A model to predict the risk of mortality in severely ill COVID-19 patients.

BACKGROUND: To investigate and select the useful prognostic parameters to develop and validate a model to predict the mortality risk for severely and critically ill patients with the coronavirus disease 2019 (COVID-19). METHODS: We established a retrospective cohort of patients with laboratory-confirmed COVID-19 (≥18 years old) from two tertiary hospitals: the People's Hospital of Wuhan University and Leishenshan Hospital between February 16, 2020, and April 14, 2020. The diagnosis of the cases was confirmed according to the WHO interim guidance. The data of consecutive severely and critically ill patients with COVID-19 admitted to these hospitals were analyzed. A total of 566 patients from the People's Hospital of Wuhan University were included in the training cohort and 436 patients from Leishenshan Hospital were included in the validation cohort. The least absolute shrinkage and selection operator (LASSO) and multivariable logistic regression were used to select the variables and build the mortality risk prediction model. RESULTS: The prediction model was presented as a nomograph and developed based on identified predictors, including age, chronic lung disease, C-reactive protein (CRP), D-dimer levels, neutrophil-to-lymphocyte ratio (NLR), creatinine, and total bilirubin. In the training cohort, the model displayed good discrimination with an AUC of 0.912 [95% confidence interval (CI): 0.884-0.940] and good calibration (intercept = 0; slope = 1). In the validation cohort, the model had an AUC of 0.922 [95% confidence interval (CI): 0.891-0.953] and a good calibration (intercept = 0.056; slope = 1.161). The decision curve analysis (DCA) demonstrated that the nomogram was clinically useful. CONCLUSION: A risk score for severely and critically ill COVID-19 patients' mortality was developed and externally validated. This model can help clinicians to identify individual patients at a high mortality risk.

Prevalence and trend of low vision among primary and secondary school students in Wuhan from 2019 to 2020

Objective: To provide a large-scale assessment the prevalence of poor vision in 2020 among children and adolescents in Wuhan City, Hubei province and to provide basis for healthy vision promotion.

Serum Uric Acid Concentrations and Risk of Adverse Outcomes in Patients With COVID-19.

Background: Although hyperuricemia frequently associates with respiratory diseases, patients with severe coronavirus disease 2019 (COVID-19) and severe acute respiratory syndrome (SARS) can show marked hypouricemia. Previous studies on the association of serum uric acid with risk of adverse outcomes related to COVID-19 have produced contradictory results. The precise relationship between admission serum uric acid and adverse outcomes in hospitalized patients is unknown. Methods: Data of patients affected by laboratory-confirmed COVID-19 and admitted to Leishenshan Hospital were retrospectively analyzed. The primary outcome was composite and comprised events, such as intensive care unit (ICU) admission, mechanical ventilation, or mortality. Logistic regression analysis was performed to explore the association between serum concentrations of uric acid and the composite outcome, as well as each of its components. To determine the association between serum uric acid and in-hospital adverse outcomes, serum uric acid was also categorized by restricted cubic spline, and the 95% confidence interval (CI) was used to estimate odds ratios (OR). Results: The study cohort included 1854 patients (mean age, 58 years; 52% women). The overall mean ± SD of serum levels of uric acid was 308 ± 96 µmol/L. Among them, 95 patients were admitted to ICU, 75 patients received mechanical ventilation, and 38 died. In total, 114 patients reached composite end-points (have either ICU admission, mechanical ventilation or death) during hospitalization. Compared with a reference group with estimated baseline serum uric acid of 279-422 µmol/L, serum uric acid values ≥ 423 µmol/L were associated with an increased risk of composite outcome (OR, 2.60; 95% CI, 1.07- 6.29) and mechanical ventilation (OR, 3.01; 95% CI, 1.06- 8.51). Serum uric acid ≤ 278 µmol/L was associated with an increased risk of the composite outcome (OR, 2.07; 95% CI, 1.18- 3.65), ICU admission (OR, 2.18; 95% CI, 1.17- 4.05]), and mechanical ventilation (OR, 2.13; 95% CI, 1.06- 4.28), as assessed by multivariate analysis. Conclusions: This study shows that the association between admission serum uric acid and composite outcome of COVID-19 patients was U-shaped. In particular, we found that compared with baseline serum uric acid levels of 279-422 µmol/L, values ≥ 423 µmol/L were associated with an increased risk of composite outcome and mechanical ventilation, whereas levels ≤ 278 µmol/L associated with increased risk of composite outcome, ICU admission and mechanical ventilation.

