Low-rank latent matrix-factor prediction modeling for generalized high-dimensional matrix-variate regression.

Motivated by diagnosing the COVID-19 disease using two-dimensional (2D) image biomarkers from computed tomography (CT) scans, we propose a novel latent matrix-factor regression model to predict responses that may come from an exponential distribution family, where covariates include high-dimensional matrix-variate biomarkers. A latent generalized matrix regression (LaGMaR) is formulated, where the latent predictor is a low-dimensional matrix factor score extracted from the low-rank signal of the matrix variate through a cutting-edge matrix factor model. Unlike the general spirit of penalizing vectorization plus the necessity of tuning parameters in the literature, instead, our prediction modeling in LaGMaR conducts dimension reduction that respects the geometric characteristic of intrinsic 2D structure of the matrix covariate and thus avoids iteration. This greatly relieves the computation burden, and meanwhile maintains structural information so that the latent matrix factor feature can perfectly replace the intractable matrix-variate owing to high-dimensionality. The estimation procedure of LaGMaR is subtly derived by transforming the bilinear form matrix factor model onto a high-dimensional vector factor model, so that the method of principle components can be applied. We establish bilinear-form consistency of the estimated matrix coefficient of the latent predictor and consistency of prediction. The proposed approach can be implemented conveniently. Through simulation experiments, the prediction capability of LaGMaR is shown to outperform some existing penalized methods under diverse scenarios of generalized matrix regressions. Through the application to a real COVID-19 dataset, the proposed approach is shown to predict efficiently the COVID-19.

Antibody- and aptamer-free SERS substrate for ultrasensitive and anti-interference detection of SARS-CoV-2 spike protein in untreated saliva.

Sensitive and anti-interference detection of targeted signal(s) in body fluids is one of the paramount tasks in biosensing. Overcoming the complication and high cost of antibody/aptamer-modification, surface-enhanced Raman spectroscopy (SERS) based on antibody/aptamer-free (AAF) substrates has shown great promise, yet with rather limited detection sensitivity. Herein, we report ultrasensitive and anti-interference detection of SARS-CoV-2 spike protein in untreated saliva by an AAF SERS substrate, applying the evanescent field induced by the high-order waveguide modes of well-defined nanorods for SERS for the first time. A detection limit of 3.6 × 10-17 M and 1.6 × 10-16 M are obtained in phosphate buffered saline and untreated saliva, respectively; the detection limits are three orders of magnitude improved than the best records from AAF substrates. This work unlocks an exciting path to design AAF SERS substrates for ultrasensitive biosensing, not limited to detection of viral antigens.

Psychological capital and alienation among patients with COVID-19 infection: the mediating role of social support.

BACKGROUND: COVID-19 infection continues all over the world, causing serious physical and psychological impacts to patients. Patients with COVID-19 infection suffer from various negative emotional experiences such as anxiety, depression, mania, and alienation, which seriously affect their normal life and is detrimental to the prognosis. Our study is aimed to investigate the effect of psychological capital on alienation among patients with COVID-19 and the mediating role of social support in this relationship. METHODS: The data were collected in China by the convenient sampling. A sample of 259 COVID-19 patients completed the psychological capital, social support and social alienation scale and the structural equation model was adopted to verify the research hypotheses. RESULTS: Psychological capital was significantly and negatively related to the COVID-19 patients' social alienation (p < .01). And social support partially mediated the correlation between psychological capital and patients' social alienation (p < .01). CONCLUSION: Psychological capital is critical to predicting COVID-19 patients' social alienation. Social support plays an intermediary role and explains how psychological capital alleviates the sense of social alienation among patients with COVID-19 infection.

SARS-CoV-2 bivalent mRNA vaccine with broad protection against variants of concern.

