Cohesive Multi-Modality Feature Learning and Fusion for COVID-19 Patient Severity Prediction.

The outbreak of coronavirus disease (COVID-19) has been a nightmare to citizens, hospitals, healthcare practitioners, and the economy in 2020. The overwhelming number of confirmed cases and suspected cases put forward an unprecedented challenge to the hospital's capacity of management and medical resource distribution. To reduce the possibility of cross-infection and attend a patient according to his severity level, expertly diagnosis and sophisticated medical examinations are often required but hard to fulfil during a pandemic. To facilitate the assessment of a patient's severity, this paper proposes a multi-modality feature learning and fusion model for end-to-end covid patient severity prediction using the blood test supported electronic medical record (EMR) and chest computerized tomography (CT) scan images. To evaluate a patient's severity by the co-occurrence of salient clinical features, the High-order Factorization Network (HoFN) is proposed to learn the impact of a set of clinical features without tedious feature engineering. On the other hand, an attention-based deep convolutional neural network (CNN) using pre-trained parameters are used to process the lung CT images. Finally, to achieve cohesion of cross-modality representation, we design a loss function to shift deep features of both-modality into the same feature space which improves the model's performance and robustness when one modality is absent. Experimental results demonstrate that the proposed multi-modality feature learning and fusion model achieves high performance in an authentic scenario.

Classification of COVID-19 by Compressed Chest CT Image through Deep Learning on a Large Patients Cohort.

Corona Virus Disease (COVID-19) has spread globally quickly, and has resulted in a large number of causalities and medical resources insufficiency in many countries. Reverse-transcriptase polymerase chain reaction (RT-PCR) testing is adopted as biopsy tool for confirmation of virus infection. However, its accuracy is as low as 60-70%, which is inefficient to uncover the infected. In comparison, the chest CT has been considered as the prior choice in diagnosis and monitoring progress of COVID-19 infection. Although the COVID-19 diagnostic systems based on artificial intelligence have been developed for assisting doctors in diagnosis, the small sample size and the excessive time consumption limit their applications. To this end, this paper proposed a diagnosis prototype system for COVID-19 infection testing. The proposed deep learning model is trained and is tested on 2267 CT sequences from 1357 patients clinically confirmed with COVID-19 and 1235 CT sequences from non-infected people. The main highlights of the prototype system are: (1) no data augmentation is needed to accurately discriminate the COVID-19 from normal controls with the specificity of 0.92 and sensitivity of 0.93; (2) the raw DICOM image is not necessary in testing. Highly compressed image like Jpeg can be used to allow a quick diagnosis; and (3) it discriminates the virus infection within 6 seconds and thus allows an online test with light cost. We also applied our model on 48 asymptomatic patients diagnosed with COVID-19. We found that: (1) the positive rate of RT-PCR assay is 63.5% (687/1082). (2) 45.8% (22/48) of the RT-PCR assay is negative for asymptomatic patients, yet the accuracy of CT scans is 95.8%. The online detection system is available: http://212.64.70.65/covid .

Prognostic value of plasma brain natriuretic peptide value for patientswith sepsis: A meta-analysis.

PURPOSE: The aim of this meta-analysis was to clarify the diagnostic role of plasma BNP and NT-proBNP in predicting mortality for septic patients. METHODS: A systematic review was conducted prior to January 2018. Summary sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR) and diagnostic odds ratio (DOR) of the prognostic value of plasma BNP and NT-proBNP for septic patients. The area under the receiver operating curves (AUROC) were used to summarize overall test performance. RESULTS: Twenty-two studies with 3417 septic patients were selected in the analysis. The summary sensitivity, specificity, PLR, NLR, DOR and the AUROC of the overall analysis of BNP were: 0.84, 0.73, 3.1, 0.22, 14, 0.85; and these values of NT-proBNP were: 0.71, 0.73, 2.6, 0.39, 7 and 0.7 respectively; Subgroup analysis and meta-regression analyses showed that the tested method and observation endpoint influenced the summary sensitivity, specificity of BNP, but the tested day, tested method or observation endpoint did not influence the summary sensitivity, specificity of NT-proBNP. CONCLUSIONS: This meta-analysis indicates that both elevated plasma BNP and NT-proBNP have moderate predicts value for the mortality of septic patients, and the tested method and observation endpoint influence the results of BNP.

HMGB1 binding to receptor for advanced glycation end products enhances inflammatory responses of human bronchial epithelial cells by activating p38 MAPK and ERK1/2.

The proinflammatory factor high mobility group box protein 1 (HMGB1) has been implicated as an important mediator of many chronic inflammatory diseases, including asthma. Human bronchial epithelial cells (HBECs) play a central role in the pathogenesis of asthma. However, the effects of HMGB1 on HBECs and the underlying mechanisms remain unknown. Here, we investigated receptor expression and proinflammatory cytokine production by primary cultures of HBECs stimulated by HMGB1. We then examined the effects of specific receptor blockade and inhibition of p38 MAPK, ERK1/2, or PI3-K on HMGB1-induced expression of proinflammatory cytokines. HMGB1 increased the expression and secretion of TNF-α, TSLP, MMP-9, and VEGF in a dose- and time-dependent manner. HMGB1 also induced elevated expression of RAGE protein. Secretion of TNF-α, VEGF, MMP-9, and TSLP was significantly decreased by RAGE blockade and p38 MAPK pathway inhibition, while a less pronounced effect was mediated by ERK1/2 inhibition. These observations suggest that HMGB1 binds RAGE and promotes activities of p38 MAPK and ERK1/2 pathways in HBECs. This then enhances the expression of TNF-α, VEGF, MMP-9, and TSLP, which are the important inflammatory factors in asthma. These results demonstrate that HMGB1 enhances the inflammatory responses of HBECs, which are involved in the modulation of inflammatory processes in asthma.

