10784 research outputs found
Sort by
Assessing the impact of vaccines on COVID-19 efficacy in survival rates: a survival analysis approach for clinical decision support
No full textPublisher Copyright: Copyright © 2024 González Rodríguez, Oprescu, Muñoz Lezcano, Cordero Ramos, Romero Cabrera, Armengol de la Hoz and Estella.Background: The global COVID-19 pandemic, caused by the SARS-CoV-2 virus, has presented significant challenges to healthcare systems worldwide. Objective: This study, based on an analysis of a cohort from the Public Health System of Andalusia (Spain), aims to evaluate how vaccination affects case-fatality rate in patients hospitalized due to COVID-19 infection in Andalusia. Methods: The cohort consists of 37,274 individuals after applying the inclusion criteria. We conducted survival analyses employing the Cox proportional hazards models and generated adjusted survival curves to examine the outcomes. The analyses were performed from three perspectives: vaccinated vs. unvaccinated patients, vaccinated and unvaccinated patients grouped by age, and stratified by vaccination status. Results: Results indicate a substantial correlation between vaccination and a 20% reduction in the risk of case-fatality. Age-specific effects reveal varying degrees of protection across different age groups. Conclusion: These findings emphasize the pivotal role of vaccination status in COVID-19 risk assessment, supporting the development of a clinical decision support system for accurate predictions and optimizing healthcare management at admission.Peer reviewe
Large-Kernel Attention for 3D Medical Image Segmentation
Get PDFPublisher Copyright: © The Author(s) 2023.Automated segmentation of multiple organs and tumors from 3D medical images such as magnetic resonance imaging (MRI) and computed tomography (CT) scans using deep learning methods can aid in diagnosing and treating cancer. However, organs often overlap and are complexly connected, characterized by extensive anatomical variation and low contrast. In addition, the diversity of tumor shape, location, and appearance, coupled with the dominance of background voxels, makes accurate 3D medical image segmentation difficult. In this paper, a novel 3D large-kernel (LK) attention module is proposed to address these problems to achieve accurate multi-organ segmentation and tumor segmentation. The advantages of biologically inspired self-attention and convolution are combined in the proposed LK attention module, including local contextual information, long-range dependencies, and channel adaptation. The module also decomposes the LK convolution to optimize the computational cost and can be easily incorporated into CNNs such as U-Net. Comprehensive ablation experiments demonstrated the feasibility of convolutional decomposition and explored the most efficient and effective network design. Among them, the best Mid-type 3D LK attention-based U-Net network was evaluated on CT-ORG and BraTS 2020 datasets, achieving state-of-the-art segmentation performance when compared to avant-garde CNN and Transformer-based methods for medical image segmentation. The performance improvement due to the proposed 3D LK attention module was statistically validated.This study was supported in part by the BHF (TG/18/5/34111, PG/16/78/32402), the ERC IMI (101005122), the H2020 (952172), the MRC (MC/PC/21013), the Royal Society (IEC/NSFC/211235), the Imperial College Undergraduate Research Opportunities Programme (UROP), the NVIDIA Academic Hardware Grant Program, the SABER project supported by Boehringer Ingelheim Ltd, NIHR Imperial Biomedical Research Centre (RDA01), and the UKRI Future Leaders Fellowship (MR/V023799/1). J. Del Ser also received funding support from the Department of Education of the Basque Government (Consolidated Research Group MATHMODE, IT1456-22).Peer reviewe
Estimation of flea beetle damage in the field using a multistage deep learning-based solution
Get PDFPublisher Copyright: © 2023Estimation of damage in plants is a key issue for crop protection. Currently, experts in the field manually assess the plots. This is a time-consuming task that can be automated thanks to the latest technology in computer vision (CV). The use of image-based systems and recently deep learning-based systems have provided good results in several agricultural applications. These image-based applications outperform expert evaluation in controlled environments, and now they are being progressively included in non-controlled field applications. A novel solution based on deep learning techniques in combination with image processing methods is proposed to tackle the estimate of plant damage in the field. The proposed solution is a two-stage algorithm. In a first stage, the single plants in the plots are detected by an object detection YOLO based model. Then a regression model is applied to estimate the damage of each individual plant. The solution has been developed and validated in oilseed rape plants to estimate the damage caused by flea beetle. The crop detection model achieves a mean precision average of 91% with a [email protected] of 0.99 and a [email protected] of 0.91 for oilseed rape specifically. The regression model to estimate up to 60% of damage degree in single plants achieves a MAE of 7.11, and R2 of 0.46 in comparison with manual evaluations done plant by plant by experts. Models are deployed in a docker, and with a REST API communication protocol they can be inferred directly for images acquired in the field from a mobile device.Peer reviewe
