Unsupervised Domain Adaptation for COVID-19 Information Service with Contrastive Adversarial Domain Mixup

In the real-world application of COVID-19 misinformation detection, a fundamental challenge is the lack of the labeled COVID data to enable supervised end-to-end training of the models, especially at the early stage of the pandemic. To address this challenge, we propose an unsupervised domain adaptation framework using contrastive learning and adversarial domain mixup to transfer the knowledge from an existing source data domain to the target COVID-19 data domain. In particular, to bridge the gap between the source domain and the target domain, our method reduces a radial basis function (RBF) based discrepancy between these two domains. Moreover, we leverage the power of domain adversarial examples to establish an intermediate domain mixup, where the latent representations of the input text from both domains could be mixed during the training process. Extensive experiments on multiple real-world datasets suggest that our method can effectively adapt misinformation detection systems to the unseen COVID-19 target domain with significant improvements compared to the state-of-the-art baselines. © 2022 IEEE.

Dynamic dependencies and return connectedness among stock, gold and Bitcoin markets: Evidence from South Asia and China

Research background:In order to examine market uncertainty, the paper depicts broad patterns of risk and systematic exposure to global equity market shocks for the major South Asian and Chinese equity markets, as well as for specific assets (gold and Bitcoin). Purpose of the article: The purpose of this paper is to investigate the dynamic correlation among the major South Asian equity markets (India and Pakistan), the Chinese equity mar-kets, the MSCI developed markets, Bitcoin, and gold markets. Methods: While applying the GARCH-Vine-Copula model and the TVP-VAR Connectedness approach, major patterns of dependency and interconnectedness between these markets are investigated. Findings & value added: We find that risk shocks from developed equity markets are critical in these dynamic links. A net return spillover from Bitcoin to the Chinese and Pakistani stock markets throughout the sample period is reported. Interestingly, gold can be applied to hedge and diversify positions in China and major South Asian markets, particularly following the COVID-19 outbreak. Our paper presents three main original add valued: (1) This paper adds global factors to the targeted study of risk transmission among South Asian and Chinese stock markets for the first time. (2)The assets of Bitcoin and gold were added to the study of risk transmission among South Asian and Chinese stock markets for the first time, enabling the research in this paper to observe the non-linear link among the South Asian and Chinese stock markets with them. (3) Our research adds to these lines of inquiry by giving empirical evidence on how COVID-19 altered the dependent structure and return spillover dynamics of Bitcoin, gold and South Asian and Chinese stock markets for the first time. Our results have critical implications for investors and policymakers to effectively understand the nature of market forces and develop risk-averse strategies. © Instytut Badań Gospodarczych / Institute of Economic Research (Poland).

A Knowledge-driven Domain Adaptive Approach to Early Misinformation Detection in an Emergent Health Domain on Social Media

This paper focuses on an important problem of early misinformation detection in an emergent health domain on social media. Current misinformation detection solutions often suffer from the lack of resources (e.g., labeled datasets, sufficient medical knowledge) in the emerging health domain to accurately identify online misinformation at an early stage. To address such a limitation, we develop a knowledge-driven domain adaptive approach that explores a good set of annotated data and reliable knowledge facts in a source domain (e.g., COVID-19) to learn the domain-invariant features that can be adapted to detect misinformation in the emergent target domain with little ground truth labels (e.g., Monkeypox). Two critical challenges exist in developing our solution: i) how to leverage the noisy knowledge facts in the source domain to obtain the medical knowledge related to the target domain? ii) How to adapt the domain discrepancy between the source and target domains to accurately assess the truthfulness of the social media posts in the target domain? To address the above challenges, we develop KAdapt, a knowledge-driven domain adaptive early misinformation detection framework that explicitly extracts rel-evant knowledge facts from the source domain and jointly learns the domain-invariant representation of the social media posts and their relevant knowledge facts to accurately identify misleading posts in the target domain. Evaluation results on five real-world datasets demonstrate that KAdapt significantly outperforms state-of-the-art baselines in terms of accurately detecting misleading Monkeypox posts on social media. © 2022 IEEE.

