Impact of Comorbidity on Mortality in COVID-19 Patients: A Single- Center Retrospective Study (preprint)

The global spread of COVID-19 has had a profound impact on human health, with millions of people infected and a significant death. Comorbidities play a pivotal role in the prognosis of COVID-19 patients. The aim of this study was to assess the impact of the comorbidity on mortality in COVID-19 patients in a Single-Centre Retrospective Study. The characteristics and results of patients with COVID-19 admitted to KLE's Hospital in Belgaum, Karnataka, India, were examined through retrospective research. 642 participants having COVID-19 diagnoses between October 1, 2020, and September 30, 2021 were enrolled in the research. Data such as, patients' clinical features, vital signs, demographic information and patients' outcomes (survived or deceased) were collected. The findings of this study showed that out of 642 patients, 256 patients had co morbidities, 62.8% of them had poly morbidity and most prevalent underlying medical conditions were hypertension, diabetes, and heart disease which affected 30.7%, 29.1, and 7.9%, respectively. Only diabetes and renal disease reported strong associations (P.value: 0.011, aOR: 1.852 95% CI: 1.148–2.988), (p.value: 0.000, aOR: 6.491 (95% CI: 2.613–16.124), respectively. Furthermore, Comorbidities such kidney disease, and diabetes mellitus can lead to more serious complications and death in COVID-19 patients. Understanding the impact of these comorbidities on COVID-19 mortality is essential for more effective patient care and resource allocation.

Changes in the senescence profile and immune checkpoints in HIV-infected individuals after COVID-19 (preprint)

Background: Both SARS-CoV-2 and HIV infection exhibit alterations in the senescence profile and immune checkpoint (IC) molecules. However, the midterm impact of SARS-CoV-2 on these profiles in people with HIV (PWH) remains unclear. This study aimed to evaluate differences in plasma biomarker levels related to ICs, the senescence-associated secretory phenotype (SASP), and pro- and anti-inflammatory cytokines in PWH following recovery from SARS-CoV-2 infection. Methods: We conducted a cross-sectional study of 95 PWH receiving antiretroviral therapy, stratified by SARS-CoV-2 infection status: a) 48 previously infected (HIV/SARS) and b) 47 controls without previous infection (HIV). Plasma biomarkers (n=44) were assessed using Procartaplex Multiplex Immunoassays. Differences were analyzed using a generalized linear model adjusted for sex and ethnicity and corrected for the false discovery rate. Significant values were defined as an adjusted arithmetic mean ratio [≥]1.2 or [≤]0.8 and a qvalue<0.1. Spearman correlation evaluated relationships between plasma biomarkers (significant correlations, rho[≥]0.3 and q value<0.1). Results: The median age of the PWH was 45 years, and 80% were men. All SARS-CoV-2-infected PWH experienced symptomatic infection; 83.3% had mild symptomatic infection, and sample collection occurred at a median of 12 weeks postdiagnosis. The HIV/SARS group showed higher levels of ICs (CD80, PDCD1LG2, CD276, PDCD1, CD47, HAVCR2, TIMD4, TNFRSF9, TNFRSF18, and TNFRSF14), SASP (LTA, CXCL8, and IL13), and inflammatory plasma biomarkers (IL4, IL12B, IL17A, CCL3, CCL4, and INF1A) than did the HIV group. Conclusions: SARS-CoV-2 infection in PWH causes significant midterm disruptions in plasma ICs and inflammatory cytokine levels, highlighting SASP-related factors, which could be risk factors for the emergence of complications in PWH.

Coronavirus Spike-RBD Variants Differentially Bind to the Human ACE2 Receptor (preprint)

The SARS-CoV-2 betacoronavirus infects people through binding the human Angiotensin Receptor 2 (ACE2), followed by import into a cell utilizing the Transmembrane Protease, Serine 2 (TMPRSS2) and Furin cofactors. Analysis of the SARS-CoV-2 extracellular spike protein has suggested critical amino acids necessary for binding within a 197-residue portion, the receptor binding domain (RBD). A cell-based assay between a membrane tethered RBD-GFP fusion protein and the membrane bound ACE2-Cherry fusion protein allowed for mutational intersection of both RBD and ACE2 proteins. Data shows Omicron BA.1 and BA.2 variants have altered dependency on the amino terminus of ACE2 protein and suggests multiple epitopes on both proteins stabilize their interactions at the Nt and internal region of ACE2. In contrast, the H-CoV-NL63 RBD is only dependent on the ACE2 internal region for binding. A peptide inhibitor approach to this internal region thus far have failed to block binding of RBDs to ACE2, suggesting that several binding regions on ACE2 are sufficient to allow functional interactions. In sum, the RBD binding surface of ACE2 appears relatively fluid and amenable to bind a range of novel variants.