SpeCluRC-NTL: Spearman's distance-based clustering Reservoir Computing solution for NTL detection in smart grids
Get PDFPublisher Copyright: © 2024 The Author(s)Smart grids are ushering in a transformative era for energy distribution and consumption, yet their emergence also brings forth novel security and fraud detection challenges. The intricacy of detecting fraud within smart grids demands sophisticated techniques for scrutinizing vast volumes of time series data. This work introduces a novel approach that integrates time series aggregation functions, time series clustering using Spearman's distance, and reservoir computing forecasting to effectively identify fraud within smart grid systems. Specifically, the proposed methodology employs a clustering approach based on Spearman's rank distance to summarize time series data. This enables the aggregation of similar daily patterns, providing highly descriptive power and simplifying forecasting through Reservoir Computing. The subsequent step classifies each prosumer behavior as regular or potentially fraudulent. The SpeCluRC-NTL methodology, as proposed, is designed to detect fraud almost in real-time with low operational costs. The effectiveness of our approach is confirmed through empirical findings gathered from the Parc Bit distribution grid. This grid is located near Palma (Balearic Islands), Spain. The results of our research highlight the demonstrated effectiveness of the proposed approach, revealing its promising potential as it undergoes testing at the ParcBit premises. In comparison to previous works, SpeCluRC-NTL showcases its ability to reduce the false positive rate while maintaining a high true positive ratio, resulting in an increased AUC score. This has substantial implications for mitigating financial losses and addressing the various impacts associated with fraudulent activities in smart grids.Peer reviewe
From experimentation to structural change: fostering institutional entrepreneurship for public engagement in research and innovation
Get PDFPublisher Copyright: © The Author(s) 2023.Many researchers experiment with participatory settings to increase public engagement in research and innovation (R&I). Because of their temporary nature, it often remains unclear how such participatory experiments can contribute to structural change. This paper empirically explores options for bridging this gap. It analyzes how participants can be supported to act as institutional entrepreneurs to actively promote public engagement in R&I. To draw lessons, we analyze empirical material gathered on nineteen Social Labs which were set up to promote the uptake of Responsible Research and Innovation in a European R&I funding program (Horizon 2020). Involvement of motivated participants, insight in their institutional context, and specific methods and management choices that enhance a sense of agency are identified as essential for organizing change. These findings and the resulting framework of interventions may prove valuable for further (action) research into the institutionalization of public engagement in R&I.Peer reviewe
Building Digital Twins to Overcome Digitalization Barriers for Automating Construction Site Management
Get PDFPublisher Copyright: © 2024 by the authors.Construction sites are highly unpredictable environments involving a wide variety of stakeholders with complex information exchanges, which lead to the well-known inefficiencies and unproductivity of the construction sector. The adoption of Building Digital Twins (BDT) in the construction site is a promising solution to this issue, by automating data acquisition and knowledge extraction processes and providing what-if scenario simulation capabilities. Furthermore, the current research sets the principles to define, replicate, and scale-up the architecture of a Building Digital Twin Platform (BDTP), conceived as a scalar ecosystem, which allows to seamlessly manage on-site construction processes, integrating cross-cutting domains for the construction site optimization (Progress monitoring, Quality control, Operational Health and Safety, Equipment control, and Production planning). The starting point of the research is a comprehensive diagnosis of on-site process inefficiencies and the barriers to its digitalization leading to the user requirements, which have been underpinned by questionnaires and interviews addressed within an open innovation user-centered approach around Living Labs. The research has been conceived following the Design Science Research (DSR) methodology and based on the Plan-Do-Check-Act (PDCA) analysis for the continuous improvement of the construction process. By means of the adoption of the standard Business Process Model and Notation (BPMN), based on the BDTP architecture, the research has resulted in BPMN workflows stemmed from the Digital Twin (DT) where the DT itself is an actor in a service-oriented data-exchange workflow. Moreover, the use of a BDTP can pave the way for the transition from user-driven construction management to hybrid management, coexisting with both human and digital actors and merging expert knowledge with artificial intelligence techniques.Peer reviewe