Introduction: The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Omicron variant has rapidly spread around the globe. With a substantial number of mutations in its Spike protein, the SARS-CoV-2 Omicron variant is prone to immune evasion and led to the reduced efficacy of approved vaccines. Thus, emerging variants have brought new challenges to the prevention of COVID-19 and updated vaccines are urgently needed to provide better protection against the Omicron variant or other highly mutated variants. Materials and methods: Here, we developed a novel bivalent mRNA vaccine, RBMRNA-405, comprising a 1:1 mix of mRNAs encoding both Delta-derived and Omicron-derived Spike proteins. We evaluated the immunogenicity of RBMRNA-405 in BALB/c mice and compared the antibody response and prophylactic efficacy induced by monovalent Delta or Omicron-specific vaccine with the bivalent RBMRNA-405 vaccine in the SARSCoV-2 variant challenge. Results: Results showed that the RBMRNA-405 vaccine could generate broader neutralizing antibody responses against both Wuhan-Hu-1 and other SARS-CoV-2 variants, including Delta, Omicron, Alpha, Beta, and Gamma. RBMRNA-405 efficiently blocked infectious viral replication and lung injury in both Omicron- and Delta-challenged K18-ACE2 mice. Conclusion: Our data suggest that RBMRNA-405 is a promising bivalent SARS-CoV-2 vaccine with broad-spectrum efficacy for further clinical development.

Early autoimmunity and outcome in virus encephalitis: a retrospective study based on tissue-based assay.

To explore the autoimmune response and outcome in the central nervous system (CNS) at the onset of viral infection and correlation between autoantibodies and viruses. METHODS: A retrospective observational study was conducted in 121 patients (2016-2021) with a CNS viral infection confirmed via cerebrospinal fluid (CSF) next-generation sequencing (cohort A). Their clinical information was analysed and CSF samples were screened for autoantibodies against monkey cerebellum by tissue-based assay. In situ hybridisation was used to detect Epstein-Barr virus (EBV) in brain tissue of 8 patients with glial fibrillar acidic protein (GFAP)-IgG and nasopharyngeal carcinoma tissue of 2 patients with GFAP-IgG as control (cohort B). RESULTS: Among cohort A (male:female=79:42; median age: 42 (14-78) years old), 61 (50.4%) participants had detectable autoantibodies in CSF. Compared with other viruses, EBV increased the odds of having GFAP-IgG (OR 18.22, 95% CI 6.54 to 50.77, p<0.001). In cohort B, EBV was found in the brain tissue from two of eight (25.0%) patients with GFAP-IgG. Autoantibody-positive patients had a higher CSF protein level (median: 1126.00 (281.00-5352.00) vs 700.00 (76.70-2899.00), p<0.001), lower CSF chloride level (mean: 119.80±6.24 vs 122.84±5.26, p=0.005), lower ratios of CSF-glucose/serum-glucose (median: 0.50[0.13-0.94] vs 0.60[0.26-1.23], p=0.003), more meningitis (26/61 (42.6%) vs 12/60 (20.0%), p=0.007) and higher follow-up modified Rankin Scale scores (1 (0-6) vs 0 (0-3), p=0.037) compared with antibody-negative patients. A Kaplan-Meier analysis revealed that autoantibody-positive patients experienced significantly worse outcomes (p=0.031). CONCLUSIONS: Autoimmune responses are found at the onset of viral encephalitis. EBV in the CNS increases the risk for autoimmunity to GFAP.

Current hotspot and study trend of innate immunity in COVID-19: a bibliometric analysis from 2020 to 2022.

Background: Since the coronavirus disease 2019 (COVID-19) has spread throughout the world, many studies on innate immunity in COVID-19 have been published, and great progress has been achieved, while bibliometric analysis on hotspots and research trends in this field remains lacking. Methods: On 17 November 2022, articles and reviews on innate immunity in COVID-19 were recruited from the Web of Science Core Collection (WoSCC) database after papers irrelevant to COVID-19 were further excluded. The number of annual publications and the average citations per paper were analyzed by Microsoft Excel. Bibliometric analysis and visualization of the most prolific contributors and hotspots in the field were performed by VOSviewer and CiteSpace software. Results: There were 1,280 publications that met the search strategy on innate immunity in COVID-19 and were published from 1 January 2020 to 31 October 2022. Nine hundred thirteen articles and reviews were included in the final analysis. The USA had the highest number of publications (Np) at 276 and number of citations without self-citations (Nc) at 7,085, as well as an H-index of 42, which contributed 30.23% of the total publications, followed by China (Np: 135, Nc: 4,798, and H-index: 23) with 14.79% contribution. Regarding Np for authors, Netea, Mihai G. (Np: 7) from the Netherlands was the most productive author, followed by Joosten, Leo A. B. (Np: 6) and Lu, Kuo-Cheng (Np: 6). The Udice French Research Universities had the most publications (Np: 31, Nc: 2,071, H-index: 13), with an average citation number (ACN) at 67. The journal Frontiers in Immunology possessed the most publications (Np: 89, Nc: 1,097, ACN: 12.52). "Evasion" (strength 1.76, 2021-2022), "neutralizing antibody" (strength 1.76, 2021-2022), "messenger RNA" (strength 1.76, 2021-2022), "mitochondrial DNA" (strength 1.51, 2021-2022), "respiratory infection" (strength 1.51, 2021-2022), and "toll-like receptors" (strength 1.51, 2021-2022) were the emerging keywords in this field. Conclusion: The study on innate immunity in COVID-19 is a hot topic. The USA was the most productive and influential country in this field, followed by China. The journal with the most publications was Frontiers in Immunology. "Messenger RNA," "mitochondrial DNA," and "toll-like receptors" are the current hotspots and potential targets in future research.