Impact of Environmental Factors on Short-Term Eye Strain Relief during COVID-19 Quarantine: A Pilot Study

Some policies implemented during the pandemic extended the time that students spend on electronic devices, increasing the risk of physical and eye strain. However, the role of different environments on eye strain recovery has not been determined. We recruited 20 undergraduate students (10 males and 10 females) from a university in eastern China and explored the restoration effects of their eye strain in different types of spaces (wayside greenspace, a playground, a square, and woodland) on campus through scale measurements. The results showed that the eye strain of the students accumulated by 15 min of e-learning was significantly relieved after 10 min of greenspace exposure compared to the indoor environment, and the recovery effect varied depending on the type of landscape. The effect of eye strain relief was found to be positively correlated with temperature, wind speed, visible sky ratio, canopy density, tree density, and solar radiation intensity, while it was negatively correlated with relative humidity. These findings enrich the research on the restoration benefits of greenspaces and provide a basis for predicting the effect of different environments on the relief of eye strain. © 2022 by the authors.

Retrospective Analysis of the Duration of Hospitalization of Imported Patients Infected with SARS-CoV-2 in Guangzhou, 2021

The pandemic of coronavirus disease 2019 is "not over,"in fact, the "dynamic clearing"policy for SARS-CoV-2 control and prevention in China has been firmly enforced. This study aimed to analyze the clinical symptoms and dynamic viral RNA changes in 2021 at Guangzhou Eighth People's Hospital. This study showed that 31.4% of the patients (695/2212) tested negative for viral RNA from admission to the final release from quarantine. Of all negative cases, 86.5% (601/695) remained in the hospital for no more than 5 days and were asymptomatic or mild. Among the remaining 402 patients who stayed for no more than 5 days, 76.4% (307/402) were viral RNA retest positive during the isolation stage. However, 96.4% of the peak viral RNA (296/307) was over Ct = 33 cycles during the isolation stage. © Wolters Kluwer Health, Inc. All rights reserved.

A grid management system for COVID-19 antigen detection based on image recognition

Objective: To develop a SARS-CoV-2 antigen detection management system for Chinese residents under community grid management, which is supported by "health information technology" and "neural network image recognition", so as to give full play to the advantages of "grid management". This system is applied to the normalized prevention and control of COVID-19 epidemic. Method(s): The model of image recognition algorithm was built based on deep learning and convolution neural network (CNN) artificial intelligence algorithm. The improved Canny edge detection algorithm was used to monitor and locate the image edge, and then the image segmentation and judgment value calculation were completed according to projection method. The system construction was completed combing with the grid number design. Result(s): The proposed method had been tested and showed the accuracy of the algorithm. With a certain robustness, the algorithm error was proved to be small. Based on the image recognition algorithm model, the development of SARS-CoV-2 antigen detection management system covering user login, paper-strip test image upload, paper-strip test management, grid management, grid warning and regional traffic management was completed. Conclusion(s): Antigen detection is an important supplementary means of COVID-19 epidemic prevention and control in the new stage. The SARS-CoV-2 antigen detection management system for Chinese residents under community grid managemen based on image recognition enables mobile communication devices to recognize the image of SARS-CoV-2 antigen detection results, which is helpful to form a grid management mode for the epidemic and improve the management framework of epidemic monitoring, detection, early warning and prevention and control.Copyright © 2023 The Authors

Fast and direct identification of SARS-CoV-2 variants via 2D InSe field-effect transistors

