Video-Based Online Interactive System Using CNN Segmentation Method

In recent years, the COVID-19 has made it difficult for people to interact with each other face-to-face, but various kinds of social interactions are still needed. Therefore, we have developed an online interactive system based on the image processing method, that allows people in different places to merge the human region of two images onto the same image in real-time. The system can be used in a variety of situations to extend its interactive applications. The system is mainly based on the task of Human Segmentation in the CNN (convolution Neural Network) method. Then the images from different locations are transmitted to the computing server through the Internet. In our design, the system ensures that the CNN method can run in real-time, allowing both side users can see the integrated image to reach 30 FPS when the network is running smoothly. © 2023 IEEE.

Assessing Mask Integrity and Fit of N95 Masks after 5 Cycles of Decontamination and Reuse in a Simulated Clinical Setting

Aim: To evaluate the effect of decontamination and reuse on N95 masks. Background(s): The coronavirus disease (COVID-19) pandemic has strained the global availability of masks. Such shortage represents a threat to healthcare workers (HCWs). Mask reprocessing and reuse may alleviate the shortage. Many laboratory studies have proven the effectiveness and feasibility of decontaminating N95 masks. However, very few had HCWs wearing them between cycles of decontamination. Our study evaluated mask integrity (assessed by qualitative mask fitting [QMF], as well as technical measures like bacterial filtration efficacy [BFE]) through five cycles of decontamination using four different modalities - steam, moist heat (MH), UV-C irradiation (UVCI), and hydrogen peroxide vaporization (HPV). Method(s): Each study cycle involved a HCW wearing a N95 mask for two hours, followed by the assigned decontamination process, and then a QMF. This was repeated for a maximum of 5 cycles, as long as the wearer passed QMF. 40 HCWs were recruited for each of the four decontamination modalities. The technical measures of mask integrity assessed were: BFE, Particulate Filtration Efficiency (PFE), Pressure Drop and Splash Resistance. Result(s): 60.6% (HPV) to 77.5% (MH) of the masks passed five cycles of wear and decontamination, as assessed by the wearers passing QMF all five times. MH-decontaminated masks retained all technical measures of integrity through all 5 cycles. HPV reduced masks' BFE after the fourth cycle while UVCI tended to increase the Pressure Drop. Conclusion(s): The results suggest that MH is a promising method for decontaminating N95 masks without compromising fit and integrity. [Figure presented] [Table presented]Copyright © 2023 Southern Society for Clinical Investigation.

COVID-19 Prediction Model Based on Pre-training and Fine-tuning Strategy

The COVID-19 pandemic has had a serious impact on the global society. Building a mathematical model to predict the number of confirmed cases will help provide a basis for public health decision-making.In a complex and changeable external environment, the infectious disease prediction model based on deep learning has become commonly researched. However, the existing models have high requirements regarding the amount of data and cannot adapt to a scene with scarce data during supervised learning. This results in the reduction of model prediction accuracy.The COVID-19 prediction model P-GRU combined with pre-training and fine-tuning strategy is constructed in this study. By adopting the pre-training strategy on the dataset obtained from a specific region, the model is exposed to more epidemic data in advance. Consequently, it can learn the implicit evolution law of COVID-19, provide more sufficient prior knowledge for model prediction, and use the fixed length series containing recent historical information to predict the number of confirmed cases in the future.During the prediction process, the impact of local restrictive policies on the epidemic trend is considered to realize an accurate prediction of the dataset in the target area. The experimental results demonstrate that the pre-training strategy can effectively improve the prediction performance.Compared to Convolution Neural Network (CNN), Recurrent Neural Network (RNN), Long and Short Term Memory (LSTM) network, and Gated Recurrent Unit (GRU) models, P-GRU model attains excellent performance regarding the Mean Absolute Percentage Error (MAPE) and Root Mean Squared Error (RMSE) evaluation indexes. Furthermore, it is more suitable for predicting the transmission trend of COVID-19. © 2022, Editorial Office of Computer Engineering. All rights reserved.