HMGB1 contributes to allergen-induced airway remodeling in a murine model of chronic asthma by modulating airway inflammation and activating lung fibroblasts.

The pro-inflammation factor high-mobility group box protein 1 (HMGB1) has been implicated in the pathogenesis of asthma. In this study, we used a murine model of chronic asthma to evaluate the effects of HMGB1 on airway remodeling. Female BALB/c mice were randomly divided into four groups: control, ovalbumin (OVA) asthmatic, OVA+isotype antibody and OVA+anti-HMGB1 antibody. Anti-HMGB1 antibody therapy was started on day 21 and was administered three times per week for 6 weeks before intranasal challenge with OVA. In this mouse model, HMGB1 expression is significantly elevated. The anti-HMGB1 antibody group exhibited decreased levels of immunoglobulin E (IgE) and inflammatory mediators and reduced inflammatory cell accumulation, airway hyperresponsiveness (AHR), mucus synthesis, smooth muscle thickness and lung collagen content compared with the OVA groups. Treatment with HMGB1 increased proliferation, migration, collagen secretion and α-smooth muscle actin (SMA) expression in MRC-5 cells. Treatment with the HMGB1/IL-1ß complex significantly increased the expression and secretion of transforming growth factor (TGF-ß1), matrix metalloproteinase (MMP)-9 and vascular endothelial growth factor (VEGF). Altogether, these results suggest that blocking HMGB1 activity may reverse airway remodeling by suppressing airway inflammation and modulating lung fibroblast phenotype and activation.

Population dynamics of an Acinetobacter baumannii clonal complex during colonization of patients.

Acinetobacter baumannii has emerged as one of the leading pathogens causing hospital-acquired infection. The success of A. baumannii as a pathogen has to a large extent been attributed to its capacity to remodel its genome. Several major epidemic clonal complexes of A. baumannii spread across different health care facilities around the world, each of which contains a subset of diversified strains. However, little is known about the population dynamics during colonization of A. baumannii within hosts. Here, whole-genome sequencing was used to analyze population dynamics of A. baumannii strains isolated from a group of patients at different time points as well as from different sites of a particular patient. Seven out of nine of the sampled A. baumannii strains belonged to the international clone II (CC92 clonal complex). While the A. baumannii strains were found to be stable in three patients, there was a change of A. baumannii strains in one patient. Comparative genomic analysis revealed that the accessory genome of these strains contained a large set of virulence-encoding genes and these virulence factors might play a role in determining population dynamics. Microscale genome modification has been revealed by analysis of single nucleotide polymorphisms (SNPs) between A. baumannii strains isolated from the same patient. Parallel evolutionary traits have been observed during genome diversification when A. baumannii colonize in different patients. Our study suggested that both antibiotic usage and host environment might impose selective forces that drive the rapid adaptive evolution in colonizing A. baumannii.

Ultrafiltration versus intravenous diuretic therapy to treat acute heart failure: a systematic review.

BACKGROUND: Patients with decompensated heart failure frequently present with volume overload, which is conventionally treated with diuretics. These drugs have been associated with several adverse effects, including increased mortality, leading some clinicians to propose ultrafiltration as a safe alternative to remove sodium and water. OBJECTIVE: The objective of our study was to compare the safety and efficacy of ultrafiltration and conventional intravenous diuretic therapy for patients with acute heart failure and volume overload. DATA SOURCES: We searched the following databases through November 2012: Cochrane Library (1993-), PubMed (1988-), OVID (1984-), EBSCO (1984-), CBM (1978-), VIP (1989-), and CNKI (1979-). In addition, we manually searched relevant references and review articles. STUDY SELECTION: Randomized controlled trials comparing the efficacy of ultrafiltration and intravenous diuretics in patients diagnosed with hypervolemic acute heart failure were included. Five trials were found to satisfy all the inclusion criteria. STUDY APPRAISAL AND SYNTHESIS METHODS: Two reviewers independently determined study eligibility, assessed methodological quality and extracted the data. We analyzed the data and pooled them, when appropriate, using Revman 5.0. We assessed the risk of bias in the included studies using guidelines in the Cochrane Handbook 5.0 for Systematic Reviews of Interventions, taking into account sequence generation, allocation concealment, blinding, incomplete outcome data, and selective outcome reporting. RESULTS: Data from the initial phase of five trials involving 477 participants were included. Meta-analysis of the pooled data showed that ultrafiltration was significantly better than diuretic drugs based on 48-h weight loss (Z = 3.72; P < 0.001, weighted mean difference [WMD] = 1.25 kg, 95 % CI 0.59-1.91) and based on 48-h fluid removal (Z = 4.23; P < 0.001, WMD = 1.06 L, 95 % CI 0.57-1.56). Adverse events did not differ significantly between the ultrafiltration and intravenous diuretic treatment groups. LIMITATIONS: There are several limitations to our review, including publication bias and selection bias. Our review included only a few studies involving relatively few participants. CONCLUSIONS: The available evidence suggests that early ultrafiltration is safe and effective for patients with hypervolemic acute heart failure. It allows greater fluid removal and weight loss by 48 h than do intravenous diuretics, with no significant increase in adverse effects.