Predictive-Cognitive Maintenance for Advanced Integrated railway Management
No full textPublisher Copyright: © 2024 11th European Workshop on Structural Health Monitoring, EWSHM 2024. All rights reserved.Railway systems play a vital role in modern transportation, and Predictive-Cognitive Maintenance (PCM) has emerged as a transformative approach in the context of Advanced Integrated Railway Management to ensure the safety, reliability, and efficiency of these systems. PCM leverages data analytics and machine learning to optimize railway system maintenance. This requires effective structural health monitoring (SHM) using low-cost sensor devices. This paper presents a prototype solar-powered wireless sensor node with a 3-axis MEMS accelerometer and energy-harvesting features for monitoring rail-track vibrations. The node contains a microcontroller that runs embedded machine learning models to preprocess the vibration data after train crossing. Abnormal vibrations indicative of defects were detected in real time using the TinyML inference at the edge. Instead of raw data, only the model results were wirelessly transmitted to a digital twin in the cloud. The digital twin aggregates data across the rail network for the system-level assessment of RUL and maintenance planning. This edge computing approach minimizes wireless transmission and cloud storage compared to raw sensor streaming. Embedded ML enables real-time damage detection, whereas cloud digital twins provide system-level prognostic insights. The solar-powered platform enables long-term remote monitoring at low cost without wiring or battery changes. A full-scale physical model was used to validate the edge node prototypes against calculation models and wired accelerometers for impulse loads. The results demonstrated that these nodes can provide a sensor layer for cost-effective PCM in railway systems. In summary, this study proposes an edge computing and embedded ML approach for SHM that integrates cloud-based digital twins to enable the predictive-cognitive maintenance of railway infrastructure. Wireless nodes demonstrate potential for low-cost, convenient, and automated rail health monitoring.Peer reviewe
Life cycle sustainability assessment of short chain carboxylic acid produced from municipal bio-wastes
Get PDFPublisher Copyright: © 2024 The AuthorsBio-based products are a fast-growing market due to increasing consumer consciousness for sustainability. Although this is per se a positive trend, it leads to a higher demand for organic feedstocks which normally comes from primary agricultural sources and can lead to undesired deforestation or other land use changes to farmland. At the same time, Europe is facing another challenge related with the treatment of organic wastes. In this context, the project CAFIPLA developed an integrated process to convert heterogeneous organic materials to building blocks for the bio-based economy. This study performs a life cycle sustainability assessment (life cycle assessment, life cycle costing and social life cycle assessment) of the production of short chain carboxylic acids (SCCA) employing municipal bio-wastes as a feedstock. In addition to a hot-spot identification to detect the main sources of impact, a comparison of the novel technology with the current benchmark is carried out applying a cradle-to-gate approach and using 1 kg of SCCA as a functional unit. Results show the great performance of CAFIPLA in all the environmental categories analysed. Furthermore, the profitability of the plant is also verified, reaching a payback period below 6 years as long as the product is sold above 0.49 €/kg. Finally, the potential social risk associated to the supply chain is also improved with CAFIPLA technology.Peer reviewe
Simplified Geometric Processing of Solar Radiation for Improved Data-Driven Modelling of Short-Term Energy & Comfort Performance in Buildings
No full textPublisher Copyright: © 2024 University of Split, FESB.This paper reviews different approaches to integrate solar radiation into data-driven energy models of buildings. A geometrical model is proposed to project solar radiation data over specific wall and window orientations. The proposed method is based on well-known engineering formulae, which is adapted for its application jointly with a short-term data-driven energy performance model. The suggested approach is tested over data from an existing building in Singapore, where the solar model is coupled with an ARX model. Overall, the model performs well, with one-hour ahead predictions with MAE in the range of 0.12-0.14 °C in the modeled rooms. The correction of solar radiation improves correlation with regards to this variable by 20%.Peer reviewe