检索结果-内蒙古大学图书馆

Evaluation of GPT and BERT-based models on identifying protein-protein interactions in biomedical text

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Rehana, Hasin Çam, Nur Bengisu Basmaci, Mert Zheng, Jie Jemiyo, Christianah He, Yongqun Özgür, Arzucan Hur, Junguk Computer Science Graduate Program University of North Dakota Grand ForksND58202 United States Department of Computer Engineering Bogazici University Istanbul34342 Turkey Unit for Laboratory Animal Medicine Department of Microbiology and Immunology University of Michigan Ann ArborMI48109 United States Department of Biomedical Sciences University of North Dakota School of Medicine and Health Sciences Grand ForksND58202 United States Center for Computational Medicine and Bioinformatics University of Michigan Ann ArborMI48109 United States

Detecting protein-protein interactions (PPIs) is crucial for understanding genetic mechanisms, disease pathogenesis, and drug design. However, with the fast-paced growth of biomedical literature, there is a growing need for automated and accurate extraction of PPIs to facilitate scientific knowledge discovery. Pre-trained language models, such as generative pre-trained transformers (GPT) and bidirectional encoder representations from transformers (BERT), have shown promising results in natural language processing (NLP) tasks. We evaluated the performance of PPI identification of multiple GPT and BERT models using three manually curated gold-standard corpora: Learning Language in Logic (LLL) with 164 PPIs in 77 sentences, Human Protein Reference Database with 163 PPIs in 145 sentences, and Interaction Extraction Performance Assessment with 335 PPIs in 486 sentences. BERT-based models achieved the best overall performance, with BioBERT achieving the highest recall (91.95%) and F1-score (86.84%) and PubMedBERT achieving the highest precision (85.25%). Interestingly, despite not being explicitly trained for biomedical texts, GPT-4 achieved commendable performance, comparable to the top-performing BERT models. It achieved a precision of 88.37%, a recall of 85.14%, and an F1-score of 86.49% on the LLL dataset. These results suggest that GPT models can effectively detect PPIs from text data, offering promising avenues for application in biomedical literature mining. Further research could explore how these models might be fine-tuned for even more specialized tasks within the biomedical domain. © 2023, CC BY.

关键词： Extraction

Multi-branch Cascaded Swin Transformers with Attention to k-space Sampling Pattern for Accelerated MRI Reconstruction

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Ekanayake, Mevan Pawar, Kamlesh Harandi, Mehrtash Egan, Gary Chen, Zhaolin Monash Biomedical Imaging Monash University ClaytonVIC3800 Australia Department of Electrical and Computer Systems Engineering Monash University ClaytonVIC3800 Australia School of Psychological Sciences Monash University ClaytonVIC3800 Australia Department of Data Science and AI Monash University ClaytonVIC3800 Australia

Global correlations are widely seen in human anatomical structures due to similarity across tissues and bones. These correlations are reflected in magnetic resonance imaging (MRI) scans as a result of close-range proton density and T1/T2 parameters in tissues. Furthermore, to achieve accelerated MRI, k-space data are undersampled which causes global aliasing artifacts. Convolutional neural network (CNN) models are widely utilized for accelerated MRI reconstruction, but those models are limited in capturing global correlations due to the intrinsic locality of the convolution operation. The recent self-attention-based transformer models are capable of capturing global correlations among image features, however, the current contributions of transformer models for MRI reconstruction are minute. The existing contributions mostly provide CNN-transformer hybrid solutions and rarely leverage the physics of MRI. In this paper, we propose a physics-based stand-alone (convolution free) transformer model titled, the Multi-branch Cascaded Swin Transformers (McSTRA) for accelerated MRI reconstruction. McSTRA combines several interconnected MRI physics-related concepts with the transformer networks: it exploits global MR features via the shifted window self-attention mechanism;it extracts MR features belonging to different spectral components separately using a multi-branch setup;it iterates between intermediate de-aliasing and k-space correction via a cascaded network with data consistency in k-space and intermediate loss computations;furthermore, we propose a novel positional embedding generation mechanism to guide self-attention utilizing the point spread function corresponding to the undersampling mask. With the combination of all these components, we propose a consolidated robust solution for MRI reconstruction which shows improved performance by significantly outperforming state-of-the-art MRI reconstruction solutions both visually and quantitatively. Further experiments dem

