DiffuseExpand: Expanding dataset for 2D medical image segmentation using diffusion models (preprint)

Dataset expansion can effectively alleviate the problem of data scarcity for medical image segmentation, due to privacy concerns and labeling difficulties. However, existing expansion algorithms still face great challenges due to their inability of guaranteeing the diversity of synthesized images with paired segmentation masks. In recent years, Diffusion Probabilistic Models (DPMs) have shown powerful image synthesis performance, even better than Generative Adversarial Networks. Based on this insight, we propose an approach called DiffuseExpand for expanding datasets for 2D medical image segmentation using DPM, which first samples a variety of masks from Gaussian noise to ensure the diversity, and then synthesizes images to ensure the alignment of images and masks. After that, DiffuseExpand chooses high-quality samples to further enhance the effectiveness of data expansion. Our comparison and ablation experiments on COVID-19 and CGMH Pelvis datasets demonstrate the effectiveness of DiffuseExpand. Our code is released at https://github.com/shaoshitong/DiffuseExpand.

Predicting unseen antibodies’ neutralizability via adaptive graph neural networks

The effects of novel antibodies are hard to predict owing to the complex interactions between antibodies and antigens. Zhang and colleagues use a graph-based method to learn a dynamic representation that allows for predictions of neutralization activity and demonstrate the method by recommending probable antibodies for human immunodeficiency virus, severe acute respiratory syndrome coronavirus 2, influenza and dengue. Most natural and synthetic antibodies are 'unseen'. That is, the demonstration of their neutralization effects with any antigen requires laborious and costly wet-lab experiments. The existing methods that learn antibody representations from known antibody-antigen interactions are unsuitable for unseen antibodies owing to the absence of interaction instances. The DeepAAI method proposed herein learns unseen antibody representations by constructing two adaptive relation graphs among antibodies and antigens and applying Laplacian smoothing between unseen and seen antibodies' representations. Rather than using static protein descriptors, DeepAAI learns representations and relation graphs 'dynamically', optimized towards the downstream tasks of neutralization prediction and 50% inhibition concentration estimation. The performance of DeepAAI is demonstrated on human immunodeficiency virus, severe acute respiratory syndrome coronavirus 2, influenza and dengue. Moreover, the relation graphs have rich interpretability. The antibody relation graph implies similarity in antibody neutralization reactions, and the antigen relation graph indicates the relation among a virus's different variants. We accordingly recommend probable broad-spectrum antibodies against new variants of these viruses.

Changes in the types of patients in Burn Unit: A single center experience

Background Burns are a common concern around the world, with the majority of cases happening in low- and middle-income nations. China is the largest developing country. With the unremitting efforts of domestic colleagues, China has taken the lead in the treatment of burn in the world. With the change of times, we have observed some noteworthy changes in the types of patients that have admitted our Burns and Plastic Surgery, the Affiliated Hospital of Yangzhou University. Methods This retrospective observational study included brought into;all patients reached to our burn unit during 2013-2021. The gathered data were descriptively examined and statistically contrasted with each other year. Results Of 4407 cases admitted to burn unit during 2013-2021, men constituted 56% of such cases, with an average age of 47.3 ± 19.3 years. Moreover, among the patients hospitalized, January and February usually admit fewer than other months. Between 2013 and 2021, both the number of patients admitted to burn unit and the expense of their hospitalization rose yearly. The percentage of burn patients admitted to burn ward of our hospital is decreasing, especially during the period of serious Coronavirus disease (COVID-19) epidemic. We also observed that during the COVID-19 pandemic, patients with superficial masses also dropped off a cliff because of government controls. Conclusion The diseases in the department show the trend of maximizing marginal disciplines, burn surgeons are facing a more complex challenge. Further research addressing the relationship between the change of patient types and economic and social development in burn department will help to foster better pinpoint hospitalization patients need, fine service for hospitalized patients.

