Search

Your search keyword '"McGuinness BE"' showing total 23,158 results
Start Over Author "McGuinness BE"

Refine your results

23,158 results on '"McGuinness BE"'

1. Skill Issues: An Analysis of CS:GO Skill Rating Systems

Author

Bober-Irizar, Mikel, Dua, Naunidh, and McGuinness, Max

Subjects
Computer Science - Artificial Intelligence, Computer Science - Machine Learning
Abstract

The meteoric rise of online games has created a need for accurate skill rating systems for tracking improvement and fair matchmaking. Although many skill rating systems are deployed, with various theoretical foundations, less work has been done at analysing the real-world performance of these algorithms. In this paper, we perform an empirical analysis of Elo, Glicko2 and TrueSkill through the lens of surrogate modelling, where skill ratings influence future matchmaking with a configurable acquisition function. We look both at overall performance and data efficiency, and perform a sensitivity analysis based on a large dataset of Counter-Strike: Global Offensive matches.

Published
2024

2. Leveraging Digital Assets: Lessons from a 14-Year-Old Isotope Tracer Course for Professional Scientists

Author

Eann Malabanan, Owen P. McGuinness, and Kendra H. Oliver

Abstract

The COVID-19 pandemic and subsequent policies (e.g., social distancing, travel restrictions) challenged both organizers for and attendees of programs typically held in person. Many scientific training programs quickly adapted to virtual formats by incorporating digital assets developed for virtual learning and remote social engagement. At the outset, the value of continuing digital elements with future in-person events was unclear. To examine how virtual resources supported heterogeneous professional training programs, we reviewed survey data for a 14-year-old training program for scientific professionals titled "Isotope Tracers in Metabolic Research: Principles and Practice of Kinetic Analysis." We found a positive relationship between survey satisfaction and the postpandemic in-person program that included digital assets held in 2022 compared to prepandemic in-person programs. To better understand the postpandemic program satisfaction, we assessed the 2021 virtual course format and survey data. We found that although there was a desire to return to in-person programs, the digital assets and approaches were valued. In examining the individual programmatic elements in the 2022 in-person course, there was better "value and understandability" of lectures over previous in-person years. These findings highlight how incorporating new digital engagement strategies for professional development benefits even the most established programs in supporting heterogeneous learners.

Published
2024
Full Text
View/download PDF

3. Gaming and Blockchain: Hype and Reality

Author

McGuinness, Max

Subjects
Computer Science - Computers and Society, Computer Science - Distributed, Parallel, and Cluster Computing
Abstract

This paper explores the adoption of blockchain technology in the gaming industry. While supporters affirm that distributed ledger technology has potential to revolutionize gaming economies and provide players with control over their virtual assets, there are practical challenges such as energy consumption and user adoption to be addressed, and detractors question whether blockchain integration is even necessary. This report characterises popular blockchain-based gaming projects like Enjin and Axie Infinity, then compares metrics such as transaction cost and player feedback to evaluate the longevity of blockchain-integrated gaming as a whole.

Published
2024

4. An accurate detection is not all you need to combat label noise in web-noisy datasets

Author

Albert, Paul, Valmadre, Jack, Arazo, Eric, Krishna, Tarun, O'Connor, Noel E., and McGuinness, Kevin

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Training a classifier on web-crawled data demands learning algorithms that are robust to annotation errors and irrelevant examples. This paper builds upon the recent empirical observation that applying unsupervised contrastive learning to noisy, web-crawled datasets yields a feature representation under which the in-distribution (ID) and out-of-distribution (OOD) samples are linearly separable. We show that direct estimation of the separating hyperplane can indeed offer an accurate detection of OOD samples, and yet, surprisingly, this detection does not translate into gains in classification accuracy. Digging deeper into this phenomenon, we discover that the near-perfect detection misses a type of clean examples that are valuable for supervised learning. These examples often represent visually simple images, which are relatively easy to identify as clean examples using standard loss- or distance-based methods despite being poorly separated from the OOD distribution using unsupervised learning. Because we further observe a low correlation with SOTA metrics, this urges us to propose a hybrid solution that alternates between noise detection using linear separation and a state-of-the-art (SOTA) small-loss approach. When combined with the SOTA algorithm PLS, we substantially improve SOTA results for real-world image classification in the presence of web noise github.com/PaulAlbert31/LSA, Comment: Accepted in the European Conference on Computer Vision (ECCV) 2024

Published
2024

5. Convergence, optimization and stability of singular eigenmaps

Author

Akwei, Bernard, Atkins, Bobita, Bailey, Rachel, Dalal, Ashka, Dinin, Natalie, Kerby-White, Jonathan, McGuinness, Tess, Patricks, Tonya, Rogers, Luke, Romanelli, Genevieve, Su, Yiheng, and Teplyaev, Alexander

Subjects
Mathematics - Probability, Mathematics - Metric Geometry, Mathematics - Statistics Theory, 60D05 28A80 62R07 65J20
Abstract

Eigenmaps are important in analysis, geometry, and machine learning, especially in nonlinear dimension reduction. Approximation of the eigenmaps of a Laplace operator depends crucially on the scaling parameter $\epsilon$. If $\epsilon$ is too small or too large, then the approximation is inaccurate or completely breaks down. However, an analytic expression for the optimal $\epsilon$ is out of reach. In our work, we use some explicitly solvable models and Monte Carlo simulations to find the approximately optimal range of $\epsilon$ that gives, on average, relatively accurate approximation of the eigenmaps. Numerically we can consider several model situations where eigen-coordinates can be computed analytically, including intervals with uniform and weighted measures, squares, tori, spheres, and the Sierpinski gasket. In broader terms, we intend to study eigen-coordinates on weighted Riemannian manifolds, possibly with boundary, and on some metric measure spaces, such as fractals.

Published
2024

6. Skew circuits and circumference in a binary matroid

Author

McGuinness, Sean

Subjects
Mathematics - Combinatorics
Abstract

Let C_1 and C_2 be skew circuits in a binary matroid having circumference c. For any positive integer k there is a constant a_k such that if min { |A| ; C_1 \subset A \subset E-A} > a_k, then |C_1| + |C_2| < 2c -k.