Evaluation of Country Dietary Habits Using Machine Learning Techniques in Relation to Deaths from COVID-19 (preprint)

COVID-19 disease has affected almost every country in the world. The large number of infected people and the different mortality rates between countries has given rise to many hypotheses about the key points that make the virus so lethal in some places. In this study, the eating habits of 170 countries were evaluated in order to find correlations between these habits and mortality rates caused by COVID-19 using machine learning techniques that group the countries together according to the different distribution of fat, energy, and protein across 23 different types of food, as well as the amount ingested in kilograms. Results shown how obesity and the high consumption of fats appear in countries with the highest death rates, whereas countries with a lower rate have a higher level of cereal consumption accompanied by a lower total average intake of kilocalories.

Microscopic models for the large-scale spread of SARS-CoV-2 virus: A Statistical Mechanics approach (preprint)

In this work, we derive a system of Boltzmann-type equations to describe the spread of SARS-CoV-2 virus at the microscopic scale, that is by modeling the human-to-human mechanisms of transmission. To this end, we consider two populations, characterized by specific distribution functions, made up of individuals without symptoms (population $1$) and infected people with symptoms (population $2$). The Boltzmann operators model the interactions between individuals within the same population and among different populations with a probability of transition from one to the other due to contagion or, vice versa, to recovery. In addition, the influence of innate and adaptive immune systems is taken into account. Then, starting from the Boltzmann microscopic description we derive a set of evolution equations for the size and mean state of each population considered. Mathematical properties of such macroscopic equations, as equilibria and their stability, are investigated and some numerical simulations are performed in order to analyze the ability of our model to reproduce the characteristic features of Covid-19.

Agent based network modelling of COVID-19 disease dynamics and vaccination uptake in a New South Wales Country Township (preprint)

We employ an agent-based contact network model to study the relationship between vaccine uptake and disease dynamics in a hypothetical country town from New South Wales, Australia, undergoing a COVID-19 epidemic, over a period of three years. We model the contact network in this hypothetical township of N = 10000 people as a scale-free network, and simulate the spread of COVID-19 and vaccination program using disease and vaccination uptake parameters typically observed in such a NSW town. We simulate the spread of the ancestral variant of COVID-19 in this town, and study the disease dynamics while the town maintains limited but non-negligible contact with the rest of the country which is assumed to be undergoing a severe COVID-19 epidemic. We also simulate a maximum three doses of Pfizer Comirnaty vaccine being administered in this town, with limited vaccine supply at first which gradually increases, and analyse how the vaccination uptake affects the disease dynamics in this town, which is captured using an extended compartmental model with epidemic parameters typical for a COVID-19 epidemic in Australia. Our results show that, in such a township, three vaccination doses are sufficient to contain but not eradicate COVID-19, and the disease essentially becomes endemic. We also show that the average degree of infected nodes (the average number of contacts for infected people) predicts the proportion of infected people. Therefore, if the hubs (people with a relatively high number of contacts) are disproportionately infected, this indicates an oncoming peak of the infection, though the lag time thereof depends on the maximum number of vaccines administered to the populace. Overall, our analysis provides interesting insights in understanding the interplay between network topology, vaccination levels, and COVID-19 disease dynamics in a typical remote NSW country town.

Perceived Risk of Developing Severe Illness or Complications Due to Covid-19 in the Brazilian Population (preprint)

Background The perception of the risk of developing serious illness as a result of COVID-19 was one of the first reports used to reflect the health condition of infected people after hospital admission. The effects of COVID-19 are more severe in individuals with chronic noncommunicable diseases (NCDs), indicating that the characteristics and implications of these diseases in people with COVID-19 need to be investigated.Methods This cross-sectional study was was carried out with 1961 people aged 18 or older living in Brazil. An open research approach (survey) was used for the sample design, which involved the use of an online questionnaire. Descriptive statistical analysis and logistic regression were applied to identify factors associated with the perceived risk of complications due to COVID-19.Results The sample was mostly composed of women (n = 1383; 70.5%), 18 to 39 years old (n = 1144; 58.3%), and white (n = 1140; 56.4%). It was possible to observe that people who perceived a risk of developing diseases or complications if they became infected with COVID-19 were more likely to have a chronic noncommunicable disease (NCD) (OR: 4.51; 95% CI: 3.61–5.65), self-perception of potential risk of becoming infected with COVID-19 (OR: 2.34; 95% CI: 1.87–2.93), self-perceived potential risk of the population becoming infected with COVID-19 (OR: 5.80; 95% CI: 3.30–10.74), wearing a protective mask (OR: 12.98; 95% CI: 5.8–31.35) during the pandemic period and having a religion (OR: 1.29; 95% CI: 1.02–1.63).Conclusions The study showed that the perception of the risk of developing a severe form of the disease was significant in certain groups, such as religious people or those with chronic noncommunicable diseases.