Targeting TRIM28-mediated SUMOylation of SARS-CoV-2 Nucleocapsid Protein Inhibits SARS-CoV-2 Infection (preprint)

Viruses, as opportunistic intracellular parasites, hijack the cellular machinery of host cells to support their survival and propagation. Consequently, numerous viral proteins are subjected to host-mediated post-translational modifications. Here, we demonstrate that the SARS-CoV-2 nucleocapsid protein (SARS2-NP) is modified by a small ubiquitin-like modifier (SUMO) on the lysine 65 residue. SARS2-NP SUMOylation is essential for executing efficiently SARS2-NP’s ability in homo-oligomerization, RNA association, liquid-liquid phase separation (LLPS), thereby the innate antiviral immune response is suppressed robustly both in vitro and in vivo . These roles played by SARS2-NP SUMOylation can be achieved through intermolecular association between SUMO conjugation and a newly identified SUMO-interacting motif (SIM) in SARS2-NP. Importantly, the widespread SARS2-NP R203K mutation in SARS-CoV-2 variants gains a novel site of SUMOylation which further increases SARS2-NP’s LLPS and immunosuppression. Notably, we discover that the SUMO E3 ligase TRIM28 is responsible for catalyzing SARS2-NP SUMOylation. An interfering peptide targeting the TRIM28 and SARS2-NP interaction was screened out to block SARS2-NP SUMOylation and LLPS, and consequently inhibit SARS-CoV-2 replication and rescue innate antiviral immunity. Collectively, these data support SARS2-NP SUMOylation as an essential modification for SARS-CoV-2 virulence, and therefore provide a strategy to antagonize SARS-CoV-2.

Research and development of intelligent question answering robot system: a case study of its application in response to COVID-19 epidemic in Jinan city

During the epidemic of coronavirus disease 2019(COVID-19), the local Centers for Disease Control were bombarded with large amounts of questions from the public and the human hotline system was unable to meet the demands. As a result, Jinan Centers for Disease Coatrd developed an "intelligent question answering robot system" to cope with this situation. This paper introduces the design of the robot system and construction and classification of the knowledge base, and evaluates its application effects. The robot system can greatly reduce pressure on the human hotline, actively record and analyze users' demands, and improve the quality and efficiency of Centers for Disease Coatrd consultation service. It is a valuable and growable operating mode of consultation service, which can provide reference for the information service in future public health events.

Neutrophil Infiltration Induces Myocardial Injury In COVID-19 Post-Mortem Cases (preprint)

Objectives: The pathological features of severe cardiac injury induced by COVID-19 and relevant clinical features is unknown.Methods: This autopsy cohort study, including hearts from 26 deceased patients hospitalized in intensive care unit due to COVID-19, was conducted at four sites in Wuhan, China. Cases were divided into neutrophil-infiltration group and no-neutrophil group according to histopathological identification of neutrophilic infiltrates or not.Results: Among 26 cases, four cases had active myocarditis with histopathological examination. All cases with myocarditis accompanied with extensive neutrophil infiltration, while cases without myocarditis did not. Detection rates of interleukin-6 (100% vs 4.6%) and tumor necrosis factor-a (100% vs 31.8%) in neutrophil-infiltration group were significantly higher compared to no-neutrophil group (p<0.05 for both). At admission, patients with neutrophil infiltration in myocardium had significantly higher baseline values of aspartate aminotransferase, D dimer and high-sensitivity C reactive protein compared to other 22 patients (p<0.05 for all). During hospitalization, patients with neutrophil infiltration had a significantly higher maximum of creatine kinase (CK)-MB (median 280.0 vs 38.7IU/L, p=0.04), and a quantitatively higher top Troponin I (median 1.112 vs 0.220ng/ml, p=0.56) than patients without neutrophil infiltration. Conclusions: In hearts from deceased patients with severe COVID-19, active myocarditis was commonly infiltrated with neutrophils. Cases with neutrophil-infiltrated myocarditis had a series of severe abnormal laboratory tests at admission, and a high maximum of CK-MB during hospitalization. Role of neutrophil on severe heart injury and even systemic condition in COVID-19 should be emphasized.

