10784 research outputs found
Sort by
Cost-Sensitive Ordinal Classification Methods to Predict SARS-CoV-2 Pneumonia Severity
No full textPublisher Copyright: © 2013 IEEE.Objective: To study the suitability of cost-sensitive ordinal artificial intelligence-machine learning (AI-ML) strategies in the prognosis of SARS-CoV-2 pneumonia severity. Materials & methods: Observational, retrospective, longitudinal, cohort study in 4 hospitals in Spain. Information regarding demographic and clinical status was supplemented by socioeconomic data and air pollution exposures. We proposed AI-ML algorithms for ordinal classification via ordinal decomposition and for cost-sensitive learning via resampling techniques. For performance-based model selection, we defined a custom score including per-class sensitivities and asymmetric misprognosis costs. 260 distinct AI-ML models were evaluated via 10 repetitions of 5 × 5 nested cross-validation with hyperparameter tuning. Model selection was followed by the calibration of predicted probabilities. Final overall performance was compared against five well-established clinical severity scores and against a 'standard' (non-cost sensitive, non-ordinal) AI-ML baseline. In our best model, we also evaluated its explainability with respect to each of the input variables. Results: The study enrolled n = 1548 patients: 712 experienced low, 238 medium, and 598 high clinical severity. d = 131 variables were collected, becoming dprime = 148 features after categorical encoding. Model selection resulted in our best-performing AI-ML pipeline having: 1)no imputation of missing data, 2)no feature selection (i.e. using the full set of dprime features), 3)'Ordered Partitions' ordinal decomposition, 4)cost-based reimbalance, and 5)a Histogram-based Gradient Boosting classifier. This best model (calibrated) obtained a median accuracy of 68.1% [67.3%, 68.8%] (95% confidence interval), a balanced accuracy of 57.0% [55.6%, 57.9%], and an overall area under the curve (AUC) 0.802 [0.795, 0.808]. In our dataset, it outperformed all five clinical severity scores and the 'standard' AI-ML baseline. Discussion & conclusion: We conducted an exhaustive exploration of AI-ML methods designed for both ordinal and cost-sensitive classification, motivated by a real-world application domain (clinical severity prognosis) in which these topics arise naturally. Our model with the best classification performance exploited successfully the ordering information of ground truth classes, coping with imbalance and asymmetric costs. However, these ordinal and cost-sensitive aspects are seldom explored in the literature.Peer reviewe
Early Modern Irish Merchant CaptainsConnecting Ports, Connecting People
No full textPublisher Copyright: © 2024 Ediciones Universidad de Salamanca. All rights reserved.Merchant ship captains played a very meaningful role in the upsurge of the early modern (trans)atlantic trade. Nevertheless, the long periods spent on board, the uncertain time for stopover, and the itinerancy led them to a way of living depicted by rootlessness and constrained personal relationships. Based on three case studies of Irish shipmasters operating from 18th century Bilbao port —Moriarty, Malaghlin and Magra—, their professional careers are shown and how they impacted on their social and emotional life, as well as the role they played interlinking people.Peer reviewe
Corrosion mechanisms of plasma welded Nickel aluminium bronze immersed in seawater
No full textPublisher Copyright: © 2024 The AuthorsNickel Aluminium Bronzes (NAB) are copper-based multi-phase alloys used extensively in marine applications. NAB is vulnerable to seawater corrosion, however the interaction between its corrosion mechanisms and real-world factors including biofouling, weld microstructure and residual stress are poorly understood. Seawater corrosion tests were performed on plasma-welded NAB in laboratory and marine environments, demonstrating that the retained β’ phase in the Heat Affected Zone (HAZ) experiences Selective Phase Corrosion (SPC), whereas crevice corrosion associated with SPC of the κIII phase occurs at biofouled and stressed areas of parent material. These factors, seldom simulated in physical tests, severely impact NAB's corrosion resistance.Peer reviewe
Multiplex real-time PCR FilmArray performance in the diagnosis of meningoencephalitis: lights and shadows
