Automatic Detection of COVID-19 Infection Using Chest X-Ray Images Through Transfer Learning

The new coronavirus (COVID-19), declared by the World Health Organization as a pandemic, has infected more than 1 million people and killed more than 50 thousand. An infection caused by COVID-19 can develop into pneumonia, which can be detected by a chest X-ray exam and should be treated appropriately. In this work, we propose an automatic detection method for COVID-19 infection based on chest X-ray images. The datasets constructed for this study are composed of 194 X-ray images of patients diagnosed with coronavirus and 194 X-ray images of healthy patients. Since few images of patients with COVID-19 are publicly available, we apply the concept of transfer learning for this task. We use different architectures of convolutional neural networks (CNNs) trained on ImageNet, and adapt them to behave as feature extractors for the X-ray images. Then, the CNNs are combined with consolidated machine learning methods, such as k-Nearest Neighbor, Bayes, Random Forest, multilayer perceptron (MLP), and support vector machine (SVM). The results show that, for one of the datasets, the extractor-classifier pair with the best performance is the MobileNet architecture with the SVM classifier using a linear kernel, which achieves an accuracy and an F1-score of 98.5%. For the other dataset, the best pair is DenseNet201 with MLP, achieving an accuracy and an F1-score of 95.6%. Thus, the proposed approach demonstrates efficiency in detecting COVID-19 in X-ray images.      

Spatial optimization of photovoltaic-based hydrogen-electricity supply chain through an integrated geographical information system and mathematical modeling approach

Clean Technologies and Environmental Policy - Hydrogen is a potential energy carrier for renewable energy as it has a clean emission when consumed. To implement hydrogen energy system in...     

Poly(ether ether sulfone)s and sulfonated poly(ether ether sulfone)s derived from functionalized 1,1‐diphenylethylene derivatives

Symmetrically disubstituted 1,1‐diphenylethylene derivatives, 1,1‐bis[4‐(t‐butyldimethylsiloxy)phenyl]ethylene and 1,1‐bis(4‐hydroxyphenyl)ethylene, were prepared by new synthesis methods and employed as functionalized monomers in the preparation of new poly(ether ether sulfone) derivatives, with the introduction of the 1,1‐diphenylethylene unit along the polymer backbone. A series of parent 1,1‐diphenylethylene based poly(ether ether sulfone)s were prepared via (a) the cesium fluoride catalysed polycondensation reactions of dihalogenated diaryl sulfones with silylated bisphenols and (b) the base catalysed aromatic nucleophilic substitution polycondensation reaction between aromatic dihalogenated diaryl sulfones with bisphenols. The post‐polymerization sulfonation reaction via the thiol‐ene addition reaction of the poly(ether ether sulfone) precursor containing the 1,1‐diphenylethylene unit along the backbone, with sodium 3‐mercapto‐1‐propane sulfonate, afforded well‐defined sulfonated poly(ether ether sulfone), with the alkyl sulfonate group introduced pendant to the polymer backbone. The organic compounds and different polymer derivatives were characterized by standard chromatography, spectroscopy, spectrometry, thermal analyses, microscopy and X‐ray diffraction measurements. © 2016 Society of Chemical Industry     

Catalytic processing of atmospheric residue of petroleum over AlSBA-15 nanomaterials with different acidity

The AlSBA-15 nanomaterials with different Si/Al ratios were investigated as a promising material for processing of atmospheric residue of petroleum (ATR). The processing was performed in a thermobalance using nitrogen as a carrier gas at different heating hates. The catalytic pyrolysis was performed with ATR sample containing 20% mass of AlSBA-15. Based on TG data by applying Ozawa-Flynn-Wall kinetic model, the activation energy for the pyrolysis of ATR was higher than for the ATR/AlSBA-15, throughout all range of decomposition, indicating that the nanostructured material is a promising material for processing petroleum residues in order to obtain liquid products in the gasoline and diesel range.     

Economic and environmental optimization of the biobutanol purification process

Current technologies for the production of biobutanol by fermentation have improved the production processes. These new technology improvements are economically viable with respect to the petrochemical pathway. For this, the aim of this paper is to compare four different process designs for the purification of biobutanol by solving a multi-objective optimization process involving two objective functions: the total annual cost and return of investment as economic functions and the associated eco-indicator 99 as an environmental function. The process associated to the routes A, B, and C consists of a steam stripping distillation and distillation columns, while the process D includes distillation columns with a liquid–liquid extraction column. Process modeling was performed in the Aspen Plus software, and the multi-objective optimization was conducted using differential evolution with tabu list as a stochastic optimization method. Results indicate that the process route D is the most profitable design and the process route C has the lowest environmental impact measured through the eco-indicator 99 method. Additionally, the use of a solar collector against steam has been compared in order to produce the required heat duty needed in every single distillation column to have a broader view about the environmental and economic impact of these devices.     

Bed Expansion and Particle Classification in Liquid Fluidized Beds with Structured Internals

The inclusion of a structured packing as internal in a liquid‐solid fluidized bed allows expansion of the liquid velocity operation range before elutriation, promoting the liquid solid contact and mixing. The bed expansion of liquid‐solid fluidized beds provided with structured packing as internals is examined, for solids denser than the liquid phase and within a wide range of operating conditions. A correlation to estimate the bed expansion in liquid‐solid fluidized beds using structured packing as internals is developed. In addition, the feasibility of employing structured packing as internals for favoring classification of different density particles is demonstrated by analyzing the mass elutriated from the column at different liquid velocities for single particles or binary mixtures.     

Gelcasting of alumina suspensions containing nanoparticles with glycerol monoacrylate

An acrylic monomer of low toxicity containing two hydroxyl groups has been synthesized and used for gelcasting in water. The results have been compared to those achieved with the use of a commercially available monomer (2-hydroxyethyl acrylate). Due to the chemical structure of the synthesized monomer, no addition of the crosslinking agent was necessary for gelation and similar results in terms of rheology of suspensions, density and microstructure of the bodies were obtained with respect to those obtained with the commercial monomer. However, higher time for gelation was observed.Two alumina powders with very different particle sizes were used in this study: a commercial submicron-sized powder (d₅₀ = 0.35 μm) and a nanometer-sized alumina obtained by freeze-drying from aluminum sulphate solutions. The rheological behavior of concentrated suspensions was studied in order to establish their stability and to analyse the effect of the different monomers used in the process. Once the suspensions were optimized, the influence of the size of the powder on the gelation process was studied. The sintered density of submicrometer-sized alumina was higher (99%) than that measured when the bimodal suspension was used due to the difficulty to obtain highly concentrated suspensions from nanometric powder.     

HPCVIEW: A Tool for Top-down Analysis of Node Performance

It is increasingly difficult for complex scientific programs to attain a significant fraction of peak performance on systems that are based on microprocessors with substantial instruction-level parallelism and deep memory hierarchies. Despite this trend, performance analysis and tuning tools are still not used regularly by algorithm and application designers. To a large extent, existing performance tools fail to meet many user needs and are cumbersome to use. To address these issues, we developed HPCVIEW—a toolkit for combining multiple sets of program profile data, correlating the data with source code, and generating a database that can be analyzed anywhere with a commodity Web browser. We argue that HPCVIEW addresses many of the issues that have limited the usability and the utility of most existing tools. We originally built HPCVIEW to facilitate our own work on data layout and optimizing compilers. Now, in addition to daily use within our group, HPCVIEW is being used by several code development teams in DoD and DoE laboratories as well as at NCSA.