As the COVID-19 pandemic evolves and new variants emerge, the development of more efficient identification approaches of variants is urgent to prevent continuous outbreaks of SARS-CoV-2. Field-effect transistors (FETs) with two-dimensional (2D) materials are viable platforms for the detection of virus nucleic acids (NAs) but cannot yet provide accurate information on NA variations. Herein, 2D Indium selenide (InSe) FETs were used to identify SARS-CoV-2 variants. The device's mobility and stability were ensured by atomic layer deposition (ALD) of Al2O3. The resulting FETs exhibited sub-fM detection limits ranging from 10(-14) M to 10(-)(8) M. The recognition of single-nucleotide variations was achieved within 15 min to enable the fast and direct identification of two core mutations (L452R, R203M) in Delta genomes (p < .01). Such capability originated from the trap states in oxidized InSe (InSe1-xOx) after ALD, resulting in traps-involved carrier transport responsive to the negative charges of NAs. In sum, the proposed approach might highly provide epidemiological information for timely surveillance of the COVID pandemic.

Energy-Efficient Intelligent Pulmonary Auscultation for Post COVID-19 Era Wearable Monitoring Enabled by Two-Stage Hybrid Neural Network

This paper proposes an energy-efficient intelligent pulmonary auscultation system for post COVID-19 era wearable monitoring. This system consists of a tightly coupled two-stage hybrid neural network (TC-TSHNN) model and a corresponding multi-task training paradigm to improve prediction accuracy and generalization ability based on the fact that the number of COVID-19 patients is far less than that of normal people. At the first stage, two-category coarse classification is performed to identify normal and abnormal lung sounds. If the lung sound is abnormal, the second stage would be triggered to perform a four-category fine-grained classification. Besides, discrete wavelet transform is utilized for feature extraction, denoising and data reduction. In addition, advanced lightweight convolutional neural networks are used to reduce the model's computation and improve the model's performance. The hybrid network model can achieve 92% computation reduction and energy saving compared with a direct four-category classification when the input lung sound is normal, which is the majority of cases. Experiment results with inter-patient classification on the COVID-19 lung sound dataset from Tongji Hospital in Wuhan City and the ICBHI'17 dataset show that the proposed TC-TSHNN model can significantly reduce power consumption while maintaining competitive performance against the state-of-the-art work. © 2022 IEEE.

Contrastive Domain Adaptation for Early Misinformation Detection: A Case Study on COVID-19

Despite recent progress in improving the performance of misinformation detection systems, classifying misinformation in an unseen domain remains an elusive challenge. To address this issue, a common approach is to introduce a domain critic and encourage domain-invariant input features. However, early misinformation often demonstrates both conditional and label shifts against existing misinformation data (e.g., class imbalance in COVID-19 datasets), rendering such methods less effective for detecting early misinformation. In this paper, we propose contrastive adaptation network for early misinformation detection (CANMD). Specifically, we leverage pseudo labeling to generate high-confidence target examples for joint training with source data. We additionally design a label correction component to estimate and correct the label shifts (i.e., class priors) between the source and target domains. Moreover, a contrastive adaptation loss is integrated in the objective function to reduce the intra-class discrepancy and enlarge the inter-class discrepancy. As such, the adapted model learns corrected class priors and an invariant conditional distribution across both domains for improved estimation of the target data distribution. To demonstrate the effectiveness of the proposed CANMD, we study the case of COVID-19 early misinformation detection and perform extensive experiments using multiple real-world datasets. The results suggest that CANMD can effectively adapt misinformation detection systems to the unseen COVID-19 target domain with significant improvements compared to the state-of-the-art baselines. © 2022 ACM.