Plasma-based antigen persistence in the post-acute phase of SARS-CoV-2 infection (preprint)

BACKGROUND: Although RNA viruses like SARS-CoV-2 are generally thought to be transient, the persistence of viral components beyond the acute phase can be driven by a variety of virologic and immunologic factors. Recent studies have suggested that SARS-CoV-2 antigens may persist following COVID-19 but were limited by a lack of comparison to a large number of true negative control samples. METHODS: Using single molecule array (Simoa) assays for SARS-CoV-2 spike, S1, and nucleocapsid antigen in plasma from 171 pandemic-era individuals in the post-acute phase of SARS-CoV-2 infection and 250 pre-pandemic control samples, we compared prevalence of antigen detection. We used logistic regression models and prevalence ratios (PRs) to assess the relationship between demographic and disease factors and antigen persistence. RESULTS: Compared to the proportion of antigen positivity in the pre-pandemic controls (2%), detection of any SARS-CoV-2 antigen was more frequent across all post-acute COVID-19 time bins (3-6 months: 12.6%, p<0.001; 6-10 months, 10.7%, p=0.0002; 10-14 months, 7.5%, p=0.017). These differences were driven by spike protein for up to 14 months and nucleocapsid in the first 6 months after infection. The co-occurrence of multiple antigens at a single timepoint was uncommon. Hospitalization for acute COVID-19 (versus not hospitalized) and worse self-reported health during acute COVID-19 among those not hospitalized (versus more benign illness) were associated with higher prevalence of post-acute antigen detection (PR 1.86, p=0.03; PR 3.5, p=0.07, respectively) in the pandemic era. CONCLUSIONS: Our findings provide strong evidence that SARS-CoV-2 antigens can persist beyond the period of acute illness. The observation that more than 10% of plasma samples for over a year following initial SARS-CoV-2 infection contain detectable viral antigen, which are potentially immunogenic, has significant implications given the sheer number of people infected with SARS-CoV-2 to date. More work will be needed to determine whether these antigens have a causal role in post-acute sequelae of SARS-CoV-2 infection (PASC).

Factors associated with SARS-CoV-2 among persons living with HIV in Zambia: Analysis of three concurrent SARS-CoV-2 prevalence surveys conducted in July 2020 in six districts of Zambia (preprint)

People living with HIV (PLHIV) are a unique population because of their altered immune systems and taking antiretroviral therapy (ART) that may modify risk of SARS-CoV-2. Evidence from sub-Saharan African countries suggests that, despite not having higher SARS-CoV-2 infection prevalence than HIV-negative persons, PLHIV suffer worse COVID-19 outcomes. We assessed the prevalence of SARS-COV-2 infection by HIV status in Zambia in July 2020. We analyzed data from three different concurrent SARS-CoV-2 prevalence surveys (household, outpatient-department, and health-worker) conducted in six districts of Zambia in July 2020. Information on demographics and medical history was collected. Nasopharyngeal swabs were used to screen for SARS-CoV-2 RNA using polymerase-chain-reaction (PCR) and blood specimens were screened for SARS-CoV-2 virus-specific antibodies using an enzyme-linked-immunosorbent-assay (ELISA). Test-specific SARS-CoV-2 prevalence was calculated. Multilevel logistic regression models were used to measure test-specific adjusted odd ratios (aORs) of SARS-CoV-2 positivity by HIV status, controlling for demographic and medical history. We analysed the outcomes of the two different tests separately. Among 7,092 participants, 4,717 (66.5%) consented to blood-draw and 4,642 (65.5%) consented to nasopharyngeal swab. Overall SARS-CoV-2 positivity was 9.4% by PCR and 3.8% by ELISA. SARS-CoV-2 prevalence detected by PCR was higher among PLHIV than HIV-negative respondents (12.4% vs 9.1%, respectively, OR: 1.4, 95% confidence intervals [CI]: 1.0-1.9) and lower by ELISA (1.9% vs 3.9%, respectively, OR: 0.5, 95%CI: 0.2-0.9). Among PLHIV, not being on ART was an independent predictor of SARS-CoV-2 PCR positivity (aOR: 5.24,95% CI: 1.19-22.22) but did not have a significant effect on ELISA results. During the first COVID-19 wave in Zambia, PLHIV were more likely to be acutely infected with SARS-CoV-2 but less likely to be seropositive than participants without HIV. Intervention programs could focus on early access to COVID-19 vaccinations, testing and ART might reduce COVID-19 morbidity among PLHIV.