Published
2024

7. The mouse metabolic phenotyping center (MMPC) live consortium: an NIH resource for in vivo characterization of mouse models of diabetes and obesity

Author

Laughlin, Maren, McIndoe, Richard, Adams, Sean H, Araiza, Renee, Ayala, Julio E, Kennedy, Lucy, Lanoue, Louise, Lantier, Louise, Macy, James, Malabanan, Eann, McGuinness, Owen P, Perry, Rachel, Port, Daniel, Qi, Nathan, Elias, Carol F, Shulman, Gerald I, Wasserman, David H, and Lloyd, KC Kent

Subjects
Biological Sciences, Genetics, Obesity, Nutrition, Behavioral and Social Science, Diabetes, Metabolic and endocrine, Good Health and Well Being, Metabolism, Mouse, Phenotyping, In vivo, Resource, Service, Tests, Animals, Mice, Diabetes Mellitus, Disease Models, Animal, Phenotype, National Institutes of Health (U.S.), United States, National Institute of Diabetes and Digestive and Kidney Diseases (U.S.), Genetics & Heredity
Abstract

The Mouse Metabolic Phenotyping Center (MMPC)Live Program was established in 2023 by the National Institute for Diabetes, Digestive and Kidney Diseases (NIDDK) at the National Institutes of Health (NIH) to advance biomedical research by providing the scientific community with standardized, high quality phenotyping services for mouse models of diabetes and obesity. Emerging as the next iteration of the MMPC Program which served the biomedical research community for 20 years (2001-2021), MMPCLive is designed as an outwardly-facing consortium of service cores that collaborate to provide reduced-cost consultation and metabolic, physiologic, and behavioral phenotyping tests on live mice for U.S. biomedical researchers. Four MMPCLive Centers located at universities around the country perform complex and often unique procedures in vivo on a fee for service basis, typically on mice shipped from the client or directly from a repository or vendor. Current areas of expertise include energy balance and body composition, insulin action and secretion, whole body carbohydrate and lipid metabolism, cardiovascular and renal function, food intake and behavior, microbiome and xenometabolism, and metabolic pathway kinetics. Additionally, an opportunity arose to reduce barriers to access and expand the diversity of the biomedical research workforce by establishing the VIBRANT Program. Directed at researchers historically underrepresented in the biomedical sciences, VIBRANT-eligible investigators have access to testing services, travel and career development awards, expert advice and experimental design consultation, and short internships to learn test technologies. Data derived from experiments run by the Centers belongs to the researchers submitting mice for testing which can be made publicly available and accessible from the MMPCLive database following publication. In addition to services, MMPCLive staff provide expertise and advice to researchers, develop and refine test protocols, engage in outreach activities, publish scientific and technical papers, and conduct educational workshops and training sessions to aid researchers in unraveling the heterogeneity of diabetes and obesity.

Published
2024

8. Reflecting on the Impact of Hybrid Work on an Academic Library Using the Socio-Ecological Model

Author

Tagge, Natalie, Martin, Lisa, and McGuinness, Susan M

Subjects
Information and Computing Sciences, Library and Information Studies, Infectious Diseases, Emerging Infectious Diseases, Information & Library Sciences
Abstract

Covid-19 led to a dramatic change in the academic library work environment, with many workers shifting to a hybrid work schedule once they returned to in person work. The authors describe the impact of this schedule shift through the novel use of the socio-ecological model. The socio-ecological model looks at the impact of choices through varying circles or levels. The authors discuss the individual, interpersonal, community, and societal levels of impact that the switch to hybrid work had within their context and discuss their role as unit leaders and the decisions made to address these varying impacts.

Published
2024

9. Effects of vitamin D supplementation on a deep learning-based mammographic evaluation in SWOG S0812.

Author

McGuinness, Julia, Anderson, Garnet, Mutasa, Simukayi, Hershman, Dawn, Terry, Mary, Tehranifar, Parisa, Lew, Danika, Yee, Monica, Brown, Eric, Kairouz, Sebastien, Kuwajerwala, Nafisa, Bevers, Therese, Doster, John, Zarwan, Corrine, Kruper, Laura, Minasian, Lori, Ford, Leslie, Arun, Banu, Neuhouser, Marian, Goodman, Gary, Brown, Patrick, Ha, Richard, and Crew, Katherine

Subjects
Humans, Female, Mammography, Deep Learning, Breast Neoplasms, Dietary Supplements, Breast Density, Middle Aged, Cholecalciferol, Adult, Vitamin D, Premenopause, Neural Networks, Computer, Risk Assessment
Abstract

Deep learning-based mammographic evaluations could noninvasively assess response to breast cancer chemoprevention. We evaluated change in a convolutional neural network-based breast cancer risk model applied to mammograms among women enrolled in SWOG S0812, which randomly assigned 208 premenopausal high-risk women to receive oral vitamin D3 20 000 IU weekly or placebo for 12 months. We applied the convolutional neural network model to mammograms collected at baseline (n = 109), 12 months (n = 97), and 24 months (n = 67) and compared changes in convolutional neural network-based risk score between treatment groups. Change in convolutional neural network-based risk score was not statistically significantly different between vitamin D and placebo groups at 12 months (0.005 vs 0.002, P = .875) or at 24 months (0.020 vs 0.001, P = .563). The findings are consistent with the primary analysis of S0812, which did not demonstrate statistically significant changes in mammographic density with vitamin D supplementation compared with placebo. There is an ongoing need to evaluate biomarkers of response to novel breast cancer chemopreventive agents.

Published
2024

10. Spatially resolved single-cell atlas unveils a distinct cellular signature of fatal lung COVID-19 in a Malawian population

Author

Nyirenda, James, Hardy, Olympia M., Silva Filho, João Da, Herder, Vanessa, Attipa, Charalampos, Ndovi, Charles, Siwombo, Memory, Namalima, Takondwa Rex, Suwedi, Leticia, Ilia, Georgios, Nyasulu, Watipenge, Ngulube, Thokozile, Nyirenda, Deborah, Mvaya, Leonard, Phiri, Joseph, Chasweka, Dennis, Eneya, Chisomo, Makwinja, Chikondi, Phiri, Chisomo, Ziwoya, Frank, Tembo, Abel, Makwangwala, Kingsley, Khoswe, Stanley, Banda, Peter, Morton, Ben, Hilton, Orla, Lawrence, Sarah, dos Reis, Monique Freire, Melo, Gisely Cardoso, de Lacerda, Marcus Vinicius Guimaraes, Trindade Maranhão Costa, Fabio, Monteiro, Wuelton Marcelo, Ferreira, Luiz Carlos de Lima, Johnson, Carla, McGuinness, Dagmara, Jambo, Kondwani, Haley, Michael, Kumwenda, Benjamin, Palmarini, Massimo, Denno, Donna M., Voskuijl, Wieger, Kamiza, Steve Bvuobvuo, Barnes, Kayla G., Couper, Kevin, Marti, Matthias, Otto, Thomas D., and Moxon, Christopher A.