High-frequency volatility connectedness and time-frequency correlation among Chinese stock and major commodity markets around COVID-19

This study examines the connectedness and time-frequency correlation of price volatility across the Chinese stock market and major commodity markets. This paper applies a DCC-GARCH-based volatility connectedness model and the cross-wavelet transform to examine the transmission of risk patterns in these markets before and during the COVID-19 outbreak, as well as the leading lag relationship and synergistic movements between different time domains. First, the findings of the DCC-GARCH connectedness model show dynamic total spillovers are stronger after the COVID-19 outbreak. Chinese stocks and corn have been net spillovers in the system throughout the sample period, but the Chinese market plays the role of a net receiver of volatility relative to other markets (net pairwise directional connectedness) in the system as a whole. In terms of wavelet results, there is some connection to the connectedness results, with all commodity markets, except soybeans and wheat, showing significant dependence on Chinese equities in the medium/long term following the COVID-19 outbreak. Secondly, the medium-to long-term frequency of the crude oil market and copper market are highly dependent on the Chinese stock market, especially after the COVID-19 outbreak. Meanwhile, the copper market is the main source of risk for the Chinese stock market, while the wheat market sends the least shocks to the Chinese stock market. The findings of this paper will have a direct impact on a number of important decisions made by investors and policymakers. © Hongjun Zeng, Ran Lu, 2022.

Psychological and Emotional Responses during Different Stages of the COVID-19 Pandemic Based on a Survey of a Mental Health Hotline

Background: The coronavirus (COVID-19) outbreak in 2019 triggered psychological and emotional responses. This research investigates the psychological status and emotional problems of those who sought psychological assistance during the epidemic period by calling a mental health hotline. Methods: This study aims to combine qualitative and quantitative research. Descriptive analysis was used for undertaking qualitative research. We analyzed the data from group 1 (n = 706), in which the people used the mental health hotline from 25 January 2020 to 23 June 2020. A self-designed questionnaire was developed in accordance with the classification and summarized items from group 1’s psychological problems and emotional status. To implement the quantitative research, we conducted a cross-sectional descriptive survey and used the self-compiled scale and HADS to investigate group 2 (n = 553) from May 2020 to June 2020. Results: Descriptive statistics and comparative analysis revealed that: ①Visitors mainly reported behavior, emotional, family relationship problems and sleep disorders. ② Anxiety, comorbidities, sleep disorders and coping problems were the most frequently reported problems. ③ There were significant differences in the number of visitors experiencing various problems or exhibiting harmful behaviors (sorrow, worry, fear, depression, sleep disorders, self-harm or suicide, and coping problems, anxiety, hypochondria, and comorbidity) in the four stages of the epidemic. ④ More than a quarter of participants still suffered from anxiety or depression in the later stages of the epidemic. Conclusion: Different problems manifested at different stages of the epidemic, and psychological interventions and assistance should be tailored to reflect this. © 2022, Tech Science Press. All rights reserved.

The efficacy and acceptance of online learning vs. offline learning in medical student education: a systematic review and meta-analysis

Background: To maintain the continuity of medical education during the COVID-19 epidemic, online learning has replaced traditional face-to-face learning. But the efficacy and acceptance of online learning for medical education remains unknown. This meta-analysis aimed to assess whether online learning improves learning outcomes and is more acceptable to medical students compared to offline learning. Methods: Four databases were searched for randomized controlled trials (RCTs) and comparative studies (non-RCTs) involving online learning published from January 1900 to October 2020. A total of twenty-seven studies comparing online and offline learning in medical students were included. The Grading of Recommendations, Assessment, Development and Evaluations (GRADE) framework and Newcastle-Ottawa Scale (NOS) were used to assess the methodological quality of RCTs and non-RCTs respectively. The data of knowledge and skills scores and course satisfaction were synthesized using a random effects model for the meta-analysis. Results: Twenty-one RCTs that were judged to be of high quality according to the GRADE framework and six non-RCTs studies which ranged from 6 to 8 (NOS) and can be considered high-quality were included in this meta-analysis. The revealed that the online learning group had significantly higher post-test scores (SMD =0.58, 95% CI: 0.25 to 0.91;P=0.0006) and pre-and post-test score gains than the offline group (SMD =1.12, 95% CI: 0.14 to 2.11, P=0.02). In addition, online education was more satisfactory to participants than the offline learning (OR: 2.02;95% CI: 1.16 to 3.52;P=0.01). Subgroup analysis was performed on knowledge and skill scores at the post-test level. The selected factors included study outcome, study design and type, participants, course type and country. No significant factors were observed in the subgroup analysis except for course type subgroup analysis. Discussion: Online learning in medical education could lead to higher post-test knowledge and skill scores than offline learning. It also has higher satisfaction ratings than offline education. In conclusion, online learning can be considered as a potential educational method during the COVID-19 pandemic. However, given the risk of bias of included studies such as the inclusion of non-randomized comparative studies, the conclusion should be made with cautions. Trial Registration: CRD42020220295. © Journal of Xiangya Medicine. All rights reserved.