Perceptions of access to harm reduction services during the COVID-19 pandemic among people who inject drugs in Chicago (preprint)

Background The COVID-19 pandemic amplified the risk environment for people who inject drugs (PWID), making continued access to harm reduction services imperative. Research has shown that some harm reduction service providers were able to continue to provide services throughout the pandemic. Most of these studies, however, focused on staff perspectives, not those of PWID. Our study examines changes in perceptions of access to harm reduction services among PWID participating in a longitudinal study conducted through the University of Illinois-Chicago's Community Outreach Intervention Project field sites during the COVID-19 pandemic. Methods Responses to a COVID-19 module added to the parent study survey that assessed the impact of COVID-19 on PWID participating in an ongoing longitudinal study were analyzed to understand how study participants self-reported access to harm reduction services changed throughout the pandemic. Mixed effects logistic regression was used to examine difficulty in syringe access as an outcome of COVID-19 phase. Results Most participants reported that access to syringes and naloxone remained the same as prior to the pandemic. Participants had significantly higher odds of reporting difficulty in accessing syringes earlier in the pandemic. Conclusions The lack of perceived changes in harm reduction access by PWID and the decrease in those reporting difficulty accessing syringes as the pandemic progressed suggests the efficacy of adaptations to harm reduction service provision (e.g., window and mobile service) during the pandemic. Further research is needed to understand how the COVID-19 pandemic may have impacted PWIDs engagement with harm reduction services.

Unravelling Causal Associations between Population Mobility and COVID-19 Cases in Spain: a Transfer Entropy Analysis (preprint)

Human mobility is a well-known factor in the spread of infectious diseases. During the COVID-19 pandemic, the rapid spread of the SARS-CoV-2 virus led to healthcare systems collapsing in numerous countries, such as Spain and Italy, resulting in a significant number of deaths. To avoid such disastrous outcomes in the future, it is vital to understand how population mobility is linked to the spread of infectious diseases. To assess that, we applied an information theoretic approach called transfer entropy (TE) to measure the influence of the number of infected people travelling between two localities on the future number of infected people in the destination. We first validated our approach using simulated data from a SIR epidemiological model and found that the mobility-based TE was effective in filtering out non-causal influences that could otherwise arise, thereby successfully recovering the epidemic's spreading patterns and the mobility network topology. We then applied the mobility-based TE to analyse the COVID-19 pandemic in Spain. We identified which regions acted as the main drivers of the pandemic at different periods, both globally and locally. Our results unravelled significant epidemiological events such as the outbreak in Lleida during the Summer of 2020, caused by the influx of temporary workers. We also analysed the effects of a non-pharmaceutical intervention in Catalunya, using mobility-based TE to compare the infection dynamics with a control region. These results help clarify how human mobility influences the dynamic spread of infectious diseases and could be used to inform future non-pharmaceutical interventions.

Clinical and laboratory features of COVID-19 illness and outcomes in immunocompromised individuals during the first pandemic wave in Sydney, Australia (preprint)

People with immunocompromising conditions are at increased risk of SARS-CoV-2 infection and mortality, however early in the pandemic it was challenging to collate data on this heterogenous population. We conducted a registry study of immunocompromised individuals with polymerase chain reaction (PCR)-confirmed SARS-CoV-2 infection from March - October 2020 in Sydney, Australia to understand clinical and laboratory outcomes in this population prior to the emergence of the Delta variant. 27 participants were enrolled into the study including people with a haematologic oncologic conditions (n=12), secondary immunosuppression (N=8) and those with primary or acquired immunodeficiency (i.e. HIV; N=7). All participants had symptomatic COVID-19 with the most common features being cough (64%), fever (52%) and headache (40%). Five patients demonstrated delayed SARS-CoV-2 clearance lasting three weeks to three months. The mortality rate in this study was 7% compared to 1.3% in the state of New South Wales Australia during the same period. This study provides data from the first eight months of the pandemic on COVID-19 outcomes in at-risk patient groups.