Neutrophil Infiltration Induces Myocardial Injury in COVID-19 Post-Mortem Cases (preprint)

Background: The pathological features of severe cardiac injury induced by COVID-19 and relevant clinical features is unknown.Methods: This autopsy cohort study, including hearts from 26 deceased patients hospitalized in intensive care unit due to COVID-19, was conducted at four sites in Wuhan, China. Cases were divided into neutrophil-infiltration group and no-neutrophil group according to histopathological identification of neutrophilic infiltrates or not.Findings: Among 26 cases, four cases had active myocarditis with histopathological examination. All cases with myocarditis accompanied with extensive neutrophil infiltration, while cases without myocarditis did not. Detection rates of interleukin-6 (100% vs 4.6%) and tumor necrosis factor-α (100% vs 31.8%) in neutrophil-infiltration group were significantly higher compared to no-neutrophil group (p<0.05 for both). At admission, patients with neutrophil infiltration in myocardium had significantly higher baseline values of aspartate aminotransferase, D dimer and high-sensitivity C reactive protein compared to other 22 patients (p<0.05 for all). During hospitalization, patients with neutrophil infiltration had a significantly higher maximum of creatine kinase (CK)-MB (median 280.0 vs 38.7IU/L, p=0.04), and a quantitatively higher top Troponin I (median 1.112 vs 0.220ng/ml, p=0.56) than patients without neutrophil infiltration.Interpretation: In hearts from deceased patients with severe COVID-19 , active myocarditis was commonly infiltrated with neutrophils. Cases with neutrophil-infiltrated myocarditis had a series of severe abnormal laboratory tests at admission, and a high maximum of CK-MB during hospitalization. Role of neutrophil on severe heart injury and even systemic condition in COVID-19 should be emphasized.Funding Information: : Emergency Key Program of Guangzhou Laboratory, Grant No. EKPG21-32. Declaration of Interests: None exist.Ethics Approval Statement: Full autopsy was performed after patient death with the approval of the ethics committees and written consent of patient relatives in accordance with regulations issued by the National Health Commission of China and the Helsinki Declaration.

SARS-CoV-2 RdRp is a versatile enzyme with proofreading activity and ability to incorporate NHC into RNA by using diphosphate form molnupiravir as a substrate (preprint)

The coronavirus disease 2019 (COVID-19) has been ravaging throughout the world for almost two years and has severely impaired both human health and the economy. The causative agent, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) employs the viral RNA-dependent RNA polymerase (RdRp) complex for genome replication and transcription, making RdRp an appealing target for antiviral drug development. Although the structure of the RdRp complex has been determined, the function of RdRp has not been fully characterized. Here we reveal that in addition to RNA dependent RNA polymerase activity, RdRp also shows exoribonuclease activity and consequently proofreading activity. We observed that RdRp and nsp14-ExoN, when combined, exhibit higher proofreading activity compared to RdRp alone. Moreover, RdRp can recognize and utilize nucleoside diphosphate (NDP) as substrate to synthesize RNA and can also incorporate {beta}-d-N4-hydroxycytidine (NHC) into RNA while using diphosphate form molnupiravir as substrate.