[Advances in clinical prediction scores for prognosis of coronavirus disease-2019].

Coronavirus Disease-2019 (COVID-19) has been a major public health issue all over the world, placing a significant burden on available healthcare resources. The most common types of COVID-19 are the mild and common forms. Although the proportion of the severe-critical types is smaller, the rate of death is significantly higher and the medical resources required tend to be greater. Thus, a variety of scores based on other disease and COVID-19 were used to assess the risk of poor prognosis on the COVID-19, including the common scores for community-acquired pneumonia, sepsis and viral pneumonia. Unfortunately, the above scores often lacked an adequate description of the applicable population or were at high risk of bias with unknown applicability. Therefore, the article summarized the existing scores, aiming to provide a reference for clinical prognostic risk assessment.

Sentiment Analysis of Autonomous Vehicles after Extreme Events Using Social Media Data

This paper aims to leverage social media data to understand the public opinion on autonomous driving after extreme events, including the Uber and Tesla crashes and the COVID-19 pandemic. Uber and Tesla crashes that happened consecutively in 2018 have posed uncertainty and the public concern toward the autonomous vehicle (AV) technology. The COVID-19 pandemic has drastically increased people's fear of taking mass transit, while the social distancing policy could easily favor contactless travel experiences provided by AVs. To understand people's attitudinal changes before and after these extreme events, three sources of social media data are leveraged: Facebook, Twitter and Reddit. Sentiment analysis is performed with BERT (Bidirectional Encoder Representation from Transformers) model to study the change in people's attitude toward AVs. Results show that after Uber and Tesla crashes, the proportion of people with a negative attitude increases, while after the pandemic, the proportion of people with a positive attitude increases. These results are quite consistent with our intuition. We then conduct regression analysis using XGBoost to analyze the impact of individual's demographic information on his/her sentiment toward AVs. We find that Age has the most significant effect on people's attitudes toward AVs. Engineers and entrepreneurs are more likely to introduce and discuss the AV technology in social media. © 2021 IEEE.

The application of KMV model in China's insurance market during the COVID-19

In 2020, the COVID-19 has a certain impact on the credit risk of China's listed of insurance companies. This paper selects relevant data of four China's listed insurance companies from the first quarter of 2019 to the third quarter of 2020, and then uses revised KMV model to measure the credit risk of these four insurance companies. The empirical results show that in the season of the outbreak, the default distance of China's listed insurance companies has decreased to varying degrees, indicating that the epidemic has caused a temporary increase in the credit risk of the insurance industry. © 2021 Institute of Physics Publishing. All rights reserved.

Correlation of the ratio of IgM/IgG concentration to days after symptom onset (IgM/T or IgG/T) with disease severity and outcome in non-critical COVID-19 patients