Published
2024
Full Text
View/download PDF

13. Progress in the Treatment of Alzheimer’s Disease Is Needed – Position Statement of European Alzheimer’s Disease Consortium (EADC) Investigators

Author

Jessen, Frank, Kramberger, M. G., Angioni, D., Aarsland, D., Balasa, M., Bennys, K., Boada, M., Boban, M., Chincarini, A., Exalto, L., Felbecker, A., Fliessbach, K., Frisoni, G. B., Garza-Martínez, A. J., Grimmer, T., Hanseeuw, B., Hort, J., Ivanoiu, A., Klöppel, S., Krajcovicova, L., McGuinness, B., Mecocci, P., de Mendonca, A., Nous, A., Ousset, P.-J., Paquet, C., Perneczky, R., Peters, O., Tabuas-Pereira, M., Piazza, F., Plantone, D., Riverol, M., Ruiz, A., Sacco, G., Santana, I., Scarmeas, N., Solje, E., Stefanova, E., Sutovsky, S., van der Flier, W., Welsh, T., Wimo, A., Winblad, B., Frölich, L., and Engelborghs, S.

Published
2024
Full Text
View/download PDF

15. Controllable suppression of the unconventional superconductivity in bulk and thin-film Sr$_{2}$RuO$_{4}$ via high-energy electron irradiation

Author

Ruf, Jacob P., Noad, Hilary M. L., Grasset, Romain, Miao, Ludi, Zhakina, Elina, McGuinness, Philippa H., Nair, Hari P., Schreiber, Nathaniel J., Kikugawa, Naoki, Sokolov, Dmitry, Konczykowski, Marcin, Schlom, Darrell G., Shen, Kyle M., and Mackenzie, Andrew P.

Subjects
Condensed Matter - Superconductivity, Condensed Matter - Materials Science
Abstract

In bulk Sr$_{2}$RuO$_{4}$, the strong sensitivity of the superconducting transition temperature $T_{\text{c}}$ to nonmagnetic impurities provides robust evidence for a superconducting order parameter that changes sign around the Fermi surface. In superconducting epitaxial thin-film Sr$_{2}$RuO$_{4}$, the relationship between $T_{\text{c}}$ and the residual resistivity $\rho_0$, which in bulk samples is taken to be a proxy for the low-temperature elastic scattering rate, is far less clear. Using high-energy electron irradiation to controllably introduce point disorder into bulk single-crystal and thin-film Sr$_{2}$RuO$_{4}$, we show that $T_{\text{c}}$ is suppressed in both systems at nearly identical rates. This suggests that part of $\rho_0$ in films comes from defects that do not contribute to superconducting pairbreaking, and establishes a quantitative link between the superconductivity of bulk and thin-film samples.

Published
2024

16. Dataset Clustering for Improved Offline Policy Learning

Author

Wang, Qiang, Deng, Yixin, Sanchez, Francisco Roldan, Wang, Keru, McGuinness, Kevin, O'Connor, Noel, and Redmond, Stephen J.

Subjects
Computer Science - Machine Learning, Computer Science - Robotics
Abstract

Offline policy learning aims to discover decision-making policies from previously-collected datasets without additional online interactions with the environment. As the training dataset is fixed, its quality becomes a crucial determining factor in the performance of the learned policy. This paper studies a dataset characteristic that we refer to as multi-behavior, indicating that the dataset is collected using multiple policies that exhibit distinct behaviors. In contrast, a uni-behavior dataset would be collected solely using one policy. We observed that policies learned from a uni-behavior dataset typically outperform those learned from multi-behavior datasets, despite the uni-behavior dataset having fewer examples and less diversity. Therefore, we propose a behavior-aware deep clustering approach that partitions multi-behavior datasets into several uni-behavior subsets, thereby benefiting downstream policy learning. Our approach is flexible and effective; it can adaptively estimate the number of clusters while demonstrating high clustering accuracy, achieving an average Adjusted Rand Index of 0.987 across various continuous control task datasets. Finally, we present improved policy learning examples using dataset clustering and discuss several potential scenarios where our approach might benefit the offline policy learning community.

Published
2024

22. Video Anomaly Detection via Spatio-Temporal Pseudo-Anomaly Generation : A Unified Approach

Author

Rai, Ayush K., Krishna, Tarun, Hu, Feiyan, Drimbarean, Alexandru, McGuinness, Kevin, Smeaton, Alan F., and O'Connor, Noel E.

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Artificial Intelligence, Computer Science - Machine Learning
Abstract

Video Anomaly Detection (VAD) is an open-set recognition task, which is usually formulated as a one-class classification (OCC) problem, where training data is comprised of videos with normal instances while test data contains both normal and anomalous instances. Recent works have investigated the creation of pseudo-anomalies (PAs) using only the normal data and making strong assumptions about real-world anomalies with regards to abnormality of objects and speed of motion to inject prior information about anomalies in an autoencoder (AE) based reconstruction model during training. This work proposes a novel method for generating generic spatio-temporal PAs by inpainting a masked out region of an image using a pre-trained Latent Diffusion Model and further perturbing the optical flow using mixup to emulate spatio-temporal distortions in the data. In addition, we present a simple unified framework to detect real-world anomalies under the OCC setting by learning three types of anomaly indicators, namely reconstruction quality, temporal irregularity and semantic inconsistency. Extensive experiments on four VAD benchmark datasets namely Ped2, Avenue, ShanghaiTech and UBnormal demonstrate that our method performs on par with other existing state-of-the-art PAs generation and reconstruction based methods under the OCC setting. Our analysis also examines the transferability and generalisation of PAs across these datasets, offering valuable insights by identifying real-world anomalies through PAs., Comment: Accepted in CVPRW 2024 - VAND Workshop

Published
2023

23. Learning and reusing primitive behaviours to improve Hindsight Experience Replay sample efficiency

Author

Sanchez, Francisco Roldan, Wang, Qiang, Bulens, David Cordova, McGuinness, Kevin, Redmond, Stephen, and O'Connor, Noel

Subjects
Computer Science - Robotics, Computer Science - Artificial Intelligence
Abstract

Hindsight Experience Replay (HER) is a technique used in reinforcement learning (RL) that has proven to be very efficient for training off-policy RL-based agents to solve goal-based robotic manipulation tasks using sparse rewards. Even though HER improves the sample efficiency of RL-based agents by learning from mistakes made in past experiences, it does not provide any guidance while exploring the environment. This leads to very large training times due to the volume of experience required to train an agent using this replay strategy. In this paper, we propose a method that uses primitive behaviours that have been previously learned to solve simple tasks in order to guide the agent toward more rewarding actions during exploration while learning other more complex tasks. This guidance, however, is not executed by a manually designed curriculum, but rather using a critic network to decide at each timestep whether or not to use the actions proposed by the previously-learned primitive policies. We evaluate our method by comparing its performance against HER and other more efficient variations of this algorithm in several block manipulation tasks. We demonstrate the agents can learn a successful policy faster when using our proposed method, both in terms of sample efficiency and computation time. Code is available at https://github.com/franroldans/qmp-her., Comment: 6 pages, 2 figures, 1 algorithm, 1 table. Version accepted to ICARA 2024