Analysis of applicable pattern and mechanism of medicine and food homologous traditional Chinese medicine in prevention and treatment of modern viral infectious diseases

Objective To explore the application pattern and mechanism of medicine and food homologous traditional Chinese medicine (TCM) against modern viral diseases. Methods The method of literature mining was applied based on the characteristics of modern viral diseases, combining with ancient books and modern prescriptions for the prevention and treatment of viral diseases to build a relevant prescription database. Then SPSS and R language were used to analyze the high-frequency medicine and food homologous TCM and high confidence medicine and food homologous prescriptions in these prescriptions, and cluster analysis was carried out. The antiviral characteristic active ingredients of high-frequency medicinal and food homologous TCN were identified and analyzed, and the action mechanism of active ingredients against modern viral diseases was evaluate by network pharmacology. Results In the prevention and treatment of modern viral diseases, Gancao (Glycyrrhizae Radix et Rhizoma)-Chenpi (Citri Reticulatae Pericarpium)-Fuling (Poria) had the highest confidence, Glycyrrhizae Radix et Rhizoma-Jiegeng (Platycodonis Radix) had the highest support. At the same time, the prescriptions were clustered and analyzed to obtain Jinyinhua (Lonicerae Japonicae Flos)-Huangqi (Astragali Radix)-Huoxiang (Agastache rugosa), Glycyrrhizae Radix et Rhizoma-Xingren (Armeniacae Semen Amarum)-Poria-Platycodonis Radix-Citri Reticulatae Pericarpium, Ganjiang (Zingiberis Rhizoma)-Renshen (Ginseng Radix et Rhizoma), Zisu (Perilla frutescens)-Gegen (Puerariae Lobatae Radix), Lugen (Phragmitis Rhizoma)-Sangye (Mori Folium), Shengjiang (Zingiberis Rhizoma Recens)-Dazao (Jujubae Fructus) clustering new prescription. The core action targets of EGFR, CASP3, VEGFA, STAT3, MMP9, HSP90AA1, mTOR, PTGS2, MMP2, TLR4, MAPK14, etc were identified. The action mechanism involved human cytomegalovirus infection, coronavirus disease-coronavirus disease 2019 (COVID-19), etc. The core action pathway were phosphatidylinositol-3/kinase protein kinase B (PI3K/Akt) signal pathway, mitogen activated protein kinase (MAPK) signal pathway, interleukin-17 (IL-17) signal pathway, Janus kinase/signal transducer and activator of transcription (JAK/STAT) signal pathway, etc. Conclusion Through data mining, six new prescriptions for preventing and controlling modern viral diseases were obtained, and the mechanism of action was preliminarily discussed, which provided some reference for the research and development of medicine and food homologous TCM prescriptions for the prevention and treatment of viral epidemics and related health products.

Insolvency Law and Policy in Times of COVID-19 and Beyond: Academic Roundtable Proceedings

The Centre for Chinese and Comparative Law and Asian Business Law Institute jointly organised an academic roundtable entitled 'Insolvency Law and Policy in Times of COVID-19 and Beyond' on 29 March 2022. The event brought together judges, scholars, experts and professionals from world-leading universities and institutions. This note sets out a summary of the proceedings held at the academic roundtable.

EVALUATING VIRTUAL GROUP VISITS FOR DIABETES CARE DURING THE COVID-19 PANDEMIC

BACKGROUND: One strategy to aid patients in managing their diabetes is group visits (GVs) that include group education and individual medical visits with a provider. Though in-person GVs have been shown to benefit patients, few studies have evaluated virtual diabetes GVs. METHODS: In this single-arm trial, adult patients with A1c ≥ 8% from six community health centers were recruited to participate in monthly virtual GVs for 6 months. Patients completed surveys about diabetes self-care, support, distress, and the group visit concept before and after they participated in GVs. Health center staff trained to lead the GVs completed surveys pre and post virtual GVs. Linear mixed effects models were used to adjust cohort-based association and model the survey data over time for the time trend effect. RESULTS: Forty-five patients enrolled in the study and thirty-eight patients completed the baseline survey. The average age was 55 (range of 36-83) and 65% of enrollees were female. 63% were black/African American, 32% were white/Caucasian, and 8% were Hispanic/Latino. Thirty-four patients attended one or more GVs and twenty-one patients completed the post GV survey. Overall satisfaction with the virtual GVs was high with 18/20 (90%) of participants being very satisfied and 20/21 (95%) saying they would attend GVs in the future. Most participants agreed that GVs helped improve diabetes self-management skills (78%), motivated them to achieve health goals (89%), and introduced them to others living with diabetes (78%). Barriers to participation were the timing of the GVs and access to a computer, tablet, phone, and internet. Patients had an increase in their diabetes knowledge (mean (SD): 3.2/ 5 (0.9) to 3.6/5 (0.7), p= 0.02) and diabetes support (3.5/5 (0.64) to 4.1/5 (0.7), p <0.001) as well as decreased diabetes distress (2.9/6 (1.5) to 1.2/6 (0.5), p=0.03) from baseline to 6 months. Thirty-five staff enrolled in the study and seventeen completed a post GV survey. Most staff agreed that GVs provided patients with social support and more frequent contact with medical providers. Staff largely agreed that virtual GVs increased opportunity for teamwork and collaboration (94%), care coordination (82%), and understanding of patients (94%). However, only 5/17 (29%) and 3/17 (18%) staff members agreed that virtual GVs increased provider productivity or led to higher reimbursement, respectively. Staff cited other priorities at the health center, difficulty recruiting patients, and concerns about access to technology as the biggest barriers to implementing virtual GVs. CONCLUSIONS: Virtual GVs show promise as evidenced by high patient satisfaction and improvements in support, distress, and diabetes knowledge in patients. Staff also perceived virtual GV benefits to patients, staff, and health centers despite concerns about logistics such as productivity, reimbursement, and the health center's ability to continue visits virtually.