BACKGROUND: Correlation of SARS-CoV-2 serum antibodies with COVID-19 development and outcome has not been fully studied. Due to the time dynamic of antibodies, the antibody concentration of the same patient varies greatly at different times during the course of the disease. Therefore, our study used IgM/T or IgG/T (the ratio of serum antibody concentration to days after symptom onset) to reflect the patient's humoral immune status, and analyzed their correlation with COVID-19 development and outcome. METHODS: Clinical data of 50 non-critical COVID-19 patients were retrospectively analyzed. Time-resolved fluorescence immunochromatography was used to quantitatively detect SARS-CoV-2 IgM and IgG. Correlation analysis was performed. RESULTS: IgM antibody was positive on day 5 of symptom onset, increased within 2 weeks, and then gradually decreased. However, IgG antibody was positive on week 2 of symptom onset and continued to increase since. Additionally, IgM/T, but not IgG/T of recovery period (Spearman rho=0.17;P=0.283), was negatively correlated with disease course in 2 weeks of symptom onset (Spearman rho=-0.860;P=0.000). IgG/T of recovery period was positively correlated with clinical classification (Spearman rho=0.432;P=0.004), number of involved lung lobes (Spearman rho=0.343;P=0.026), and lung lesions (Spearman rho=0.472;P=0.002). CONCLUSIONS: Within 2 weeks of symptom onset, higher IgM/T indicates faster recovery and shorter disease course. In recovery period, higher IgG/T suggests more serious disease. IgM/T or IgG/T may predict disease severity and outcome in non-critical COVID-19 patients.

Diagnostic value of combined nucleic acid and antibody detection in suspected COVID-19 cases.

OBJECTIVES: Nucleic acid testing is the gold standard method for the diagnosis of coronavirus disease 2019 (COVID-19); however, large numbers of false-negative results have been reported. In this study, nucleic acid detection and antibody detection (IgG and IgM) were combined to improve the testing accuracy of patients with suspected COVID-19. STUDY DESIGN: The positive rate of nucleic acid detection and antibody detection (IgG and IgM) were compared in suspected COVID-19 patients. METHODS: A total of 71 patients with suspected COVID-19 were selected to participate in this study, which included a retrospective analysis of clinical features, imaging examination, laboratory biochemical examination and nucleic acid detection and specific antibody (IgM and IgG) detection. RESULTS: The majority of participants with suspected COVID-19 presented with fever (67.61%) and cough (54.93%), and the imaging results showed multiple small patches and ground-glass opacity in both lungs, with less common infiltration and consolidation opacity (23.94%). Routine blood tests were mostly normal (69.01%), although only a few patients had lymphopenia (4.23%) or leucopenia (12.68%). There was no statistical difference in the double-positive rate between nucleic acid detection (46.48%) and specific antibody (IgG and IgM) detection (42.25%) (P = 0.612), both of which were also poorly consistent with each other (kappa = 0.231). The positive rate of combined nucleic acid detection and antibody detection (63.38%) was significantly increased, compared with that of nucleic acid detection (46.48%) and that of specific antibody (IgG and IgM) detection (42.25%), and the differences were statistically significant (P = 0.043 and P = 0.012, respectively). CONCLUSIONS: Nucleic acid detection and specific antibody (IgG and IgM) detection had similar positive rates, and their combination could improve the positive rate of COVID-19 detection, which is of great significance for diagnosis and epidemic control.

[Discussion on diagnosis and treatment of hepatobiliary malignancies during the outbreak of COVID-19].

Objective: From December 2019, the new coronavirus pneumonia (COVID-19) broke out in Wuhan, Hubei, and spread rapidly to the nationwide. On January 20, 2020, the National Health Committee classified COVID-19 pneumonia as one of B class infectious diseases and treated it as class A infectious disease. During the epidemic period, the routine diagnosis and treatment of tumor patients was affected with varying degrees. In this special period, we performed the superiority of the multi-disciplinary team of diagnosis and treatment, achieved accurate diagnosis and treatment of patients with hepatobiliary malignant tumors, provided support for these patients with limited medical resources, and helped them to survive during the epidemic period.On the basis of fully understanding the new coronavirus pneumonia, the treatment strategy should be changed timely during the epidemic, and more appropriate treatment methods should be adopted to minimize the adverse effect of the epidemic on tumor treatment.