Published
2023

25. Simvastatin in Critically Ill Patients with Covid-19.

Author

Hills, Thomas, Lorenzi, Elizabeth, Berry, Lindsay, Shyamsundar, Murali, Al-Beidh, Farah, Annane, Djillali, Arabi, Yaseen, Aryal, Diptesh, Au, Carly, Beane, Abigail, Bhimani, Zahra, Bonten, Marc, Bradbury, Charlotte, Brunkhorst, Frank, Burrell, Aidan, Buxton, Meredith, Calfee, Carolyn, Cecconi, Maurizio, Cheng, Allen, Cove, Matthew, Detry, Michelle, Estcourt, Lise, Fitzgerald, Mark, Goligher, Ewan, Goossens, Herman, Green, Cameron, Haniffa, Rashan, Harrison, David, Hashmi, Madiha, Higgins, Alisa, Huang, David, Ichihara, Nao, Jayakumar, Deva, Kruger, Peter, Lamontagne, François, Lampro, Lamprini, Lawler, Patrick, Marshall, John, Mason, Alexina, McGlothlin, Anna, McGuinness, Shay, McQuilten, Zoe, McVerry, Bryan, Mouncey, Paul, Murthy, Srinivas, Neal, Matthew, Nichol, Alistair, OKane, Cecilia, Parke, Rachael, Parker, Jane, Rabindrarajan, Ebenezer, Reyes, Luis, Rowan, Kathryn, Saito, Hiroki, Santos, Marlene, Saunders, Christina, Seymour, Christopher, Shankar-Hari, Manu, Sinha, Pratik, Thompson, B, Turgeon, Alexis, Turner, Anne, van de Veerdonk, Frank, Weis, Sebastian, Young, Ian, Zarychanski, Ryan, McArthur, Colin, Angus, Derek, Berry, Scott, Derde, Lennie, Webb, Steve, Gordon, Anthony, McAuley, Daniel, and Lewis, Roger

Subjects
Humans, Bayes Theorem, COVID-19, COVID-19 Drug Treatment, Critical Illness, Hospital Mortality, Hydroxymethylglutaryl-CoA Reductase Inhibitors, Simvastatin, Treatment Outcome
Abstract

BACKGROUND: The efficacy of simvastatin in critically ill patients with coronavirus disease 2019 (Covid-19) is unclear. METHODS: In an ongoing international, multifactorial, adaptive platform, randomized, controlled trial, we evaluated simvastatin (80 mg daily) as compared with no statin (control) in critically ill patients with Covid-19 who were not receiving statins at baseline. The primary outcome was respiratory and cardiovascular organ support-free days, assessed on an ordinal scale combining in-hospital death (assigned a value of -1) and days free of organ support through day 21 in survivors; the analyis used a Bayesian hierarchical ordinal model. The adaptive design included prespecified statistical stopping criteria for superiority (>99% posterior probability that the odds ratio was >1) and futility (>95% posterior probability that the odds ratio was

Published
2023

26. Northern Ireland Pupils Transcend Cultural Difference through Transformed Integrated Schools: 'We Don't Think about Religion When We're Passing the Ball, We Just Do It'

Author

Lesley Abbott and Samuel McGuinness

Abstract

Schooling for Northern Ireland children has over decades been in denominationally separate schools, until an integrated system was instigated by concerned parents in the late 1970s amidst growing political violence. By educating together Catholic and Protestant pupils and those of other religions or none, the hope was to contribute to peace in a conflict-prone society. The first integrated schools were planned and government funded. Then, in the early 1990s, some segregated schools sought to transform to integrated status through a detailed, formal process. In individual, face-to-face interviews, 11- and 16-year old pupils (n = 20), representing both school sectors and both the main cultural backgrounds, described their experiences of learning side by side for the first time with peers of a different tradition. Both younger and older pupils could discern the differences between their previous (segregated) and current (integrated) settings, saw the benefits of accepting and acceptance, and perceived sameness as well as difference. They could appreciate the global perspective, understood the purpose of outward, visible changes such as emblems and school displays, and welcomed curriculum change, particularly in Physical Education. Pupils understood the value of having difficult conversations without acrimony at the same time as learning about and respecting others' viewpoints.

Published
2024
Full Text
View/download PDF

27. Multi-site Integrated Optical Addressing of Trapped Ions

Author

Kwon, Joonhyuk, Setzer, William J., Gehl, Michael, Karl, Nicholas, Van Der Wall, Jay, Law, Ryan, Blain, Matthew G., Stick, Daniel, and McGuinness, Hayden J.

Subjects
Quantum Physics
Abstract

One of the most effective ways to advance the performance of quantum computers and quantum sensors is to increase the number of qubits or quantum resources in the system. A major technical challenge that must be solved to realize this goal for trapped-ion systems is scaling the delivery of optical signals to many individual ions. In this paper we demonstrate an approach employing waveguides and multi-mode interferometer splitters to optically address multiple $^{171}\textrm{Yb}^+$ ions in a surface trap by delivering all wavelengths required for full qubit control. Measurements of hyperfine spectra and Rabi flopping were performed on the E2 clock transition, using integrated waveguides for delivering the light needed for Doppler cooling, state preparation, coherent operations, and detection. We describe the use of splitters to address multiple ions using a single optical input per wavelength and use them to demonstrate simultaneous Rabi flopping on two different transitions occurring at distinct trap sites. This work represents an important step towards the realization of scalable integrated photonics for atomic clocks and trapped-ion quantum information systems., Comment: 10 pages, 4 figures (+3 supplementary figures)

Published
2023
Full Text
View/download PDF

28. Cyclic Orderings of Paving Matroids

Author

McGuinness, Sean

Subjects
Mathematics - Combinatorics
Abstract

A matroid M is cyclically orderable if there is a cyclic permutation of the elements of M such that any r consecutive elements form a basis in M. An old conjecture of Kajitani, Miyano, and Ueno states that a matroid M is cyclically orderable if and only if for all nonempty subsets X in E(M), |X|/r(M) is less than or equal to |E(M)|/r(M). In this paper, we verify this conjecture for all paving matroids.