Functional Foods: A Promising Strategy for Restoring Gut Microbiota Diversity Impacted by SARS-CoV-2 Variants.

Natural herbs and functional foods contain bioactive molecules capable of augmenting the immune system and mediating anti-viral functions. Functional foods, such as prebiotics, probiotics, and dietary fibers, have been shown to have positive effects on gut microbiota diversity and immune function. The use of functional foods has been linked to enhanced immunity, regeneration, improved cognitive function, maintenance of gut microbiota, and significant improvement in overall health. The gut microbiota plays a critical role in maintaining overall health and immune function, and disruptions to its balance have been linked to various health problems. SARS-CoV-2 infection has been shown to affect gut microbiota diversity, and the emergence of variants poses new challenges to combat the virus. SARS-CoV-2 recognizes and infects human cells through ACE2 receptors prevalent in lung and gut epithelial cells. Humans are prone to SARS-CoV-2 infection because their respiratory and gastrointestinal tracts are rich in microbial diversity and contain high levels of ACE2 and TMPRSS2. This review article explores the potential use of functional foods in mitigating the impact of SARS-CoV-2 variants on gut microbiota diversity, and the potential use of functional foods as a strategy to combat these effects.

Endomembrane remodeling in SARS-CoV-2 infection.

During severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, the viral proteins intimately interact with host factors to remodel the endomembrane system at various steps of the viral lifecycle. The entry of SARS-CoV-2 can be mediated by endocytosis-mediated internalization. Virus-containing endosomes then fuse with lysosomes, in which the viral S protein is cleaved to trigger membrane fusion. Double-membrane vesicles generated from the ER serve as platforms for viral replication and transcription. Virions are assembled at the ER-Golgi intermediate compartment and released through the secretory pathway and/or lysosome-mediated exocytosis. In this review, we will focus on how SARS-CoV-2 viral proteins collaborate with host factors to remodel the endomembrane system for viral entry, replication, assembly and egress. We will also describe how viral proteins hijack the host cell surveillance system-the autophagic degradation pathway-to evade destruction and benefit virus production. Finally, potential antiviral therapies targeting the host cell endomembrane system will be discussed.

Classification of tuberculosis-related programmed cell death-related patient subgroups and associated immune cell profiling.

Background: Tuberculosis (TB) is the deadliest communicable disease in the world with the exception of the ongoing COVID-19 pandemic. Programmed cell death (PCD) patterns play key roles in the development and progression of many disease states such that they may offer value as effective biomarkers or therapeutic targets that can aid in identifying and treating TB patients. Materials and methods: The Gene Expression Omnibus (GEO) was used to gather TB-related datasets after which immune cell profiles in these data were analyzed to examine the potential TB-related loss of immune homeostasis. Profiling of differentially expressed PCD-related genes was performed, after which candidate hub PCD-associated genes were selected via a machine learning approach. TB patients were then stratified into two subsets based on the expression of PCD-related genes via consensus clustering. The potential roles of these PCD-associated genes in other TB-related diseases were further examined. Results: In total, 14 PCD-related differentially expressed genes (DEGs) were identified and highly expressed in TB patient samples and significantly correlated with the abundance of many immune cell types. Machine learning algorithms enabled the selection of seven hub PCD-related genes that were used to establish PCD-associated patient subgroups, followed by the validation of these subgroups in independent datasets. These findings, together with GSVA results, indicated that immune-related pathways were significantly enriched in TB patients exhibiting high levels of PCD-related gene expression, whereas metabolic pathways were significantly enriched in the other patient group. Single cell RNA-seq (scRNA-seq) further highlighted significant differences in the immune status of these different TB patient samples. Furthermore, we used CMap to predict five potential drugs for TB-related diseases. Conclusion: These results highlight clear enrichment of PCD-related gene expression in TB patients and suggest that this PCD activity is closely associated with immune cell abundance. This thus indicates that PCD may play a role in TB progression through the induction or dysregulation of an immune response. These findings provide a foundation for further research aimed at clarifying the molecular drivers of TB, the selection of appropriate diagnostic biomarkers, and the design of novel therapeutic interventions aimed at treating this deadly infectious disease.