关键词： Image reconstruction

Smartphone-based Biofeedback Games for Shielding of Immune System against COVID-19: The CovidShield Game Suite design approach 2020

学校读者我要写书评

暂无评论

Smartphone-based Biofeedback Games for Shielding of Immune S...

9th International Conference on Software Development and Technologies for Enhancing Accessibility and Fighting Info-exclusion

作者： Sofia Dias Pedro Silva Ilias Chrysovergis Vasileios Charisis Georgios Tsoumalis Zafeirios Bampos Athanasios Papazoglou Dimitrios Iakovakis Stelios Hadjidimitriou Leontios Hadjileontiadis Herbert Jelinek Faculdade de Motricidade Humana Universidade de Lisboa PT Master-D Training CO Lisbon Portugal PT Department of Electrical & Computer Engineering Aristotle University of Thessaloniki GR 54124 Thessaloniki Greece GR Department of Electrical Engineering and Computer Science/Department of Biomedical Engineering Khalifa University of Science and Technology PO BOX: 127788 Abu Dhabi UAE GR Khalifa University of Science Technology & Research (KUSTAR) AE

ISBN: (纸本)9781450389372

Game-based biofeedback applications on smartphone devices can be played anywhere, giving users, at the same time, a tangible and real-time external measure of their behavior/response. The CovidShield Game Suite (CGS) design process approach presented here, as part of the CovidShield project, introduces a holistic interactive approach to shield the human immune system against COVID-19, by targeting optimal control of all the relevant factors, i.e., physical and mental, in an unobtrusive, gamified and user friendly, low-cost way, allowing for personalization and adaptation to each user's physiological and mental profile. From the methodological point of view, to facilitate the visualization of the constructed breathing game-scenarios, the storyboarding technique was adopted. Overall, the realization of the CGS sets the basis for establishing a holistic framework that could improve and/or change the quality of life of the user/player by following the learning process of breathing patterns that the CGS can offer.

关键词： CovidShield Game Suite Biofeedback Games Game design COVID-19 pandemic Immune System Smartphone

System Model Extraction of MEMS Beam Resonator for Colpitts Crystal Oscillator

学校读者我要写书评

暂无评论

System Model Extraction of MEMS Beam Resonator for Colpitts ...

International Conference on engineering and Emerging Technologies (ICEET)

作者： Faisal Nawaz Muhammad Sulaiman Umar Jamil Rana Iqtidar Shakoor Department of Electrical & Computer Engineering Air University Islamabad Pakistan Department of Mechatronics & Biomedical Engineering Air University Islamabad Pakistan

ISBN: (数字)9781728145815

ISBN: (纸本)9781728145822

A complementary metal oxide semiconductor-micro electro-mechanical systems (CMOS-MEMS) platform is applicable for integrating MEMS resonators and CMOS integrated circuits on a single chip. This provides a monolithic connection of MEMS with electronics. With this integration, different configurations of monolithically integrated MEMS resonators and CMOS amplifiers have been developed for commercial multi user purpose. It also provides variety of design flexibilities for resonator applications. In this paper, the design and analysis of a 2.14MHz MEMS beam resonator for Colpitts crystal oscillator circuit is presented. A MEMS beam resonator is designed to function as an oscillator. The proposed MEMSColpitts oscillator circuit can generate 1 to 5MHz clocks, when operated in different modes. This oscillator works specifically on 2.14MHz by macro model extraction for 9 th mode only. The beam resonator requires approximately 10V dc to operate. The actuation was simulated in IntelliSuite (thermo-electro mechanical module). A macro model of resonator is extracted by using system model extraction (SME) and utilized in electrical circuit of Colpitts crystal oscillator. The comparison with other research works is also presented.