Published
2023

29. Seeing the Fruit for the Leaves: Robotically Mapping Apple Fruitlets in a Commercial Orchard

Author

Qureshi, Ans, Smith, David, Gee, Trevor, Nejati, Mahla, Shahabi, Jalil, Lim, JongYoon, Ahn, Ho Seok, McGuinness, Ben, Downes, Catherine, Jangali, Rahul, Black, Kale, Lim, Hin, Duke, Mike, MacDonald, Bruce, and Williams, Henry

Subjects
Computer Science - Robotics
Abstract

Aotearoa New Zealand has a strong and growing apple industry but struggles to access workers to complete skilled, seasonal tasks such as thinning. To ensure effective thinning and make informed decisions on a per-tree basis, it is crucial to accurately measure the crop load of individual apple trees. However, this task poses challenges due to the dense foliage that hides the fruitlets within the tree structure. In this paper, we introduce the vision system of an automated apple fruitlet thinning robot, developed to tackle the labor shortage issue. This paper presents the initial design, implementation,and evaluation specifics of the system. The platform straddles the 3.4 m tall 2D apple canopy structures to create an accurate map of the fruitlets on each tree. We show that this platform can measure the fruitlet load on an apple tree by scanning through both sides of the branch. The requirement of an overarching platform was justified since two-sided scans had a higher counting accuracy of 81.17 % than one-sided scans at 73.7 %. The system was also demonstrated to produce size estimates within 5.9% RMSE of their true size., Comment: Accepted at the International Conference on Intelligent Robots and Systems (IROS 2023)

Published
2023

31. Learning-Rate-Free Learning: Dissecting D-Adaptation and Probabilistic Line Search

Author

McGuinness, Max

Subjects
Computer Science - Machine Learning, Mathematics - Optimization and Control
Abstract

This paper explores two recent methods for learning rate optimisation in stochastic gradient descent: D-Adaptation (arXiv:2301.07733) and probabilistic line search (arXiv:1502.02846). These approaches aim to alleviate the burden of selecting an initial learning rate by incorporating distance metrics and Gaussian process posterior estimates, respectively. In this report, I provide an intuitive overview of both methods, discuss their shared design goals, and devise scope for merging the two algorithms.

Published
2023

32. Self-Supervised and Semi-Supervised Polyp Segmentation using Synthetic Data

Author

Moreu, Enric, Arazo, Eric, McGuinness, Kevin, and O'Connor, Noel E.

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Early detection of colorectal polyps is of utmost importance for their treatment and for colorectal cancer prevention. Computer vision techniques have the potential to aid professionals in the diagnosis stage, where colonoscopies are manually carried out to examine the entirety of the patient's colon. The main challenge in medical imaging is the lack of data, and a further challenge specific to polyp segmentation approaches is the difficulty of manually labeling the available data: the annotation process for segmentation tasks is very time-consuming. While most recent approaches address the data availability challenge with sophisticated techniques to better exploit the available labeled data, few of them explore the self-supervised or semi-supervised paradigm, where the amount of labeling required is greatly reduced. To address both challenges, we leverage synthetic data and propose an end-to-end model for polyp segmentation that integrates real and synthetic data to artificially increase the size of the datasets and aid the training when unlabeled samples are available. Concretely, our model, Pl-CUT-Seg, transforms synthetic images with an image-to-image translation module and combines the resulting images with real images to train a segmentation model, where we use model predictions as pseudo-labels to better leverage unlabeled samples. Additionally, we propose PL-CUT-Seg+, an improved version of the model that incorporates targeted regularization to address the domain gap between real and synthetic images. The models are evaluated on standard benchmarks for polyp segmentation and reach state-of-the-art results in the self- and semi-supervised setups.

Published
2023

33. Enhancing CLIP with GPT-4: Harnessing Visual Descriptions as Prompts

Author

Maniparambil, Mayug, Vorster, Chris, Molloy, Derek, Murphy, Noel, McGuinness, Kevin, and O'Connor, Noel E.

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Artificial Intelligence, Computer Science - Computation and Language, Computer Science - Machine Learning
Abstract

Contrastive pretrained large Vision-Language Models (VLMs) like CLIP have revolutionized visual representation learning by providing good performance on downstream datasets. VLMs are 0-shot adapted to a downstream dataset by designing prompts that are relevant to the dataset. Such prompt engineering makes use of domain expertise and a validation dataset. Meanwhile, recent developments in generative pretrained models like GPT-4 mean they can be used as advanced internet search tools. They can also be manipulated to provide visual information in any structure. In this work, we show that GPT-4 can be used to generate text that is visually descriptive and how this can be used to adapt CLIP to downstream tasks. We show considerable improvements in 0-shot transfer accuracy on specialized fine-grained datasets like EuroSAT (~7%), DTD (~7%), SUN397 (~4.6%), and CUB (~3.3%) when compared to CLIP's default prompt. We also design a simple few-shot adapter that learns to choose the best possible sentences to construct generalizable classifiers that outperform the recently proposed CoCoOP by ~2% on average and by over 4% on 4 specialized fine-grained datasets. The code, prompts, and auxiliary text dataset is available at https://github.com/mayug/VDT-Adapter., Comment: Paper accepted at ICCV-W 2023. V2 contains additional comparisons with concurrent works

Published
2023

34. Joint one-sided synthetic unpaired image translation and segmentation for colorectal cancer prevention

Author

Moreu, Enric, Arazo, Eric, McGuinness, Kevin, and O'Connor, Noel E.

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Artificial Intelligence
Abstract

Deep learning has shown excellent performance in analysing medical images. However, datasets are difficult to obtain due privacy issues, standardization problems, and lack of annotations. We address these problems by producing realistic synthetic images using a combination of 3D technologies and generative adversarial networks. We propose CUT-seg, a joint training where a segmentation model and a generative model are jointly trained to produce realistic images while learning to segment polyps. We take advantage of recent one-sided translation models because they use significantly less memory, allowing us to add a segmentation model in the training loop. CUT-seg performs better, is computationally less expensive, and requires less real images than other memory-intensive image translation approaches that require two stage training. Promising results are achieved on five real polyp segmentation datasets using only one real image and zero real annotations. As a part of this study we release Synth-Colon, an entirely synthetic dataset that includes 20000 realistic colon images and additional details about depth and 3D geometry: https://enric1994.github.io/synth-colon, Comment: arXiv admin note: substantial text overlap with arXiv:2202.08680

Published
2023
Full Text
View/download PDF

35. Quantum squeezing cannot beat the standard quantum limit

Author

McGuinness, Liam P.