In silico design of miniprotein to inhibit SARS-CoV-2 variant Omicron spike protein.

Omicron is a novel variant of SARS-CoV-2 that is currently spreading globally as the dominant strain. The virus first enters the host cell through the receptor binding domain (RBD) of the spike protein by interacting with the angiotensin-converting enzyme 2 (ACE2). Thus, the RBD protein is an ideal target for the design of drugs against the Omicron variant. Here, we designed several miniprotein inhibitors in silico to combat the SARS-CoV-2 Omicron variant using single- and double-point mutation approaches, based on the structure of the initial inhibitor AHB2. Also, two parallel molecular dynamics (MD) simulations were performed for each system to reproduce the calculated results, and the binding free energy was evaluated with the MM/PBSA method. The evaluated values showed that all inhibitors, including AHB2, M7E, M7E + M43W, and M7E + M43Y, were energetically more beneficial to the binding with the RBD than ACE2. In particular, the mutant inhibitor M7E + M43Y possessed the highest binding affinity to RBD and was selected as the most promising "best" inhibitor among all inhibitors. In addition, the combination of multiple analysis methods, such as free energy landscape analysis (FEL), principal component analysis (PCA), dynamic cross-correlation matrix analysis (DCCM), and hydrogen bond, salt bridge, and hydrophobic interaction analysis, also demonstrated that the mutations significantly affect the dynamical behavior and binding pattern of the inhibitor binding to the RBD protein. The current work suggested that miniprotein inhibitors can form stable complex structures with the RBD protein and exert a blocking or inhibitory effect on the SARS-CoV-2 variant Omicron. In conclusion, this study has identified several novel mutant inhibitors with enhanced affinity to the RBD protein, and provided potential guidance and insights for the rational design of therapeutic approaches for the new SARS-CoV-2 variant Omicron.

Simultaneous detection and phylogenetic analysis of porcine epidemic diarrhea virus and porcine circovirus 4 in Henan province, China.

Porcine circovirus 4 (PCV4) is a recently discovered circovirus that was first reported in 2019 in several pigs in Hunan province of China and has also been identified in pigs infected with porcine epidemic diarrhea virus (PEDV). To further investigate the coinfection and genetic diversity of these two viruses, 65 clinical samples (including feces and intestinal tissues) were collected from diseased piglets on 19 large-scale pig farms in Henan province of China, and a duplex SYBR Green I-based quantitative real-time polymerase chain reaction (qPCR) assay was developed for detecting PEDV and PCV4 simultaneously. The results showed that the limit of detection was 55.2 copies/µL and 44.1 copies/µL for PEDV and PCV4, respectively. The detection rate for PEDV and PCV4 was 40% (26/65) and 38% (25/65), respectively, and the coinfection rate for the two viruses was 34% (22/65). Subsequently, the full-length spike (S) gene of eight PEDV strains and a portion of the genome containing the capsid (Cap) gene of three PCV4 strains were sequenced and analyzed. Phylogenetic analysis showed that all of the PEDV strains from the present study clustered in the G2a subgroup and were closely related to most of the PEDV reference strains from China from 2011 to 2021, but they differed genetically from a vaccine strain (CV777), a Korean strain (virulent DR1), and two Chinese strains (SD-M and LZC). It is noteworthy that two PEDV strains (HEXX-24 and HNXX-24XIA) were identified in one sample, and the HNXX-24XIA strain had a large deletion at amino acids 31-229 of the S protein. Moreover, a recombination event was observed in strain HEXX-24. Phylogenetic analysis based on the amino acid sequence of the PCV4 Cap protein revealed that PCV4 strains were divided into three genotypes: PCV4a1, PCV4a2, and PCV4b. Three strains in the present study belonged to PCV4a1, and they had a high degree of sequence similarity (>98% identity) to other PCV4 reference strains. This study not only provides technical support for field investigation of PEDV and PCV4 coinfection but also provides data for their prevention and control.

