out without revealing users' sensitive content by a degenerate, irreversible, encoding scheme and an efficient FM-index search routine, generally used in the next-generation sequencing. In our solution, Sapiens Aperio Veritas Engine (S.A.V.E.), every word in the query is encoded into one of 12 "amino acids" comprising a pseudo-biological sequence (PBS) at users' local machines. The PBS-mediated plagiarism detection is done by users' submission of locally encoded PBS through our cloud service to locate duplicates in the collected web contents, currently including all the English and Chinese Wikipedia pages and Open Access journal articles, as of April 2021, which had been encoded in the same way as the query. It is found that PBSs with a length longer than 12, comprising a combination of more than 12 "amino acids", can return correct results with a false positive rate © 2021, CC BY-NC-ND.

关键词： Search engines

Pollen and asthma morbidity in Atlanta: A 26-year time-series study

学校读者我要写书评

暂无评论

Environment International 2023年 177卷 107998页

作者： Lappe, Brooke L. Ebelt, Stefanie D'Souza, Rohan R. Manangan, Arie Brown, Claudia Saha, Shubhayu Harris, Drew Chang, Howard H. Sole, Adam Scovronick, Noah Gangarosa Department of Environmental Health Rollins School of Public Health Emory University AtlantaGA United States Department of Biostatistics and Bioinformatics Rollins School of Public Health Emory University AtlantaGA United States Climate and Health Program Division of Environmental Health Science and Practice National Center for Environmental Health Centers for Disease Control and Prevention AtlantaGA United States Office of Climate Change and Health Equity Department of Health and Human Services WashingtonDC United States Division of Pulmonary and Critical Care Medicine University of Virginia CharlottesvilleVA United States

Background: Compared to many environmental risk factors, the relationship between pollen and asthma is understudied, including how associations may differ by pollen type and between subgroups, and how associations may be changing over time. Objectives: We evaluated the association between ambient pollen concentrations and emergency department (ED) visits for asthma and wheeze in Atlanta, Georgia during 1993–2018. We estimated overall associations for 13 individual pollen taxa, as well as associations by decade, race, age (5–17, 18–64, 65+), and insurance status (Medicaid vs non-Medicaid). Methods: Speciated pollen data were acquired from Atlanta Allergy & Asthma, a nationally certified pollen counting station. ED visit data were obtained from individual hospitals and from the Georgia Hospital Association. We performed time-series analyses using quasi-Poisson distributed lag models, with primary analyses assessing 3-day (lag 0–2 days) pollen levels. Models controlled for day of week, holidays, air temperature, month, year, and month-by-year interactions. Results: From 1993 to 2018, there were 686,259 ED visits for asthma and wheeze in the dataset, and the number of ED visits increased over time. We observed positive associations of asthma and wheeze ED visits with nine of the 13 pollen taxa: trees (maple, birch, pine, oak, willow, sycamore, and mulberry), two weeds (nettle and pigweed), and grasses. Rate ratios indicated 1–8% increases in asthma and wheeze ED visits per standard deviation increases in pollen. In general, we observed stronger associations in the earliest period (1993–2000), in younger people, and in Black patients;however, results varied by pollen taxa. Conclusions: Some, but not all, types of pollen are associated with increased ED visits for asthma/wheeze. Associations are generally higher in Black and younger patients and appear to have decreased over time. © 2023

关键词： Hospitals

SCALE RELIANT INFERENCE

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Nixon, Michelle Pistner McGovern, Kyle C. Letourneau, Jeffrey David, Lawrence A. Lazar, Nicole A. Mukherjee, Sayan Silverman, Justin D. College of Information Sciences and Technology The Pennsylvania State University United States Program in Bioinformatics and Genomics The Pennsylvania State University United States Department of Molecular Genetics and Microbiology Duke University United States Department of Statistics The Pennsylvania State University United States Huck Institutes for the Life Sciences The Pennsylvania State University United States Departments of Statistical Science Mathematics Computer Science Biostatistics & Bioinformatics Duke University United States Center for Scalable Data Analytics and Artificial Intelligence University of Leipzig Germany Max Planck Institute for Mathematics in the Natural Sciences Germany Department of Medicine The Pennsylvania State University United States

Many scientific fields, including human gut microbiome science, collect multivariate count data where the sum of the counts is unrelated to the scale of the underlying system being measured (e.g., total microbial load in a subject’s colon). This disconnect complicates downstream analyses such as differential analysis in case-control studies. This article is motivated by a novel study of in vitro human gut microbiome models. Popular tools for analyzing these data led to dramatically elevated rates of both false positives and false negatives. To understand those failures, we provide a formal problem statement that frames these challenges of scale in terms of the classical theory of identifiability. We call this the problem of Scale Reliant Inference (SRI). We use this formulation to prove fundamental limits on SRI in terms of criteria such as consistency and type-I error control. We show that the failures of existing methods stem from a fundamental failure to properly quantify uncertainty in the system scale. We demonstrate that a particular type of Bayesian model called a Bayesian Partially Identified Model (PIMs) can correctly quantify uncertainty in SRI. We introduce Scale Simulation Random Variables (SSRVs) as a flexible and efficient approach to specifying and inferring Bayesian PIMs. In the context of both real and simulated data, we find SSRVs drastically decrease type-I and type-II error rates. © 2022, CC BY.

关键词： Bayesian networks