Impact of age and sex on neuroinflammation following SARS-CoV-2 infection in a murine model (preprint)

Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), the etiological agent for the worldwide COVID-19 pandemic, is known to infect people of all ages and both sexes. Senior populations have the greatest risk of severe disease, and sexual dimorphism in clinical outcomes has been reported in COVID-19. SARS-CoV-2 infection in humans can cause damage to multiple organ systems, including the brain. Neurological symptoms are widely observed in patients with COVID-19, with many survivors suffering from persistent neurological and cognitive impairment, potentially accelerating Alzheimer's disease. The present study aims to investigate the impact of age and sex on the neuroinflammatory response to SARS-CoV-2 infection using a mouse model. Wild-type C57BL/6 mice were inoculated, by intranasal route, with SARS-CoV-2 lineage B.1.351 variant known to infect mice. Older animals and in particular males exhibited a significantly greater weight loss starting at 4 dpi. In addition, male animals exhibited higher viral RNA loads and higher titers of infectious virus in the lung, which was particularly evident in males at 16 months of age. Notably, no viral RNA was detected in the brains of infected mice, regardless of age or sex. Nevertheless, expression of IL-6, TNF-, and CCL-2 in the lung and brain was increased with viral infection. An unbiased brain RNA-seq/transcriptomic analysis showed that SARS-CoV-2 infection caused significant changes in gene expression profiles in the brain, with innate immunity, defense response to virus, cerebravascular and neuronal functions, as the major molecular networks affected. The data presented in this study show that SARS-CoV-2 infection triggers a neuroinflammatory response despite the lack of detectable virus in the brain. Age and sex have a modifying effect on this pathogenic process. Aberrant activation of innate immune response, disruption of blood-brain barrier and endothelial cell integrity, and supression of neuronal activity and axonogenesis underlie the impact of SARS-CoV-2 infection on the brain. Understanding the role of these affected pathways in SARS-CoV-2 pathogenesis helps identify appropriate points of therapeutic interventions to alleviate neurological dysfunction observed during COVID-19.

Structure, Dynamics and Free Energy Studies on the Effect of Spot Mutations on SARS-CoV-2 Spike Protein Binding with ACE2 Receptor (preprint)

The ongoing COVID-19 pandemic continues to infect people worldwide, and the virus continues to evolve in significant ways which can pose challenges to the efficiency of available vaccines and therapeutic drugs and cause future pandemic. Therefore, it is important to investigate the binding and interaction of ACE2 with different RBD variants. A comparative study using all-atom MD simulations was conducted on ACE2 binding with 8 different RBD variants, including N501Y, E484K, P479S, T478I, S477N, N439K, K417N and N501Y-E484K-K417N on RBD. Based on the RMSD, RMSF, and DSSP results, the overall the binding of RBD variants with ACE2 is stable, and the secondary structure of RBD and ACE2 are consistent after the spot mutation. Besides that, a similar buried surface area, a consistent binding interface and a similar amount of hydrogen bonds formed between RBD with ACE2 although the exact residue pairs on the binding interface were modified. The change of binding free energy from spot mutation was predicted using the free energy perturbation (FEP) method. It is found that N501Y, N439K, and K417N can strengthen the binding of RBD with ACE2, while E484K and P479S weaken the binding, and S477N and T478I have negligible effect on the binding. Spot mutations modified the dynamic correlation of residues in RBD based on the dihedral angle covariance matrix calculation. Doing dynamic network analysis, a common intrinsic network community extending from the tail of RBD to central, then to the binding interface region was found, which could communicate the dynamics in the binding interface region to the tail thus to the other sections of S protein. The result can supply unique methodology and molecular insight on studying the molecular structure and dynamics of possible future pandemics and design novel drugs.

A Population Level Study on the Determinants of COVID-19 Vaccine Hesitancy at the U.S. County Level (preprint)

Multiple COVID-19 vaccines were proven to be safe and effective in curbing severe illness, but despite vaccine availability, uptake rates were relatively low in the United States (U.S.), primarily due to vaccine hesitancy. To better understand factors associated with COVID-19 vaccine hesitancy in the U.S., our study provides a comprehensive, data-driven population-level statistical analysis at the county level. We find that political affiliation, as determined by the proportion of votes received by the Republican candidate in the 2020 presidential election, has the strongest association with COVID-19 vaccine hesitancy. The next strongest association was median household income, which has a negative association. The percentage of Black people and the average number of vehicles per household are also positively associated with vaccine hesitancy. In contrast, COVID-19 infection rate, percentage of Hispanic people, postsecondary education percentage, median age, and prior non-COVID-19 childhood vaccination coverage are other factors negatively associated with vaccine hesitancy. Unlike previous studies, we do not find significant relationships between cable TV news viewership or Twitter misinformation variables with COVID-19 vaccine hesitancy. These results shed light on some factors that may impact vaccination choice in the U.S. and can be used to target specific populations for educational outreach and vaccine campaign strategies in efforts to reduce vaccine hesitancy.