Clinicopathological and immunological features of new onset kidney disease: a rare event after SARS-CoV-2 vaccination.

The onset of various kidney diseases has been reported after severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) vaccination. However, detailed clinical and pathological features are lacking. We screened and analyzed patients with newly diagnosed kidney diseases after inactivated SARS-CoV-2 vaccination in Peking University First Hospital from January 2021 to August 2021, and compared them with the reported cases in the literature. We obtained samples of blood, urine and renal biopsy tissues. Clinical and laboratory information, as well as light microscopy, immunostaining and ultrastructural observations, were described. The SARS-CoV-2 spike protein and nucleoprotein were stained using the immunofluorescence technique in the kidney biopsy samples. SARS-CoV-2 specific antibodies were tested using magnetic particle chemiluminescence immunoassay. The study group included 17 patients with a range of conditions including immune-complex-mediated kidney diseases (IgA nephropathy, membranous nephropathy and lupus nephritis), podocytopathy (minimal change disease and focal segmental glomerulosclerosis) and others (antineutrophil-cytoplasmic-antibody-associated vasculitis, anti-glomerular basement membrane nephritis, acute tubulointerstitial nephritis and thrombotic microangiopathy). Seven patients (41.18%) developed renal disease after the first dose and ten (58.82%) after the second dose. The kidney disease spectrum as well as clinicopathological features are similar across different types of SARS-CoV-2 vaccines. We found no definitive evidence of SARS-CoV-2 spike protein or nucleoprotein deposition in the kidney biopsy samples. Seropositive markers implicated abnormal immune responses in predisposed individuals. Treatment and follow-up (median = 86 days) showed that biopsy diagnosis informed treatment and prognosis in all patients. In conclusion, we observed various kidney diseases following SARS-CoV-2 vaccine administration, which show a high consistency across different types of SARS-CoV-2 vaccines. Our findings provide evidence against direct vaccine protein deposition as the major pathomechanism, but implicate abnormal immune responses in predisposed individuals. These findings expand our understanding of SARS-CoV-2 vaccine renal safety.

AIGAN: Attention-encoding Integrated Generative Adversarial Network for the reconstruction of low-dose CT and low-dose PET images.

X-ray computed tomography (CT) and positron emission tomography (PET) are two of the most commonly used medical imaging technologies for the evaluation of many diseases. Full-dose imaging for CT and PET ensures the image quality but usually raises concerns about the potential health risks of radiation exposure. The contradiction between reducing the radiation exposure and remaining diagnostic performance can be addressed effectively by reconstructing the low-dose CT (L-CT) and low-dose PET (L-PET) images to the same high-quality ones as full-dose (F-CT and F-PET). In this paper, we propose an Attention-encoding Integrated Generative Adversarial Network (AIGAN) to achieve efficient and universal full-dose reconstruction for L-CT and L-PET images. AIGAN consists of three modules: the cascade generator, the dual-scale discriminator and the multi-scale spatial fusion module (MSFM). A sequence of consecutive L-CT (L-PET) slices is first fed into the cascade generator that integrates with a generation-encoding-generation pipeline. The generator plays the zero-sum game with the dual-scale discriminator for two stages: the coarse and fine stages. In both stages, the generator generates the estimated F-CT (F-PET) images as like the original F-CT (F-PET) images as possible. After the fine stage, the estimated fine full-dose images are then fed into the MSFM, which fully explores the inter- and intra-slice structural information, to output the final generated full-dose images. Experimental results show that the proposed AIGAN achieves the state-of-the-art performances on commonly used metrics and satisfies the reconstruction needs for clinical standards.

The relationship between eHealth literacy and palliative care knowledge, attitudes, and practice among nurses: a cross-sectional study.