No full textPublisher Copyright: © 2023, The Author(s).Purpose: We aimed to evaluate the performance of the FilmArray (FA) meningitis/encephalitis (ME) panel. Secondarily, we analyzed the false positive (FP) and false negative (FN) results, as well as the predictive values of the technique, regarding the cerebrospinal fluid (CSF) characteristics. Methods: FA is a multiplex real-time PCR detecting 14 of the most common ME pathogens in CSF. All FA performed at our hospital (2018–2022) were retrospectively reviewed. FA was compared to conventional techniques and its performance was assessed based on the final diagnosis of the episode. Results: FA was performed in 313 patients with suspicion of ME. Most patients had altered mental status (65.2%) and fever (61%). Regarding CSF characteristics, 49.8% and 53.7% presented high CSF proteins and pleocytosis, respectively. There were 84 (26.8%) positive FA results, mainly for HSV-1 (10.9%), VZV (5.1%), Enterovirus (2.6%), and S. pneumoniae (1.9%). In the 136 cases where both FA and routine methods were performed, there was a 25.7% lack of agreement. We identified 6.6% FN results, but 28.6% FP, mainly due to HSV-1. This resulted in a high negative predictive value (NPV) of 93.4%, but a positive predictive value (PPV) of 73%. Remarkably, PPV as low as 36.9%, and 70.2%, were found in cases without pleocytosis, or lack of high CSF protein levels, respectively. Conclusion: FA was associated with high NPV, but frequent FP results and low PPV, particularly for HSV-1, and especially in patients without high CSF protein levels or pleocytosis.Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. P.P.-A. (JR20/00012, PI21/00498, and ICI21/00103) and C.G.-V. (FIS PI21/01640 and ICI21/00103) have received research grants funded by Instituto de Salud Carlos III (ISCIII) and co-funded by the European Union. The funders had neither a specific role in study design or collection of data, nor in writing of the paper or decision to submit.Peer reviewe
Photoisomerization-induced LMR shift for UV radiation detection
No full textPublisher Copyright: © 2024 Elsevier B.V.This work presents an optical sensor for ultraviolet radiation (UV) detection, based on the combined effects of Lossy Mode Resonance (LMR) in the mid infrared (MIR) spectral region and the photoisomerization of a polymeric dye coating. LMR was obtained by means of a sputtered SnO2 thin film on a tetrafluoroethylene-perfluoro (or alkoxy Vinyl Ether, PFA) substrate, along with a photo sensitive coating based on poly R-478. Obtained devices shown response and recovery times of 12 and 43 s, respectively, for an UV excitation of 71 mW at 365 nm. Sensitivity as a function of the excitation wavelength was studied with the highest value of 26 nm/mW obtained at 280 nm. For this excitation wavelength, the limit of detection (LoD) obtained was 0.024 mW. Four different excitation wavelengths were used to cover all UV regions (280, 310, 365 and 395 nm). All measurements were performed at room temperature and humidity (25 °C ± 1 °C and 13% R.H. ± 2% R.H. respectively). As far as we know, it is the first time that the LMR effect has been recognized in combination with a photoisomerization process.Peer reviewe
Thermal conductivity of Portlandite: Molecular dynamics based approach
No full textPublisher Copyright: © 2023Energy storage provides a greener path of efficient energy utilization. Recent trends of research suggest concrete as a potential thermal energy storage (TES) material. Its cheap commercial availability makes it one of the most deserving candidates. Cement paste is the key glue of concrete, hence performance of its major components in thermal conduction seeks thorough scientific study. Portlandite or calcium hydroxide is the second major component of cement paste, microscopically visible at a scale length of <10−4 m. Molecular dynamics (MD) simulation has been utilized to investigate the thermal transport mechanism of portlandite at an atomistic scale. The thermal conductivity of this component has been computed using three major force fields which are compared with the experimental outcome using modulated differential scanning calorimetry (MDSC). It has been observed that the simulation outcomes are in order with the experimental value but the results are sensitive to the choice of force fields. Excitation of molecules during thermal transport is predominantly governed by lower-frequency molecular vibration indicating the existence of Boson Peaks. This manuscript presents the full thermal conductivity tensor of portlandite along with the possible mechanism associated with thermal transport.This research was also carried out under the umbrella of the project PoroPCM ( PCI2019-103657 ) funded by MCIN/ AEI /10.13039/501100011033 and co-founded by the European Union (Programación Conjunta Internacional 2019) and the project NRG-STORAGE ( GA 870114 ) funded by the European Commission . We also acknowledge the Transnational Common Laboratory (LTC) Aquitaine-Euskadi Network in Green Concrete. This work was born under the umbrella of the project “Photonic Metaconcrete with Infrared RAdiative Cooling Capacity for Large Energy Savings (MIRACLE)” funded by the EU .Peer reviewe
