Search

Current standard protocols used in the clinic for diagnosing COVID-19 include molecular or antigen tests, generally complemented by a plain chest X-Ray. The combined analysis aims to reduce the significant number of false negatives of these tests and provide complementary evidence about the presence and severity of the disease. However, the procedure is not free of errors, and the interpretation of the chest X-Ray is only restricted to radiologists due to its complexity. With the long term goal to provide new evidence for the diagnosis, this paper presents an evaluation of different methods based on a deep neural network. These are the first steps to develop an automatic COVID-19 diagnosis tool using chest X-Ray images to differentiate between controls, pneumonia, or COVID-19 groups. The paper describes the process followed to train a Convolutional Neural Network with a dataset of more than 79, 500 X-Ray images compiled from different sources, including more than 8, 500 COVID-19 examples. Three different experiments following three preprocessing schemes are carried out to evaluate and compare the developed models. The aim is to evaluate how preprocessing the data affects the results and improves its explainability. Likewise, a critical analysis of different variability issues that might compromise the system and its effects is performed. With the employed methodology, a 91.5% classification accuracy is obtained, with an 87.4% average recall for the worst but most explainable experiment, which requires a previous automatic segmentation of the lung region., (This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.)

In this paper we present a computational study of aggregation in aqueous solutions of α-1,3-glucan captured using a coarse-grained (CG) model that can be extended to other polysaccharides. This CG model captures atomistic geometry (i.e., relative placement of the hydrogen bonding donors and acceptors within the monomer) of the α-1,3-glucan monomer, the directional interactions due to the donor-acceptor hydrogen bonds, and their effect on aggregation of multiple α-1,3-glucan chains without the extensive computational resources needed for simulations with atomistic models. Using this CG model, we conduct molecular dynamics simulations to assess the effect of varying α-1,3-glucan chain length and hydrogen bond interaction strengths on the aggregation of multiple chains at finite concentrations in implicit solvent. We quantify the hydrogen bonding strength needed for multiple chains to aggregate, the distribution of inter- and intra-chain hydrogen bonds within the aggregate and in some cases, the shapes of the aggregate. We also explore the effect of substitution/silencing of some randomly selected or specific hydrogen bonding sites in the chain on the aggregation and aggregate structure. In the unmodified α-1,3-glucan solution, the inter-chain hydrogen bonds cause the chains to aggregate into sheets. Random silencing of hydrogen bonding donor sites only increases the hydrogen bond strength needed for aggregation but retains the same aggregate structure as the unmodified chains. Specific silencing of the hydrogen-bonding site on the C6 carbon leads to the chains aggregating into planar sheets that then fold over to form hollow cylinders at intermediate hydrogen bond strength - 4.7 to 5.3 kcal mol-1. These cylindrical aggregates assemble end-to-end to form larger aggregates at higher hydrogen bond strengths.

Hollow silica particles (HSPs) have become the focus of interest in many laboratories recently, because of their versatility, stemming from the ability to control their size and shape, as well as surface functionalization. Determining the mechanical stability of hollow particles is essential for their use, both in applications in which they need to retain their structure, as well as those in which they need to break down. We have synthesized a series of HSPs (inner diameter of 231 nm) with increasing wall thickness (7-25 nm), using a template approach. Their mechanical stability was measured using mercury intrusion porosimetry (MIP), which represents the novel application of the technique for these materials. The samples with complete shells break at progressively higher pressures, and samples with wall thickness ≥21 nm remain stable to the highest pressure applied (414 MPa). Other characterization methods, namely microscopy, gas adsorption, and small-angle X-ray scattering, shed light on the size parameters of the particles, as well as the porosity of the silica walls. By varying the amount of silica precursor used in the template coating step, we were able to produce hollow silicas with variable stability, thereby allowing for control of their mechanical properties.

The microstructure of perfluorinated sulfonic acid proton-exchange membranes such as Nafion significantly affects their transport properties and performance in a vanadium redox-flow battery (VRB). In this work, Nafion membranes with various equivalent weights ranging from 1000 to 1500 are prepared and the morphology-property-performance relationship is investigated. NMR and small-angle X-ray scattering studies revealed their composition and morphology variances, which lead to major differences in key transport properties related to proton conduction and vanadium-ion permeation. Their performances are further characterized as VRB membranes. On the basis of this understanding, a new perfluorosulfonic acid membrane is designed with optimal pore geometry and thickness, leading to higher ion selectivity and lower cost compared with the widely used Nafion 115. Excellent VRB single-cell performance (89.3% energy efficiency at 50 mA·cm -2 ) was achieved along with a stable cyclical capacity over prolonged cycling.