Intelligent diagnosis of the severity of disease conditions in COVID-19 patients based on the LASSO method.

Purpose: The purpose of this study is to develop an intelligent diagnosis model based on the LASSO method to predict the severity of COVID-19 patients. Methods: The study uses the clinical data of 500 COVID-19 patients from a designated hospital in Guangzhou, China, and selects eight features, including age, sex, dyspnea, comorbidity, complication, lymphocytes (LYM), CRP, and lung injury score, as the most important predictors of COVID-19 severity. The study applies the LASSO method to perform feature selection and regularization, and compares the LASSO method with other machine learning methods, such as ridge regression, support vector machine, and random forest. Results: The study finds that the ridge regression model has the best performance among the four models, with an AUROC of 0.92 in the internal validation and 0.91 in the external validation. Conclusion: The study provides a simple, robust, and interpretable model for the intelligent diagnosis of COVID-19 severity, and a convenient and practical tool for the public and the health care workers to assess COVID-19 severity. However, the study also has some limitations and directions for future research, such as the need for more data from different sources and settings, and from prospective, longitudinal, multi-class classification models. The study hopes to contribute to the prevention and control of COVID-19, and to the improvement of the diagnosis and treatment of COVID-19 patients.

[Value of CT Quantitative Parameters in Prediction of Pathological Types
of Lung Ground Glass Nodules].

BACKGROUND: The pathological types of lung ground glass nodules (GGNs) show great significance to the clinical treatment. This study was aimed to predict pathological types of GGNs based on computed tomography (CT) quantitative parameters. METHODS: 389 GGNs confirmed by postoperative pathology were selected, including 138 cases of precursor glandular lesions [atypical adenomatous hyperplasia (AAH) and adenocarcinoma in situ (AIS)], 109 cases of microinvasive adenocarcinoma (MIA) and 142 cases of invasive adenocarcinoma (IAC). The morphological characteristics of nodules were evaluated subjectively by radiologist, as well as artificial intelligence (AI). RESULTS: In the subjective CT signs, the maximum diameter of nodule and the frequency of spiculation, lobulation and pleural traction increased from AAH+AIS, MIA to IAC. In the AI quantitative parameters, parameters related to size and CT value, proportion of solid component, energy and entropy increased from AAH+AIS, MIA to IAC. There was no significant difference between AI quantitative parameters and the subjective CT signs for distinguishing the pathological types of GGNs. CONCLUSIONS: AI quantitative parameters were valuable in distinguishing the pathological types of GGNs.

Plasmon-Induced Radical-Radical Heterocoupling Boosts Photodriven Oxidative Esterification of Benzyl Alcohol over Nitrogen-Doped Carbon-Encapsulated Cobalt Nanoparticles.

Selective oxidation of alcohols under mild conditions remains a long-standing challenge in the bulk and fine chemical industry, which usually requires environmentally unfriendly oxidants and bases that are difficult to separate. Here, a plasmonic catalyst of nitrogen-doped carbon-encapsulated metallic Co nanoparticles (Co@NC) with an excellent catalytic activity towards selective oxidation of alcohols is demonstrated. With light as only energy input, the plasmonic Co@NC catalyst effectively operates via combining action of the localized surface-plasmon resonance (LSPR) and the photothermal effects to achieve a factor of 7.8 times improvement compared with the activity of thermocatalysis. A high turnover frequency (TOF) of 15.6 h-1 is obtained under base-free conditions, which surpasses all the reported catalytic performances of thermocatalytic analogues in the literature. Detailed characterization reveals that the d states of metallic Co gain the absorbed light energy, so the excitation of interband d-to-s transitions generates energetic electrons. LSPR-mediated charge injection to the Co@NC surface activates molecular oxygen and alcohol molecules adsorbed on its surface to generate the corresponding radical species (e.g., ⋅O2 - , CH3 O⋅ and R-⋅CH-OH). The formation of multi-type radical species creates a direct and forward pathway of oxidative esterification of benzyl alcohol to speed up the production of esters.

Generalized Synthesis of Ternary Sulfide Hollow Structures with Enhanced Photocatalytic Performance for Degradation and Hydrogen Evolution.

A series of ternary sulfide hollow structures have been successfully prepared by a facile glutathione (GSH)-assisted one-step hydrothermal route, where GSH acts as the source of sulfur and bubble template. We demonstrate the feasibility and versatility of this in situ gas-bubble template strategy by the fabrication of novel hollow structures of MIn2S4 (M = Cd, Zn, Ca, Mg, and Mn). Interestingly, with the reaction time varying, the hierarchical CdIn2S4 microspheres with controlled internal structures can be regulated from yolk-shell, smaller yolk-shell (yolk-shell with shrunk yolk), hollow, to solid. Under visible-light irradiation, all of our prepared CdIn2S4 samples with different morphologies were photoactivated. In virtue of the appealing hierarchical hollow structure, the yolk-shell-structured CdIn2S4 microspheres exhibited the optimal photocatalytic activity and excellent durability for both the X3B degradation and H2 evolution, which can be ascribed to the synergy-promoting effect of the small crystallite size together with the unique structural advantages of the yolk-shell structure. Thus, we hypothesize that this proof-of-concept strategy paves an example of rational design of hollow structured ternary or multinary sulfides with superior photochemical performance, holding great potential for future multifunctional applications.