关键词：

Identifying Cellular Neighborhood Phenotypes Differentiating Normal and Quiescent Crohn's Disease via MxIF

学校读者我要写书评

暂无评论

Identifying Cellular Neighborhood Phenotypes Differentiating...

Medical Imaging 2025: Digital and Computational Pathology

作者： Rudravaram, Gaurav Bao, Shunxing Krishnan, Aravind R. Remedios, Lucas W. Kim, Michael E. Saunders, Adam M. Liu, Qi Lau, Ken S. Roland, Joseph T. Washington, Mary K. Coburn, Lori A. Wilson, Keith T. Huo, Yuankai Landman, Bennett A. Department of Electrical and Computer Engineering Vanderbilt University NashvilleTN United States Department of Computer Science Vanderbilt University NashvilleTN United States Division of Gastroenterology Hepatology and Nutrition Department of Medicine Vanderbilt University Medical Center NashvilleTN United States Vanderbilt Center for Mucosal Inflammation and Cancer Vanderbilt University Medical Center NashvilleTN United States Program in Cancer Biology Vanderbilt University School of Medicine NashvilleTN United States Veterans Affairs Tennessee Valley Healthcare System NashvilleTN United States Department of Pathology Microbiology and Immunology Vanderbilt University Medical Center NashvilleTN United States Center for Quantitative Sciences Vanderbilt University Medical Center NashvilleTN United States Epithelial Biology Center Vanderbilt University Medical Center NashvilleTN United States Department of Cell and Developmental Biology Vanderbilt University School of Medicine NashvilleTN United States Vanderbilt University Department of Biomedical Engineering Nashville United States Department of Biostatistics Vanderbilt University Medical Center Nashville United States

ISBN: (纸本)9781510686045

Crohn's disease (CD) is a chronic inflammatory condition of the gastrointestinal tract characterized by alternating periods of activity and remission. This study investigates whether the cellular environments of quiescent (inactive disease) and normal tissues show similarities or differences, exploring potential residual effects from disease and tissue recovery patterns. While active CD can typically be characterized by high neutrophil density, whereas quiescent tissues present less pronounced inflammatory markers. This study leverages multiplexed immunofluorescence (MxIF) data to examine the distributions and patterns of various nuclei subtypes in quiescent versus normal tissues, aiming to determine how similar or different these tissues are in the absence of active inflammation. In our study, we analyze four tissue samples labeled as normal and five as quiescent from CD patients, as determined by a pathologist. We use rigid and deformable registration of MxIF images to analyze these tissues in a unified space and classify the nuclei into 13 subtypes. Using a KD tree, we identify the 10 nearest neighbors for each nucleus, building a combined distance-weighted neighborhood matrix representing both normal and quiescent tissues. We projected these neighborhoods into 2 dimensions using t-distributed stochastic nearest neighbor(t-SNE), clustered them, and calculated the enrichment for each subtype between normal and quiescent states within each of 20 and 40 clusters. In the 20-cluster analysis, we find monocytes enrich in one cluster and leukocytes enrich in two clusters for normal and quiescent tissues respectively. For the 40-cluster analysis, we observe monocytes, macrophages, and leukocytes enrich in one cluster each in normal tissues, and in the quiescent condition, one cluster enriched with enteroendocrine cells, and two clusters enriched with leukocytes. The results were statistically significant as determined by a chi-squared test (p © 2025 SPIE.

关键词： Tissues White blood cells Diseases and disorders Biological samples Matrices Statistical analysis Image segmentation Image registration Critical dimension metrology Inflammation

A Tissue Mechanics Based Method to Improve Tissue Displacement Estimation in Ultrasound Elastography*

学校读者我要写书评

暂无评论

A Tissue Mechanics Based Method to Improve Tissue Displaceme...