52 Weeks Later: Attitudes Towards COVID-19 Apps for Different Purposes Over Time (preprint)

The COVID-19 pandemic has prompted countries around the world to introduce smartphone apps to support disease control efforts. Their purposes range from digital contact tracing to quarantine enforcement to vaccination passports, and their effectiveness often depends on widespread adoption. While previous work has identified factors that promote or hinder adoption, it has typically examined data collected at a single point in time or focused exclusively on digital contact tracing apps. In this work, we conduct the first representative study that examines changes in people's attitudes towards COVID-19-related smartphone apps for five different purposes over the first 1.5 years of the pandemic. In three survey rounds conducted between Summer 2020 and Summer 2021 in the United States and Germany, with approximately 1,000 participants per round and country, we investigate people's willingness to use such apps, their perceived utility, and people's attitudes towards them in different stages of the pandemic. Our results indicate that privacy is a consistent concern for participants, even in a public health crisis, and the collection of identity-related data significantly decreases acceptance of COVID-19 apps. Trust in authorities is essential to increase confidence in government-backed apps and foster citizens' willingness to contribute to crisis management. There is a need for continuous communication with app users to emphasize the benefits of health crisis apps both for individuals and society, thus counteracting decreasing willingness to use them and perceived usefulness as the pandemic evolves.

Research on the infection status of and factors influencing COVID- 19 based on a questionnaire survey (preprint)

Purpose To investigate the infection situation of coronavirus disease 2019 (COVID-19) in the population after the lifting of epidemic prevention and control measures and to study and analyze its related influencing factors.Methods From March 11 to March 20, 2023, questionnaires on COVID-19 infection were distributed on the Questionnaire Star platform, and SPSS 19.0 software was used for statistical processing.Results The results showed that the infection rate of COVID-19 among the respondents reached 72.24%. Fever, fatigue, cough were the main symptoms during infection, among which 58.81% of infected people reported post COVID-19 symptoms. An unconditional logistic regression model showed that age (P = 0.002), area of residence (P༜0.001), vaccine doses (P = 0.025) were statistically significant.The infection risk was the lowest (OR = 0.40; 95% CI = 0.14–1.11) in the ≤ 17 age group and the highest (OR = 1.40; 95% CI = 0.63–3.10) in the 18–44 age group.The risk of infection was higher in urban areas (OR = 2.28; 95% CI = 1.69 ~ 3.35). The risk of infection among one dose of vaccine was the highest (OR = 1.77; 95% CI = 0.43 ~ 7.28).Conclusions COVID-19 had a high infection rate, children had a lower risk of COVID-19.People who lived in cities were more susceptible to COVID-19 and it is necessary to increase the number of vaccine doses.

CSI-Based Efficient Self-Quarantine Monitoring System Using Branchy Convolution Neural Network (preprint)

Nowadays, Coronavirus disease (COVID-19) has become a global pandemic because of its fast spread in various countries. To build an anti-epidemic barrier, self-isolation is required for people who have been to any at-risk places or have been in close contact with infected people. However, existing camera or wearable device-based monitoring systems may present privacy leakage risks or cause user inconvenience in some cases. In this paper, we propose a Wi-Fi-based device-free self-quarantine monitoring system. Specifically, we exploit channel state information (CSI) derived from Wi-Fi signals as human activity features. We collect CSI data in a simulated self-quarantine scenario and present BranchyGhostNet, a lightweight convolution neural network (CNN) with an early exit prediction branch, for the efficient joint task of room occupancy detection (ROD) and human activity recognition (HAR). The early exiting branch is used for ROD, and the final one is used for HAR. Our experimental results indicate that the proposed model can achieve an average accuracy of 98.19% for classifying five different human activities. They also confirm that after leveraging the early exit prediction mechanism, the inference latency for ROD can be significantly reduced by 54.04% when compared with the final exiting branch while guaranteeing the accuracy of ROD.