Subjects
Quantum Physics
Abstract

Quantum entanglement between particles is expected to allow one to perform tasks that would otherwise be impossible. In quantum sensing and metrology, entanglement is often claimed to enable a precision that cannot be attained with the same number of particles and time, forgoing entanglement. Two distinct approaches exist: creation of entangled states that either i) respond quicker to the signal, or ii) are associated with lower noise and uncertainty. The second class of states are generally called squeezed states. Here we show that if our definition of success is -- a precision that is impossible to achieve using the same resources but without entanglement -- then the second approach cannot succeed. In doing so we show that a single non-separable squeezed state provides fundamentally no better precision, per unit time, than a single particle.

Published
2023

36. Impact of Hemoglobin Levels on Composite Cardiac Arrest or Stroke Outcome in Patients With Respiratory Failure Due to COVID-19

Author

Shi Nan Feng, BSPH, Thu-Lan Kelly, PhD, John F. Fraser, MD, PhD, Gianluigi Li Bassi, MD, PhD, Jacky Suen, PhD, Akram Zaaqoq, MD, MPH, Matthew J. Griffee, MD, Rakesh C. Arora, MD, Nicole White, PhD, Glenn Whitman, MD, Chiara Robba, MD, PhD, Denise Battaglini, MD, PhD, Sung-Min Cho, DO, MHS, on behalf of COVID-19 Critical Care Consortium (CCCC), Robert Bartlett, John F. Fraser, Gianluigi Li Bassi, Jacky Y. Suen, Heidi J. Dalton, John Laffey, Daniel Brodie, Eddy Fan, Antoni Torres, Davide Chiumello, Alyaa Elhazm, Carol Hodgson, Shingo Ichiba, Carlos Luna, Srinivas Murthy, Alistair Nichol, Pauline Yeung Ng, Mark Ogino, Aidan Burrell, Antonio Pesenti, Tala Al-Dabbous, Huda Alfoudri, Mohammed Shamsah, Subbarao Elapavaluru, Ashley Berg, Christina Horn, Yunis Mayasi, Stephan Schroll, Dan Meyer, Jorge Velazco, Ludmyla Ploskanych, Wanda Fikes, Rohini Bagewadi, Marvin Dao, Haley White, Alondra Berrios Laviena, Ashley Ehlers Maysoon, Shalabi-McGuire, Trent Witt, Lorenzo Grazioli, Luca Lorini, E. Wilson Grandin, Jose Nunez, Tiago Reyes, Diarmuid O’Briain, Stephanie Hunter, Mahesh Ramanan, Julia Affleck, Hemanth Hurkadli Veerendra, Sumeet Rai, Josie Russell-Brown, Mary Nourse, Mark Joseph, Brook Mitchell, Martha Tenzer, Ryuzo Abe, Hwa Jin Cho, In Seok Jeong, Nadeem Rahman, Vivek Kakar, Andres Oswaldo Razo Vazquez, Nicolas Brozzi, Omar Mehkri, Sudhir Krishnan Abhijit, Duggal Stuart Houltham, Jerónimo Graf, Roderigo Diaz, Roderigo Orrego, Camila Delgado, Joyce González, Maria Soledad Sanchez, Michael Piagnerelli, Josefa Valenzuela Sarrazin, A/Prof. Gustavo Zabert, Lucio Espinosa, Paulo Delgado, Victoria Delgado, Diego Fernando, Bautista Rincón, Angela Maria Marulanda Yanten, Melissa Bustamante Duque, Alyaa Elhazmi, Abdullah Al-Hudaib, Maria Callahan, M. Azhari Taufik, Elizabeth Yasmin Wardoyo, Margaretha Gunawan, Nurindah S Trisnaningrum, Vera Irawany, Muhammad Rayhan, Mauro Panigada, Alberto Zanella, Giacomo Grasselli, Sebastiano Colombo, Chiara Martinet, Gaetano Florio, Massimo Antonelli, Simone Carelli, Domenico L. Grieco, Motohiro Asaki, Kota Hoshino, Leonardo Salazar, Mary Alejandra Mendoza Monsalve, Bairbre McNicholas, David Cosgrave, Joseph McCaffrey, Allison Bone, Yusuff Hakeem, James Winearls, Mandy Tallott, David Thomson, Christel Arnold-Day, Jerome Cupido, Zainap Fanie, Malcom Miller, Lisa Seymore, Dawid van Straaten, Ali Ait Hssain, Jeffrey Aliudin, Al-Reem Alqahtani, Khoulod Mohamed, Ahmed Mohamed, Darwin Tan, Joy Villanueva, Ahmed Zaqout, Ethan Kurtzman, Arben Ademi, Ana Dobrita, Khadija El Aoudi, Juliet Segura, Gezy Giwangkancana, Shinichiro Ohshimo, Javier Osatnik, Anne Joosten, Minlan Yang, Ana Motos, Francisco Arancibia, Virginie Williams, Alexandre Noel, Nestor Luque, Marina Fantini, Ruth Noemi Jorge García, Enrique Chicote Alvarez, Anna Greti, Adrian Ceccato, Angel Sanchez, Ana Loza Vazquez, Ferran Roche-Campo, Diego Franch-Llasat, Divina Tuazon, Marcelo Amato, Luciana Cassimiro, Flavio Pola, Francis Ribeiro, Guilherme Fonseca, Heidi Dalton, Mehul Desai, Erik Osborn Hala Deeb, Antonio Arcadipane, Gennaro Martucci, Giovanna Panarello, Chiara Vitiello, Claudia Bianco, Giovanna Occhipinti, Matteo Rossetti, Raffaele Cuffaro, Sung-Min Cho, Glenn Whitman, Hiroaki Shimizu, Naoki Moriyama, Jae-Burm Kim, Nobuya Kitamura, Johannes Gebauer, Toshiki Yokoyama, Abdulrahman Al-Fares, Sarah Buabbas, Esam Alamad, Fatma Alawadhi, Kalthoum Alawadi, Hiro Tanaka, Satoru Hashimoto, Masaki Yamazaki, Tak-Hyuck Oh, Mark Epler, Cathleen Forney, Louise Kruse, Jared Feister, Joelle Williamson, Katherine Grobengieser, Eric Gnall, Sasha Golden, Mara Caroline, Timothy Shapiro, Colleen Karaj, Lisa Thome, Lynn Sher, Mark Vanderland, Mary Welch, Sherry McDermott, Matthew Brain, Sarah Mineall, Dai Kimura, Luca Brazzi, Gabriele Sales, Giorgia Montrucchio, Tawnya Ogston, Dave Nagpal, Karlee Fischer, Roberto Lorusso, Rajavardhan Rangappa, Sujin Rai, Argin Appu, Mariano Esperatti, Nora Angélica Fuentes, Maria Eugenia Gonzalez, Edmund G. Carton, Ayan Sen, Amanda Palacios, Deborah Rainey, Gordan Samoukoviv, Josie Campisi, Lucia Durham, Emily Neumann, Cassandra Seefeldt, Octavio Falcucci, Amanda Emmrich, Jennifer Guy, Carling Johns, Kelly Potzner, Catherine Zimmermann, Angelia Espinal, Nina Buchtele, Michael Schwameis, Andrea Korhnfehl, Roman Brock, Thomas Staudinger, Stephanie-Susanne, Stecher Michaela Barnikel, Sófia Antón, Alexandra Pawlikowski, Akram Zaaqoq, Lan Anh Galloway, Caitlin Merley, Marc Csete, Luisa Quesada, Isabela Saba, Daisuke Kasugai, Hiroaki Hiraiwa, Taku Tanaka, Eva Marwali, Yoel Purnama, Santi Rahayu Dewayanti, Ardiyan, Dafsah Arifa Juzar, Debby Siagian, Yih-Sharng Chen, Indrek Ratsep, Andra-Maris Post, Piret Sillaots, Anneli Krund, Merili-Helen Lehiste, Tanel Lepik, Frank Manetta, Effe Mihelis, Iam Claire Sarmiento, Mangala Narasimhan, Michael Varrone, Mamoru Komats, Julia Garcia-Diaz, Catherine Harmon, S. Veena Satyapriya, Amar Bhatt, Nahush A. Mokadam, Alberto Uribe, Alicia Gonzalez, Haixia Shi, Johnny McKeown, Joshua Pasek, Juan Fiorda, Marco Echeverria, Rita Moreno, Bishoy Zakhary, Marco Cavana, Alberto Cucino, Giuseppe Foti, Marco Giani, Benedetta Fumagalli, Valentina Castagna, Andrea Dell’Amore, Paolo Navalesi, Hoi-Ping Shum, Alain Vuysteke, Asad Usman, Andrew Acker, Benjamin