BACKGROUND: The crucial role that nurses play in offering palliative care to patients with life-threatening diseases is widely acknowledged, but the correlation between their eHealth literacy and their knowledge, attitudes, and practice in this domain has yet to be investigated. This study is conducted to investigate the status of eHealth literacy and knowledge, attitudes, and practice regarding palliative care among nurses, and to examine their relationship. METHODS: A cross-sectional study design was conducted among 546 nurses selected from the first-class tertiary hospitals located both inside and outside of Zhejiang Province between May 12 and May 20, 2022. The online survey of eHealth literacy scale (eHEALS) and scale of knowledge, attitudes, and practice (KAP) regarding palliative care was performed using snowball sampling through the WeChat mini program "Questionnaire Star". The Spearman rank correlation and binary logistic regression model were used to analyze the independent association between eHealth literacy and KAP toward palliative care. RESULTS: The median scores of eHEALS and KAP regarding palliative care were 32 (interquartile range[IQR] 29 to 38) and 82 (IQR 54 to 106) points. The results of correlation analysis showed that the KAP regarding palliative care was significantly correlated with eHEALS (rho = 0.189, P < 0.001). In addition, the results of binary logistic regression analysis demonstrated that the eHEALS score was independently associated with the KAP score regarding palliative care when controlling for sociodemographic factors (OR = 2.109; P < 0.001). CONCLUSION: Nurses who worked in first-class tertiary hospitals have good levels of eHealth literacy, while the overall level of KAP regarding palliative care is moderate. Our findings highlight that the eHEALS score is independently associated with the KAP score regarding palliative care. Therefore, nursing managers should adopt multiple measures to comprehensively improve eHealth literacy among nurses, further enrich their knowledge of palliative care, promote a positive transformation of attitudes towards palliative care, and efficiently implement palliative care practice, in order to promote high-quality development of palliative care.

Does Preoperative Hookwire Localization Influence Postoperative Acute and Chronic Pain After Video-Assisted Thoracoscopic Surgery: A Prospective Cohort Study.

Purpose: This study aimed to investigate whether preoperative computerized tomography-guided hookwire localization-associated pain could affect acute and chronic postsurgical pain (CPSP) in patients undergoing video-assisted thoracoscopic surgery (VATS). Methods: We enrolled 161 adult patients who underwent elective VATS; sixty-nine patients experienced hookwire localization (Group A) and 69 did not (Group B). Group A was further subdivided into the multiple localization group (n=35, Group Amultiple) and the single localization group (n=34, Group Asingle) according to the number of hookwires. The numerical rating scale (NRS) was used preoperatively, during recovery at the post-anesthesia care unit (PACU), and the first two days, 3 and 6 months postoperatively. Furthermore, multivariate regression analysis was used to identify the risk factors associated with CPSP. The postoperative adverse events, length of hospital stay, and satisfaction in pain management were also recorded. Results: The incidence and severity of acute postoperative pain were similar between Group A and Group B (p > 0.05). The incidence (56.5% vs 30.4%, p = 0.002) and the NRS scores (2.0 [2.0-3.0] vs 1.0 [1.0-2.0], p = 0.011) for CPSP were significantly higher in Group A than in Group B at 3 months postoperatively. On subgroup analysis, compared with Group Asingle, the intensity of CPSP (2.0 [2.0-3.0] vs 2.0 [1.0-2.0], p = 0.005) in Group Amultiple was slightly higher at 3 months postoperatively. Conversely, the CPSP incidence (60.0% vs 29.4%, p = 0.011) was significantly higher at 6 months postoperatively in Group Amultiple. The multivariate regression analysis further validated hookwire localization as a risk factor for CPSP (odds ratio: 6.199, 95% confidence interval 2.049-18.749, p = 0.001). Patient satisfaction relating to pain management at 3 months postoperatively was lower in Group A (p = 0.034). Conclusion: The preoperative pain stress of hookwire localization increased the incidence and intensity of CPSP rather than acute pain at 3 months postoperatively, especially in patients with multiple hookwires.

Computed tomography–based COVID–19 triage through a deep neural network using mask–weighted global average pooling

Background There is an urgent need to find an effective and accurate method for triaging coronavirus disease 2019 (COVID-19) patients from millions or billions of people. Therefore, this study aimed to develop a novel deep-learning approach for COVID-19 triage based on chest computed tomography (CT) images, including normal, pneumonia, and COVID-19 cases. Methods A total of 2,809 chest CT scans (1,105 COVID-19, 854 normal, and 850 non-3COVID-19 pneumonia cases) were acquired for this study and classified into the training set (n = 2,329) and test set (n = 480). A U-net-based convolutional neural network was used for lung segmentation, and a mask-weighted global average pooling (GAP) method was proposed for the deep neural network to improve the performance of COVID-19 classification between COVID-19 and normal or common pneumonia cases. Results The results for lung segmentation reached a dice value of 96.5% on 30 independent CT scans. The performance of the mask-weighted GAP method achieved the COVID-19 triage with a sensitivity of 96.5% and specificity of 87.8% using the testing dataset. The mask-weighted GAP method demonstrated 0.9% and 2% improvements in sensitivity and specificity, respectively, compared with the normal GAP. In addition, fusion images between the CT images and the highlighted area from the deep learning model using the Grad-CAM method, indicating the lesion region detected using the deep learning method, were drawn and could also be confirmed by radiologists. Conclusions This study proposed a mask-weighted GAP-based deep learning method and obtained promising results for COVID-19 triage based on chest CT images. Furthermore, it can be considered a convenient tool to assist doctors in diagnosing COVID-19.