Annual International Conference of the IEEE engineering in Medicine and Biology Society (EMBC)

作者： Niusha Kheirkhah Sergio C. H. Dempsey Hassan Rivaz Abbas Samani Ali Sadeghi-Naini School of Biomedical Engineering Western University London Canada Department of Electrical and Computer Engineering Concordia University Montreal Canada Departments of Medical Biophysics and Electrical and Computer Engineering Western University London Canada Department of Electrical Engineering and Computer Science York University Toronto Canada

ISBN: (数字)9781728119908

ISBN: (纸本)9781728119915

Cancer is known to induce significant structural changes to tissue. In most cancers, including breast cancer, such changes yield tissue stiffening. As such, imaging tissue stiffness can be used effectively for cancer diagnosis. One such imaging technique, ultrasound elastography, has emerged with the aim of providing a low-cost imaging modality for effective breast cancer diagnosis. In quasi-static breast ultrasound elastography, the breast is stimulated by ultrasound probe, leading to tissue deformation. The tissue displacement data can be estimated using a pair of acquired ultrasound radiofrequency (RF) data pertaining to pre- and post-deformation states. The data can then be used within a mathematical framework to construct an image of the tissue stiffness distribution. Ultrasound RF data is known to include significant noise which lead to corruption of estimated displacement fields, especially the lateral displacements. In this study, we propose a tissue mechanics-based method aiming at improving the quality of estimated displacement data. We applied the method to RF data acquired from a tissue-mimicking phantom. The results indicated that the method is effective in improving the quality of the displacement data.

关键词： Mathematical model Ultrasonic imaging Strain Estimation Radio frequency Elastography

Tracing the emergence of gendered language in childhood 42

学校读者我要写书评

暂无评论

Tracing the emergence of gendered language in childhood

42nd Annual Meeting of the Cognitive Science Society: Developing a Mind: Learning in Humans, Animals, and Machines, CogSci 2020

作者： Prystawski, Ben Grant, Erin Nematzadeh, Aida Lee, Spike W.S. Stevenson, Suzanne Xu, Yang Computer Science and Cognitive Science Programs University of Toronto Canada Department of Electrical Engineering & Computer Sciences University of California Berkeley United States DeepMind Rotman School of Management Department of Psychology University of Toronto Canada Department of Computer Science University of Toronto Canada Department of Computer Science Cognitive Science Program University of Toronto Canada

Are gender associations in general language reflected in the words spoken to and by children? Previous work has suggested that language reveals gender differences in discourse, speech style, language use and acquisition. Work in artificial intelligence has shown that word embeddings trained on large corpora reflect human gender associations. We connect this work to developmental psychology by exploring whether gender associations in word embeddings are present in the linguistic input and output of children, and if so, how early gendered language emerges. We present a computational method that quantifies the gender associations of words and use a corpus of child-caretaker speech to show that these gender associations correlate significantly with those in word embeddings. We discover that gendered word use emerges in English-speaking children around age 2, and the gender associations cannot be explained solely by variables including word length, frequency, concreteness, and valence. © 2020 The Author(s)

关键词： Embeddings

Towards Robust and Scalable Dispatch Modeling of Long-Duration Energy Storage

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Guerra, Omar J. Dalvi, Sourabh Thatte, Amogh A. Cowiestoll, Brady Jorgenson, Jennie Hodge, Bri-Mathias National Renewable Energy Laboratory 15013 Denver West Parkway GoldenCO80401 United States Colorado School of Mines Advanced Energy Systems Graduate Program 1500 Illinois Street GoldenCO80401 United States Renewable and Sustainable Energy Institute University of Colorado BoulderCO80309 United States Department of Electrical Computer & Energy Engineering University of Colorado BoulderCO80309 United States