Remove of triclosan from aqueous solutions by nanoflower MnO2: Insight into the mechanism of oxidation and adsorption

Triclosan (TCS) has been proved to have a harmful effect on human health and ecological environment, especially during the COVID-19 epidemic, when plentiful antibacterial hand sanitizers were discharged. Manganese dioxide (MnO2) showed a good effect on the removal of TCS. The morphology of MnO2 was regulated in this study to increase the active sites for removing TCS and improve the removal effect. The results showed that nanoflower ε-MnO2 exhibited best removal efficiency due to its high oxygen vacancy, high Mn3+ content, easily released lattice oxygen and unique tunnel structure which make its Mn-O bond easier to activate. Further study of the mechanism revealed that the process of removing TCS by MnO2 was the first adsorption and then oxidation process and the detailed reaction process was clarified. 3-chlorophenol and 2,4-dichlorophenol were proved to be their oxidative product. Additionally, it was verified that oxidation dominated in the removal of TCS by MnO2 rather than adsorption through Density functional theory (DFT) calculations analysis. It is determined that nanoflower MnO2 was a promising material for removing TCS.

Machine learning discovery of distinguishing laboratory features for severity classification of COVID-19 patients

Abstract The exponential spread of COVID-19 worldwide is evident, with devastating outbreaks primarily in the United States, Spain, Italy, the United Kingdom, France, Germany, Turkey and Russia. As of 1 May 2020, a total of 3,308,386 confirmed cases have been reported worldwide, with an accumulative mortality of 233,093. Due to the complexity and uncertainty of the pathology of COVID-19, it is not easy for front-line doctors to categorise severity levels of clinical COVID-19 that are general and severe/critical cases, with consistency. The more than 300 laboratory features, coupled with underlying disease, all combine to complicate proper and rapid patient diagnosis. However, such screening is necessary for early triage, diagnosis, assignment of appropriate level of care facility, and institution of timely intervention. A machine learning analysis was carried out with confirmed COVID-19 patient data from 10 January to 18 February 2020, who were admitted to Tongji Hospital, in Wuhan, China. A softmax neural network-based machine learning model was established to categorise patient severity levels. According to the analysis of 2662 cases using clinical and laboratory data, the present model can be used to reveal the top 30 of more than 300 laboratory features, yielding 86.30% blind test accuracy, 0.8195 F1-score, and 100% consistency using a two-way patient classification of severe/critical to general. For severe/critical cases, F1-score is 0.9081 (i.e. recall is 0.9050, and precision is 0.9113). This model for classification can be accomplished at a mini-second-level computational cost (in contrast to minute-level manual). Based on available COVID-19 patient diagnosis and therapy, an artificial intelligence model paradigm can help doctors quickly classify patients with a high degree of accuracy and 100% consistency to significantly improve diagnostic and classification efficiency. The discovered top 30 laboratory features can be used for greater differentiation to serve as an essential supplement to current guidelines, thus creating a more comprehensive assessment of COVID-19 cases during the early stages of infection. Such early differentiation will help the assignment of the appropriate level of care for individual patients.

Multiscale Attention Guided Network for COVID-19 Diagnosis Using Chest X-ray Images (preprint)

Coronavirus disease 2019 (COVID-19) is one of the most destructive pandemic after millennium, forcing the world to tackle a health crisis. Automated lung infections classification using chest X-ray (CXR) images could strengthen diagnostic capability when handling COVID-19. However, classifying COVID-19 from pneumonia cases using CXR image is a difficult task because of shared spatial characteristics, high feature variation and contrast diversity between cases. Moreover, massive data collection is impractical for a newly emerged disease, which limited the performance of data thirsty deep learning models. To address these challenges, Multiscale Attention Guided deep network with Soft Distance regularization (MAG-SD) is proposed to automatically classify COVID-19 from pneumonia CXR images. In MAG-SD, MA-Net is used to produce prediction vector and attention from multiscale feature maps. To improve the robustness of trained model and relieve the shortage of training data, attention guided augmentations along with a soft distance regularization are posed, which aims at generating meaningful augmentations and reduce noise. Our multiscale attention model achieves better classification performance on our pneumonia CXR image dataset. Plentiful experiments are proposed for MAG-SD which demonstrates its unique advantage in pneumonia classification over cutting-edge models. The code is available at https://github.com/JasonLeeGHub/MAG-SD.