Reconsidering adaptive industrialized construction in Chinese rural areas: responding to the challenge of COVID-19

The spread of COVID-19 has caused an increasing demand for public medical rooms, especially in Chinese rural regions. Industrialized building techniques have been shown as capable of fulfilling this demand through the case of the Leishenshan Hospital. However, industrialized construction requires developed technologies and infrastructures, which are often non-existent in rural areas, thus making it difficult to replicate such a feat. Therefore, more suitable solutions for Chinese rural project delivery in the pandemic scenario are needed. Considering the constraints of pandemic prevention and rural applicability, the adaptive industrialized construction (AIC) method has potential as an alternative. This study evaluates the application of AIC by comparing simulated results using AIC and a conventional method, based on five evaluation indicators: construction speed, labourer distribution, material consumption, equipment utilization, and cost. Taking an actual project as the sample building, the results indicate that the AIC method has several advantages. These include a shorter construction period, less labourer gathering onsite, and a lower cost, suggesting it may be an effective solution for rural project delivery during the pandemic. Architects and contractors could employ the same evaluation method to explore more solutions and optimize the construction schedule for future rapid construction needs in rural areas in a pandemic.

Entry, egress and vertical transmission of SARS-CoV-2.

The high infectivity and pathogenicity of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) have caused the COVID-19 outbreak, one of the most devastating pandemics in more than a century. This pandemic has already left a trail of destruction, including enormous loss of life, a global economic slump, and widespread psychological damage. Despite assiduous world-wide endeavors, an effective cure for COVID-19 is still lacking. Surprisingly, infected neonates and children have relatively mild clinical manifestations and a much lower fatality rate than elderly adults. Recent studies have unambiguously demonstrated the vertical transmission of SARS-CoV-2 from infected pregnant women to fetuses, which creates yet another challenge for disease prevention. In this review, we will summarize the molecular mechanism for entry of SARS-CoV-2 into host cells, the basis for the failure of the lungs and other organs in severe acute cases, and the evidence for congenital transmission.

Computed tomography-based COVID-19 triage through a deep neural network using mask-weighted global average pooling.

Background: There is an urgent need to find an effective and accurate method for triaging coronavirus disease 2019 (COVID-19) patients from millions or billions of people. Therefore, this study aimed to develop a novel deep-learning approach for COVID-19 triage based on chest computed tomography (CT) images, including normal, pneumonia, and COVID-19 cases. Methods: A total of 2,809 chest CT scans (1,105 COVID-19, 854 normal, and 850 non-3COVID-19 pneumonia cases) were acquired for this study and classified into the training set (n = 2,329) and test set (n = 480). A U-net-based convolutional neural network was used for lung segmentation, and a mask-weighted global average pooling (GAP) method was proposed for the deep neural network to improve the performance of COVID-19 classification between COVID-19 and normal or common pneumonia cases. Results: The results for lung segmentation reached a dice value of 96.5% on 30 independent CT scans. The performance of the mask-weighted GAP method achieved the COVID-19 triage with a sensitivity of 96.5% and specificity of 87.8% using the testing dataset. The mask-weighted GAP method demonstrated 0.9% and 2% improvements in sensitivity and specificity, respectively, compared with the normal GAP. In addition, fusion images between the CT images and the highlighted area from the deep learning model using the Grad-CAM method, indicating the lesion region detected using the deep learning method, were drawn and could also be confirmed by radiologists. Conclusions: This study proposed a mask-weighted GAP-based deep learning method and obtained promising results for COVID-19 triage based on chest CT images. Furthermore, it can be considered a convenient tool to assist doctors in diagnosing COVID-19.