IMPACT OF VIRTUAL DIABETES GROUP VISITS ON PATIENT OUTCOMES DURING THE COVID-19 PANDEMIC IN MIDWESTERN HEALTH CENTERS

BACKGROUND: Diabetes group visits (GVs) or shared medical appointments have been shown to improve clinical outcomes, but few have reported results from virtual diabetes GVs. No studies have evaluated virtual GVs among community health center patients across a region of the U.S. METHODS: Six health center sites across five states conducted six monthly virtual GVs with up to 12 adult patients with type 2 diabetes and suboptimal glycemic control (glycosylated hemoglobin (A1C) ≥8%). Virtual group visits consisted of six monthly 60 to 90 minute-long diabetes education sessions led by health center staff via a videoconferencing platform. GV patients enrolled at the site also had an appointment with their primary care physician within two weeks of each monthly virtual group visit. Primary outcome was change in patients' A1C from baseline to 6- months. Secondary outcomes were changes in patients' blood pressure, low density lipoproteins (LDL) and weight. Patients also completed surveys at baseline and 6-months describing their diabetes self-care behaviors and satisfaction with the virtual GVs. Generalized linear mixed models and linear mixed models were used to test the effects of GVs, time points and their interaction. RESULTS: Forty eight patients were enrolled (mean age 55 ± 12 years, 67% female, 63% black/African American, 32% white/Caucasian, and 8% Hispanic/Latino, 88% had public health insurance, mean baseline A1C of 9.84% ± 1.78%, 35% with A1c <9%). 34 patients completed one or more virtual GVs;14 patients attended no virtual group visits. At 6-months, average A1C was 8.96 ±1.82;A1C decreased by -0.56% ± 0.31 compared to baseline which was borderline significant (p=0.08). At 6-months, 58% of patients had an A1C < 9% which was borderline significantly decreased (p=0.055) compared to baseline. For patients with an A1C at baseline >9%, there was a significant decrease in A1C at 6 months (-1.06 ±0.45, p=0.03). There was no significant difference in blood pressure, LDL or weight from baseline to 6- months or association of number of visits attended and change in A1C. There were no significant changes in foot self-exams, blood sugar testing, nor exercise, but patients did report more days of healthy eating in the past week at 6-months compared to baseline (4.5 ±2.3 vs. 3.2 +2.7 days, p=0.02). Overall satisfaction with the virtual GVs was high with 90%of participants being very satisfied and 95% saying they would attend GVs in the future. CONCLUSIONS: Virtual GVs show high patient satisfaction and promise for improving A1C among patients with poor glycemic control who receive care in community health centers. Future studies are needed with a larger patient sample size and a control comparison group to determine which patients and health centers are best suited for virtual GVs.

Disentangled Representation Learning in Heterogeneous Information Network for Large-scale Android Malware Detection in the COVID-19 Era and Beyond

In the fight against the COVID-19 pandemic, many social activities have moved online;society's overwhelming reliance on the complex cyberspace makes its security more important than ever. In this paper, we propose and develop an intelligent system named Dr.HIN to protect users against the evolving Android malware attacks in the COVID-19 era and beyond. In Dr.HIN, besides app content, we propose to consider higher-level semantics and social relations among apps, developers and mobile devices to comprehensively depict Android apps;and then we introduce a structured heterogeneous information network (HIN) to model the complex relations and exploit meta-path guided strategy to learn node (i.e., app) representations from HIN. As the representations of malware could be highly entangled with benign apps in the complex ecosystem of development, it poses a new challenge of learning the latent explanatory factors hidden in the HIN embeddings to detect the evolving malware. To address this challenge, we propose to integrate domain priors generated from different views (i.e., app content, app authorship, app installation) to devise an adversarial disentangler to separate the distinct, informative factors of variations hidden in the HIN embeddings for large-scale Android malware detection. This is the first attempt of disentangled representation learning in HIN data. Promising experimental results based on real sample collections from security industry demonstrate the performance of Dr.HIN in evolving Android malware detection, by comparison with baselines and popular mobile security products.