Energy storage technologies, including short-duration, long-duration, and seasonal storage, are seen as technologies that can facilitate the integration of larger shares of variable renewable energy, such as wind and solar photovoltaics, in power systems. However, despite recent advances in the techno-economic modeling of energy storage (particularly for short-duration applications), the operation and economics of long-duration energy storage are still incomplete in power systems modeling platforms. For instance, existing modeling approaches for long-duration storage are often based either on an oversimplified representation of power system operations or limited representation of storage technologies, e.g., evaluation of only a single application. This manuscript presents an overview of the challenges of modeling long-duration energy storage technologies, as well as a discussion regarding the capabilities and limitations of existing approaches. We used two test power systems with high shares of both solar photovoltaics- and wind (70% - 90% annual variable renewable energy shares) to assess long-duration energy storage dispatch approaches. Our results estimate that better dispatch modeling of long-duration energy storage could increase the associated operational value by 4% - 14% and increase the standard capacity credit by 14% - 34%. Thus, a better long-duration energy storage dispatch could represent significant cost saving opportunities for electric utilities and system operators. In addition, existing long-duration dispatch modeling approaches were tested in terms of both improved system value (e.g., based on production cost and standard capacity credit) and scalability (e.g., based on central processing unit time and peak memory usage). Both copper plate and nodal representations of the power system were considered. Although the end volume target dispatch approach, i.e., based on mid-term scheduling, showed promising performance in terms of both improved system

关键词： Scalability

Deep Learning for Detection and Localization of B-Lines in Lung Ultrasound

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Lucassen, Ruben T. Jafari, Mohammad H. Duggan, Nicole M. Jowkar, Nick Mehrtash, Alireza Fischetti, Chanel Bernier, Denie Prentice, Kira Duhaime, Erik P. Jin, Mike Abolmaesumi, Purang Heslinga, Friso G. Veta, Mitko Duran-Mendicuti, Maria A. Frisken, Sarah Shyn, Paul B. Golby, Alexandra J. Boyer, Edward Wells, William M. Goldsmith, Andrew J. Kapur, Tina Department of Radiology Brigham and Women’s Hospital Harvard Medical School BostonMA02115 United States The Department of Biomedical Engineering Eindhoven University of Technology Eindhoven5612 Netherlands The Department of Emergency Medicine Brigham and Women’s Hospital BostonMA02115 United States Centaur Labs BostonMA02116 United States The Department of Electrical and Computer Engineering The University of British Columbia VancouverBCV5T 1Z4 Canada The Departments of Neurosurgery and Radiology Brigham and Women’s Hospital Harvard Medical School BostonMA02115 United States

Lung ultrasound (LUS) is an important imaging modality used by emergency physicians to assess pulmonary congestion at the patient bedside. B-line artifacts in LUS videos are key findings associated with pulmonary congestion. Not only can the interpretation of LUS be challenging for novice operators, but visual quantification of B-lines remains subject to observer variability. In this work, we investigate the strengths and weaknesses of multiple deep learning approaches for automated B-line detection and localization in LUS videos. We curate and publish, BEDLUS, a new ultrasound dataset comprising 1,419 videos from 113 patients with a total of 15,755 expert-annotated B-lines. Based on this dataset, we present a benchmark of established deep learning methods applied to the task of B-line detection. To pave the way for interpretable quantification of B-lines, we propose a novel "single-point" approach to B-line localization using only the point of origin. Our results show that (a) the area under the receiver operating characteristic curve ranges from 0.864 to 0.955 for the benchmarked detection methods, (b) within this range, the best performance is achieved by models that leverage multiple successive frames as input, and (c) the proposed single-point approach for B-line localization reaches an F1-score of 0.65, performing on par with the inter-observer agreement. The dataset and developed methods can facilitate further biomedical research on automated interpretation of lung ultrasound with the potential to expand the clinical utility. Copyright © 2023, The Authors. All rights reserved.

关键词： COVID-19