Comparative Study Of FAST And Conventional Sintering Of Cutting Tool Material
No full textPublisher Copyright: © European Powder Metallurgy Association (EPMA)This study explores the effectiveness of Field Assisted Sintering Technology (FAST) in producing cutting tools using a Bronze 90/10, Co, Fe, and W powder mix in granulated and non-granulated conditions. The goal is to evaluate and demonstrate FAST efficiency by comparing results obtained on the same system using a conventional sintering approach. For the characterization of the samples, a Rietveld refinement XRD patterns combined with microstructural analysis, based on SEM and EDXS, was considered. The study revealed that the sintering process significantly influenced microstructural characteristics. The FAST process, known for rapid heating and short sintering periods, maintained a consistent ratio of starting phases, unlike conventional sintering prone to reactions, diffusion, and intermetallic compound formation. Conventional sintering resulted in cobalt polymorphs also exhibited a lattice expansion with respect to the starting powder, resulting mainly from the atomic rearrangement during prolonged sintering process. FAST samples showed a lower volume change, due to short sintering time.Peer reviewe
Microstructural Characterization and Optimization of CoCrCuFeMnNi High Entropy Alloy to be in Contact with Hydrogen
No full textPublisher Copyright: © 2024 75th World Foundry Congress, WFC 2024. All rights reserved.This study aims to develop a High Entropy Alloy (HEA) with reduced susceptibility to Hydrogen embrittlement (HE) and improved resistance to oxidation, potentially surpassing existing materials in hydrogen-rich environments. Microstructural analysis revealed a dendritic structure where two FCC phases have been formed: a dendritic zone with the designed composition and a Cu-rich interdendritic structure with reduced Mn content. The second phase has been eliminated with the application of heat treatment but Cr-rich segregation in grain boundary appeared during a prolongated exposure at 1100℃, which could correspond to σ phase. Segregation of σ phase might be eliminated with an optimization of the composition or heat treatment conditions, achieving a single-phase solid solution.Peer reviewe
Process Optimization and Wearing Characterization of As-Cast and Heat-Treated Cocrfemnni(Alx) High Entropy Alloy
No full textPublisher Copyright: © 2024 75th World Foundry Congress, WFC 2024. All rights reserved.This study aims to study a CoCrFeMnNi(Alx) High Entropy Alloy (HEA) with high wear resistance produced by standard induction melting from scrap and ferroalloys. To the HEA base alloy, Al was added to promote a decrease in grain size and improve the final wearing properties The thermal treatment (TT) was optimized to increase the solubility of alloying elements at 1,100ºC for 36 hours and immediately water cooling. The coefficient of friction (COF) and wear rate values are like those obtained from pure metals and with multiple remelting phases by vacuum arc melting. Adding Al in the alloy promotes a decrease in hardness and COF, and an increase in wear rate values, which is anomalous in metals. TT increases hardness and decreases COF and wear rate. The study of samples explains the abnormal behavior by the finer microstructure in Al as-cast samples and the presence of more precipitates in TT samples. The obtained processes and alloys can have future use in automotive or energy applications.Peer reviewe
Ensuring Trustworthiness of Hybrid AI-Based Robotics Systems
No full textPublisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.Hybrid Artificial Intelligence (HAI) algorithms are well adapted to the industrial environment since they require significantly less data than data-driven Artificial Intelligence (AI) algorithms, matching the industry constraints. However, HAI does not fully address the issue of trust (validity, transparency, explainability, and ethics) which must be tackled to achieve world-class HAI beneficial to humans individually, organisationally, and societally. This paper focuses on describing the methods used to collect and elicit trustworthiness requirements in the European ULTIMATE project. It includes requirements register and a trustworthiness glossary of terms.Peer reviewe