Smood, Blake Mergler, Federico Sertic, Madhu Subramanian, Alexandra Sperry, Nicolas Rizer, Erlina Burhan, Menaldi Rasmin, Ernita Akmal, Faya Sitompul, Navy Lolong, Bhat Naivedh, Simon Erickson, Peter Barrett, David Dean, Julia Daugherty, Antonio Loforte, Irfan Khan, Mohammed Abraar Quraishi, Olivia DeSantis, Dominic So, Darshana Kandamby, Jose M. Mandei, Hans Natanael, Eka YudhaLantang, Anastasia Lantang, Surya Oto Wijaya, Anna Jung, George Ng, Wing Yiu Ng, Shu Fang, Alexis Tabah, Megan Ratcliffe, Maree Duroux, Shingo Adachi, Shota Nakao, Pablo Blanco, Ana Prieto, Jesús Sánchez, Meghan Nicholson, Warwick Butt, Alyssa Serratore, Carmel Delzoppo, Pierre Janin, Elizabeth Yarad, Richard Totaro, Jennifer Coles, Bambang Pujo, Robert Balk, Andy Vissing, Esha Kapania, James Hays, Samuel Fox, Garrett Yantosh, Pavel Mishin, Saptadi Yuliarto, Kohar Hari Santoso, Susanthy Djajalaksana, Arie Zainul Fatoni, Masahiro Fukuda, Keibun Liu, Paolo Pelosi, Denise Battaglini, Juan Fernando Masa Jiménez, Diego Bastos, Sérgio Gaião, Desy Rusmawatiningtyas, Young-Jae Cho, Su Hwan Lee, Tatsuya Kawasaki, Laveena Munshi, Pranya Sakiyalak, Prompak Nitayavardhana, Tamara Seitz, Rakesh Arora, David Kent, Daniel Marino, Swapnil Parwar, Andrew Cheng, Jennene Miller, Shigeki Fujitani, Naoki Shimizu, Jai Madhok, Clark Owyang, Hergen Buscher, Claire Reynolds, Olavi Maasikas, Aleksan Beljantsev, Vladislav Mihnovits, Takako Akimoto, Mariko Aizawa, Kanako Horibe, Ryota Onodera, Meredith Young, Timothy George, Kiran Shekar, Niki McGuinness, Lacey Irvine, Brigid Flynn, Tomoyuki Endo, Kazuhiro Sugiyama, Keiki Shimizu, Kathleen Exconde, Leslie Lussier, Gösta Lotz, Maximilian Malfertheiner, Lars Maier, Esther Dreier, Neurinda Permata Kusumastuti, Colin McCloskey, Al-Awwab Dabaliz, Tarek B Elshazly, Josiah Smith, Konstanty S. Szuldrzynski, Piotr Bielański, Keith Wille, Ken Kuljit, S. Parhar, Kirsten M. Fiest, Cassidy Codan, Anmol Shahid, Mohamed Fayed, Timothy Evans, Rebekah Garcia, Ashley Gutierrez, Tae Song, Rebecca Rose, Suzanne Bennett, Denise Richardson, Giles Peek, Lovkesh Arora, Kristina Rappapport, Kristina Rudolph, Zita Sibenaller, Lori Stout, Alicia Walter, Daniel Herr, Nazli Vedadi, Shaun Thompson, Julie Hoffman, Xiaonan Ying, Ryan Kennedy, Muhammed Elhadi, Matthew Griffee, Anna Ciullo, Yuri Kida, Ricard Ferrer Roca, JordI Riera, Sofia Contreras, Cynthia Alegre, Christy Kay, Irene Fischer, Elizabeth Renner, Hayato Taniguci, John Fraser, Jacky Suen, Adrian Barnett, Nicole White, Kristen Gibbons, Simon Forsyth, Amanda Corley, India Pearse, Samuel Hinton, Gabriella Abbate, Halah Hassan, Silver Heinsar, Varun A Karnik, Katrina Ki, Hollier F. O’Neill, Nchafatso Obonyo, Leticia Pretti Pimenta, Janice D. Reid, Kei Sato, Aapeli Vuorinen, Karin S. Wildi, Emily S. Wilson, Stephanie Yerkovich, James Lee, Daniel Plotkin, Barbara Wanjiru Citarella, Laura Merson, Emma Hartley, Bastian Lubis, Takanari Ikeyama, Balu Bhaskar, Jae-Seung Jung, Shay McGuinness, Glenn Eastwood, Sandra Rossi Marta, Fabio Guarracino, Stacy Gerle, Emily Coxon, Bruno Claro, Daniel Loverde, Namrata Patil, Vieri Parrini, Angela McBride, Kathryn Negaard, Angela Ratsch, Ahmad Abdelaziz, Juan David Uribe, Adriano Peris, Mark Sanders, Dominic Emerson, Muhammad Kamal, Pedro Povoa, Roland Francis, Ali Cherif, Sunimol Joseph, Matteo Di Nardo, Micheal Heard, Kimberly Kyle, Ray A Blackwell, Patrick Biston, Hye Won Jeong, Reanna Smith, Yogi Prawira, Arturo Huerta Garcia, Nahikari Salterain, Bart Meyns, Marsha Moreno, Rajat Walia, Amit Mehta, Annette Schweda, Moh Supriatna, Cenk Kirakli, Melissa Williams, Kyung Hoon Kim, Alexandra Assad, Estefania Giraldo, Wojtek Karolak, Martin Balik, Elizabeth Pocock, Evan Gajkowski, Kanamoto Masafumi, Nicholas Barrett, Yoshihiro Takeyama, Sunghoon Park, Faizan Amin, Fina Meilyana Andriyani, Serhii Sudakevych, Magdalena Vera, Rodrigo Cornejo, Patrícia Schwarz, Ana Carolina Mardini, Thais de Paula, Ary Serpa Neto, Andrea Villoldo, Alexandre Siciliano Colafranceschi, Alejandro Ubeda Iglesias, Juan Granjean, Lívia Maria Garcia Melro, Giovana Fioravante Romualdo, Diego Gaia, Helmgton Souza, Filomena Galas, Rafael Máñez Mendiluce, Alejandra Sosa, Ignacio Martinez, Hiroshi Kurosawa, Juan Salgado, Beate Hugi-Mayr, Eric Charbonneau, Vitor Salvatore Barzilai, Veronica Monteiro, Rodrigo Ribeiro de Souza, Michael Harper, Hiroyuki Suzuki, Celina Adams, Jorge Brieva, George Nyale, Faisal Saleem Eltatar, Jihan Fatani, Husam Baeissa, Ayman AL Masri, Ahmed Rabie, Mok Yee Hui, Masahiro Yamane, Hanna Jung, Ayorinde Mojisola Margaret, Newell Nacpil, Katja Ruck, Rhonda Bakken, Claire Jara, Tim Felton, Lorenzo Berra, Bobby Shah, Arpan Chakraborty, Monika Cardona, Gerry Capatos, Bindu Akkanti, Abiodun Orija, Harsh Jain, Asami Ito, Brahim Housni, Sennen Low, Koji Iihara, Joselito Chavez, Kollengode Ramanathan, Gustavo Zabert, Krubin Naidoo, Ian Seppelt, Marlice VanDyk, Sarah MacDonald, Randy McGregor, Teka Siebenaler, Hannah Flynn, Kristi Lofton, Toshiyuki Aokage, Kazuaki Shigemitsu, Andrea Moscatelli, Giuseppe Fiorentino, Matthias Baumgaertel, Serge Eddy Mba, Jana Assy, Amelya Hutahaean, Holly Roush, Kay A Sichting, Francesco Alessandri, Debra Burns, Gavin Salt, Carl P. Garabedian, Jonathan Millar, Malcolm Sim, Adrian Mattke, Danny McAuley, Jawad Tadili, Tim Frenzel, Yaron Bar-Lavie, Aaron Blandino Ortiz, Jackie Stone, Antony Attokaran, Michael Farquharson, Brij Patel, Derek Gunning, Kenneth Baillie, Pia Watson, Kenji Tamai, Gede Ketut Sajinadiyasa, Dyah Kanyawati, Marcello Salgado, Assad Sassine, Bhirowo Yudo, Scott McCaul, Bongjin Lee, Sang Min Lee, Arnon Afek, Yoshiaki Iwashita, Bambang Pujo Semedi, Jack Metiva, Nicole Van Belle, Ignacio Martin-Loeches, Lenny Ivatt, Chia Yew Woon, Hyun Mi Kang, Timothy Smith, Erskine James, Nawar Al-Rawas, Yudai Iwasaki, Kenny Chan King-Chung, Vadim Gudzenko, Fabio Taccone, Fajar Perdhana, Yoan Lamarche, Joao Miguel Ribeiro, Nikola Bradic, Klaartje Van den Bossche, Oude Lansink, Gurmeet Singh, Gerdy Debeuckelaere, Henry T. Stelfox, Cassia Yi, Jennifer Elia, Thomas Tribble, Shyam Shankar, Raj Padmanabhan, Bill Hallinan, Luca Paoletti, Yolanda Leyva, Tatuma Fykuda, Jenelle Badulak, Jillian Koch, Amy Hackman, Lisa Janowaik, Deb Hernandez, Jennifer Osofsky, Katia Donadello, Aizah Lawang, Josh Fine, and Benjamin Davidson