An Analysis of the Influence of Novel Coronavirus Pneumonia Epidemic on Middle School Students' Mental Health (preprint)

Objective: To explore the psychological health status of middle school students in China during the recent epidemic outbreak of “Novel Coronavirus Pneumonia” (NCP) and its influencing factors, and to provide a basis and suggestions for the adjustment of mental health of middle school students during the epidemic. Methods: A total of 2,007 middle school students were randomly selected nationwide to fill out a questionnaire. The questionnaire included basic personal and family information, life status and family relations during the epidemic, attitudes and pressures in the face of the epidemic, created by Chinese professor Zung. The anxiety self-assessment scale (SAS) compiled by SPSS statistical analysis software was used for data analysis to explore the changes in mental health during the epidemic. Results: Of the 2,007 valid questionnaires, according to the data using the self-rating anxiety scale, 424 students had anxiety, accounting for 21.1% of the total number; anxiety level of boys was higher than that of girls, and the anxiety level increased as the grade level increased. The student's personality, grade level, problem-solving methods, awareness and attention to the epidemic, parents' educational background, marital status, and family economic status all had an impact on the student's psychological status during the epidemic.Conclusion: The Novel Coronavirus Pneumonia epidemic has a great impact on the mental health of some middle school students. Family, school and social departments should pay attention to this phenomenon, care for their mental health, and take reasonable measures to actively intervene.

An Antioxidant Enzyme Therapeutic for COVID-19 (preprint)

The COVID-19 pandemic has taken a significant toll on people worldwide, and there are currently no specific antivirus drugs or vaccines. We report herein a therapeutic based on catalase, an antioxidant enzyme that can effectively breakdown hydrogen peroxide and minimize the downstream reactive oxygen species, which are excessively produced resulting from the infection and inflammatory process. Catalase assists to regulate production of cytokines, protect oxidative injury, and repress replication of SARS-CoV-2, as demonstrated in human leukocytes and alveolar epithelial cells, and rhesus macaques, without noticeable toxicity. Such a therapeutic can be readily manufactured at low cost as a potential treatment for COVID-19.

AXL Promotes SARS-CoV-2 Infection of Pulmonary and Bronchial Epithelial Cells (preprint)

The current coronavirus disease 2019 (COVID-19) pandemic presents a global public health challenge. The viral pathogen responsible, Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), binds to a host receptor ACE2 through its spike (S) glycoprotein, which mediates membrane fusion and virus entry. Although the role of ACE2 as a receptor for SARS-CoV-2 is clear, studies have shown that ACE2 expression across different human tissues is extremely low, especially in pulmonary and bronchial cells. Thus, other host receptors and/or co-receptors that promote the entry of SARS-CoV-2 into cells of the respiratory system might exist. In this study, we have identified tyrosine-protein kinase receptor UFO (AXL), specifically interacts with SARS-CoV-2 S on the host cell membrane. When overexpressed in cells that do not highly express either AXL or ACE2, AXL promotes virus entry as efficiently as ACE2. Strikingly, deleting AXL, but not ACE2, significantly reduces infection of pulmonary cells by the SARS-CoV-2 virus pseudotype. Soluble human recombinant AXL, but not ACE2, blocks SARS-CoV-2 virus pseudotype infection in pulmonary cells. Taken together, our findings suggest AXL may play an important role in promoting SARS-CoV-2 infection of the human respiratory system and is a potential target in future clinical intervention strategies.