CheXLocNet: Automatic localization of pneumothorax in chest radiographs using deep convolutional neural networks.

BACKGROUND: Pneumothorax can lead to a life-threatening emergency. The experienced radiologists can offer precise diagnosis according to the chest radiographs. The localization of the pneumothorax lesions will help to quickly diagnose, which will be benefit for the patients in the underdevelopment areas lack of the experienced radiologists. In recent years, with the development of large neural network architectures and medical imaging datasets, deep learning methods have become a methodology of choice for analyzing medical images. The objective of this study was to the construct convolutional neural networks to localize the pneumothorax lesions in chest radiographs. METHODS AND FINDINGS: We developed a convolutional neural network, called CheXLocNet, for the segmentation of pneumothorax lesions. The SIIM-ACR Pneumothorax Segmentation dataset was used to train and validate CheXLocNets. The training dataset contained 2079 radiographs with the annotated lesion areas. We trained six CheXLocNets with various hyperparameters. Another 300 annotated radiographs were used to select parameters of these CheXLocNets as the validation set. We determined the optimal parameters by the AP50 (average precision at the intersection over union (IoU) equal to 0.50), a segmentation evaluation metric used by several well-known competitions. Then CheXLocNets were evaluated by a test set (1082 normal radiographs and 290 disease radiographs), based on the classification metrics: area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and positive predictive value (PPV); segmentation metrics: IoU and Dice score. For the classification, CheXLocNet with best sensitivity produced an AUC of 0.87, sensitivity of 0.78 (95% CI 0.73-0.83), and specificity of 0.78 (95% CI 0.76-0.81). CheXLocNet with best specificity produced an AUC of 0.79, sensitivity of 0.46 (95% CI 0.40-0.52), and specificity of 0.92 (95% CI 0.90-0.94). For the segmentation, CheXLocNet with best sensitivity produced an IoU of 0.69 and Dice score of 0.72. CheXLocNet with best specificity produced an IoU of 0.77 and Dice score of 0.79. We combined them to form an ensemble CheXLocNet. The ensemble CheXLocNet produced an IoU of 0.81 and Dice score of 0.82. Our CheXLocNet succeeded in automatically detecting pneumothorax lesions, without any human guidance. CONCLUSIONS: In this study, we proposed a deep learning network, called, CheXLocNet, for the automatic segmentation of chest radiographs to detect pneumothorax. Our CheXLocNets generated accurate classification results and high-quality segmentation masks for the pneumothorax at the same time. This technology has the potential to improve healthcare delivery and increase access to chest radiograph expertise for the detection of diseases. Furthermore, the segmentation results can offer comprehensive geometric information of lesions, which can benefit monitoring the sequential development of lesions with high accuracy. Thus, CheXLocNets can be further extended to be a reliable clinical decision support tool. Although we used transfer learning in training CheXLocNet, the parameters of CheXLocNet was still large for the radiograph dataset. Further work is necessary to prune CheXLocNet suitable for the radiograph dataset.

A Mouse Model of SARS-CoV-2 Infection and Pathogenesis.

Since December 2019, a novel coronavirus SARS-CoV-2 has emerged and rapidly spread throughout the world, resulting in a global public health emergency. The lack of vaccine and antivirals has brought an urgent need for an animal model. Human angiotensin-converting enzyme II (ACE2) has been identified as a functional receptor for SARS-CoV-2. In this study, we generated a mouse model expressing human ACE2 (hACE2) by using CRISPR/Cas9 knockin technology. In comparison with wild-type C57BL/6 mice, both young and aged hACE2 mice sustained high viral loads in lung, trachea, and brain upon intranasal infection. Although fatalities were not observed, interstitial pneumonia and elevated cytokines were seen in SARS-CoV-2 infected-aged hACE2 mice. Interestingly, intragastric inoculation of SARS-CoV-2 was seen to cause productive infection and lead to pulmonary pathological changes in hACE2 mice. Overall, this animal model described here provides a useful tool for studying SARS-CoV-2 transmission and pathogenesis and evaluating COVID-19 vaccines and therapeutics.