Subjects
Medical emergencies. Critical care. Intensive care. First aid, RC86-88.9
Abstract

OBJECTIVES:. Anemia has been associated with an increased risk of both cardiac arrest and stroke, frequent complications of COVID-19. The effect of hemoglobin level at ICU admission on a composite outcome of cardiac arrest or stroke in an international cohort of COVID-19 patients was investigated. DESIGN:. Retrospective analysis of prospectively collected database. SETTING:. A registry of COVID-19 patients admitted to ICUs at over 370 international sites was reviewed for patients diagnosed with cardiac arrest or stroke up to 30 days after ICU admission. Anemia was defined as: normal (hemoglobin ≥ 12.0 g/dL for women, ≥ 13.5 g/dL for men), mild (hemoglobin 10.0–11.9 g/dL for women, 10.0–13.4 g/dL for men), moderate (hemoglobin ≥ 8.0 and < 10.0 g/dL for women and men), and severe (hemoglobin < 8.0 g/dL for women and men). PATIENTS:. Patients older than 18 years with acute COVID-19 infection in the ICU. INTERVENTIONS:. None. MEASUREMENTS AND MAIN RESULTS:. Of 6926 patients (median age = 59 yr, male = 65%), 760 patients (11.0%) experienced stroke (2.0%) and/or cardiac arrest (9.4%). Cardiac arrest or stroke was more common in patients with low hemoglobin, occurring in 12.8% of patients with normal hemoglobin, 13.3% of patients with mild anemia, and 16.7% of patients with moderate/severe anemia. Time to stroke or cardiac arrest by anemia status was analyzed using Cox proportional hazards regression with death as a competing risk. Covariates selected through clinical knowledge were age, sex, comorbidities (diabetes, hypertension, obesity, and cardiac or neurologic conditions), pandemic era, country income, mechanical ventilation, and extracorporeal membrane oxygenation. Moderate/severe anemia was associated with a higher risk of cardiac arrest or stroke (hazard ratio, 1.32; 95% CI, 1.05–1.67). CONCLUSIONS:. In an international registry of ICU patients with COVID-19, moderate/severe anemia was associated with increased hazard of cardiac arrest or stroke.

Published
2024
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results