Clinical features and the maternal and neonatal outcomes of pregnant women with coronavirus disease 2019 (preprint)

BACKGROUND There is little information about the coronavirus disease 2019 (Covid-19) during pregnancy. This study aimed to determine the clinical features and the maternal and neonatal outcomes of pregnant women with Covid-19. METHODS In this retrospective analysis from five hospitals, we included pregnant women with Covid-19 from January 1 to February 20, 2020. The primary composite endpoints were admission to an intensive care unit (ICU), the use of mechanical ventilation, or death. Secondary endpoints included the clinical severity of Covid-19, neonatal mortality, admission to neonatal intensive care unit (NICU), and the incidence of acute respiratory distress syndrome (ARDS) of pregnant women and newborns. RESULTS Thirty-three pregnant women with Covid-19 and 28 newborns were identified. One (3%) pregnant woman needed the use of mechanical ventilation. No pregnant women admitted to the ICU. There were no moralities among pregnant women or newborns. The percentages of pregnant women with mild, moderate, and severe symptoms were 13 (39.4%),19(57.6%), and 1(3%). One (3.6%) newborn developed ARDS and was admitted to the NICU. The rate of perinatal transmission of SARS-CoV-2 was 3.6%. CONCLUSIONS This report suggests that pregnant women are not at increased risk for severe illness or mortality with Covid-19 compared with the general population. The SARS-CoV-2 infection during pregnancy might not be associated with as adverse obstetrical and neonatal outcomes that are seen with the severe acute respiratory syndrome coronavirus (SARS-CoV) and Middle East respiratory syndrome coronavirus (MERS-CoV) infection during pregnancy. (Funded by the National Key Research and Development Program.)

A deep learning algorithm using CT images to screen for Corona Virus Disease (COVID-19) (preprint)

BackgroundThe outbreak of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-COV-2) has caused more than 2.5 million cases of Corona Virus Disease (COVID-19) in the world so far, with that number continuing to grow. To control the spread of the disease, screening large numbers of suspected cases for appropriate quarantine and treatment is a priority. Pathogenic laboratory testing is the gold standard but is time-consuming with significant false negative results. Therefore, alternative diagnostic methods are urgently needed to combat the disease. Based on COVID-19 radiographical changes in CT images, we hypothesized that Artificial Intelligences deep learning methods might be able to extract COVID-19s specific graphical features and provide a clinical diagnosis ahead of the pathogenic test, thus saving critical time for disease control. Methods and FindingsWe collected 1,065 CT images of pathogen-confirmed COVID-19 cases (325 images) along with those previously diagnosed with typical viral pneumonia (740 images). We modified the Inception transfer-learning model to establish the algorithm, followed by internal and external validation. The internal validation achieved a total accuracy of 89.5% with specificity of 0.88 and sensitivity of 0.87. The external testing dataset showed a total accuracy of 79.3% with specificity of 0.83 and sensitivity of 0.67. In addition, in 54 COVID-19 images that first two nucleic acid test results were negative, 46 were predicted as COVID-19 positive by the algorithm, with the accuracy of 85.2%. ConclusionThese results demonstrate the proof-of-principle for using artificial intelligence to extract radiological features for timely and accurate COVID-19 diagnosis. Author summaryTo control the spread of the COVID-19, screening large numbers of suspected cases for appropriate quarantine and treatment measures is a priority. Pathogenic laboratory testing is the gold standard but is time-consuming with significant false negative results. Therefore, alternative diagnostic methods are urgently needed to combat the disease. We hypothesized that Artificial Intelligences deep learning methods might be able to extract COVID-19s specific graphical features and provide a clinical diagnosis ahead of the pathogenic test, thus saving critical time. We collected 1,065 CT images of pathogen-confirmed COVID-19 cases along with those previously diagnosed with typical viral pneumonia. We modified the Inception transfer-learning model to establish the algorithm. The internal validation achieved a total accuracy of 89.5% with specificity of 0.88 and sensitivity of 0.87. The external testing dataset showed a total accuracy of 79.3% with specificity of 0.83 and sensitivity of 0.67. In addition, in 54 COVID-19 images that first two nucleic acid test results were negative, 46 were predicted as COVID-19 positive by the algorithm, with the accuracy of 85.2%. Our study represents the first study to apply artificial intelligence to CT images for effectively screening for COVID-19.