A Closer Look at Spatial-Slice Features Learning for COVID-19 Detection (preprint)

Conventional Computed Tomography (CT) imaging recognition faces two significant challenges: (1) There is often considerable variability in the resolution and size of each CT scan, necessitating strict requirements for the input size and adaptability of models. (2) CT-scan contains large number of out-of-distribution (OOD) slices. The crucial features may only be present in specific spatial regions and slices of the entire CT scan. How can we effectively figure out where these are located? To deal with this, we introduce an enhanced Spatial-Slice Feature Learning (SSFL++) framework specifically designed for CT scan. It aim to filter out a OOD data within whole CT scan, enabling our to select crucial spatial-slice for analysis by reducing 70% redundancy totally. Meanwhile, we proposed Kernel-Density-based slice Sampling (KDS) method to improve the stability when training and inference stage, therefore speeding up the rate of convergence and boosting performance. As a result, the experiments demonstrate the promising performance of our model using a simple EfficientNet-2D (E2D) model, even with only 1% of the training data. The efficacy of our approach has been validated on the COVID-19-CT-DB datasets provided by the DEF-AI-MIA workshop, in conjunction with CVPR 2024. Our source code will be made available.

Simple 2D Convolutional Neural Network-based Approach for COVID-19 Detection (preprint)

This study explores the use of deep learning techniques for analyzing lung Computed Tomography (CT) images. Classic deep learning approaches face challenges with varying slice counts and resolutions in CT images, a diversity arising from the utilization of assorted scanning equipment. Typically, predictions are made on single slices which are then combined for a comprehensive outcome. Yet, this method does not incorporate learning features specific to each slice, leading to a compromise in effectiveness. To address these challenges, we propose an advanced Spatial-Slice Feature Learning (SSFL++) framework specifically tailored for CT scans. It aims to filter out out-of-distribution (OOD) data within the entire CT scan, allowing us to select essential spatial-slice features for analysis by reducing data redundancy by 70\%. Additionally, we introduce a Kernel-Density-based slice Sampling (KDS) method to enhance stability during training and inference phases, thereby accelerating convergence and enhancing overall performance. Remarkably, our experiments reveal that our model achieves promising results with a simple EfficientNet-2D (E2D) model. The effectiveness of our approach is confirmed on the COVID-19-CT-DB datasets provided by the DEF-AI-MIA workshop.

Using smartphone TechnolOGy to support an EffecTive Home ExeRcise intervention to prevent falls amongst community dwelling older adults The TOGETHER feasibility RCT.

Introduction Falls have major implications for quality-of-life, independence and cost to health services. Strength and balance training has been found to be effective in reducing the rate/risk of falls, as long as there is adequate fidelity to the evidence-based programme. The aims of this study were to: (1) assess the feasibility of using the 'Motivate Me' and 'My Activity Programme' intervention to support falls rehabilitation when delivered in practice (2) assess study design and trial procedures for the evaluation of the intervention. Methods A two-arm, pragmatic feasibility randomised controlled trial was conducted with five health service providers in the UK. Patients aged 50+ years eligible for a falls rehabilitation exercise programme from community services were recruited and received either: (1) standard service with a smartphone for outcome measurement only or (2) standard service plus the 'Motivate Me' and 'My Activity Programme' apps. The primary outcome was feasibility of the intervention, study design and procedures (including recruitment rate, adherence and drop-out). Outcome measures include balance, function, falls, strength, fear of falling, health related quality of life, resource use and adherence, measured at baseline, three and six month post-randomisation. Blinded assessors collected the outcome measures. Results 24 patients were randomised to control group, 26 to intervention group, mean age 77.6 (Range 62 to 92) years. We recruited 37.5% of eligible participants across the five clinical sites. 77% in the intervention group completed their full exercise programme (including the use of the app). Response rate for outcome measures at six months were 77%-80% across outcome measures, but this was effected by the COVID19 pandemic. There was a mean 2.6 ± 1.9 point difference between groups in change in BERG balance score from baseline to three months and mean 4.4 ± 2.7 point difference from baseline to six months in favour of the intervention group. Less falls (1.8 ± 2.8 vs 9.1 ± 32.6) and less injurious falls (0.1 ± 0.5 vs 0.4 ± 0.6) in the intervention group and higher adherence scores at three (17.7 ± 6.8 vs 13.1 ± 6.5) and six months (15.3 ± 7.8 vs 14.9 ± 7.8). There were no related adverse events. Health professionals and patients had few technical issues with the apps. Conclusions The motivational apps and trial procedures were feasible for health professionals and patients. There are positive indications from outcome measures in the feasibility trial and key criteria for progression to full trial were met.

The influence of changes in the Chinese COVID-19 prevention and control policies on mental health of medical staff: A network analysis.

BACKGROUND: Shortly after the first outbreak of COVID-19 in Wuhan, the disease spread rapidly around the world. Previous findings reported an increase in mental health problems among Chinese medical staff, but there was a lack of research following changes in COVID-19 prevention and control policies. METHODS: Medical staff were recruited separately in China from 15 to 16 December 2022 (N = 765, wave 1) and from 5 to 8 January 2023 (N = 690, wave 2). All participants completed the assessments of Generalized Anxiety Disorder-7, the Patient Health Questionnaire-9 and the Euthymia Scale. Network analysis was used to explore the relationships between symptoms both within and across depression, anxiety and euthymia. RESULTS: Medical staff showed worse anxiety, depression and euthymia at wave 2 than at wave 1. Depression, motor, restlessness and uncontrollable worrying showed high centrality (i.e., strength, expected influence, closeness) at wave 1, but higher at peak. Meanwhile, motor symptoms and restlessness showed the strongest connection between different mental disorders at both wave 1 and wave 2. The network structure was stable over time after the relaxation of the infection policy. LIMITATIONS: Our participants were not a random sample and the assessments were based on self-reports. CONCLUSIONS: This study indicated the changes in central and bridging symptoms in medical staff at different stages after lifting of restrictions and the withdrawal of testing requirements, which provided management suggestions for the Chinese government and hospitals, as well as clinical guidance for psychological interventions.

Prediction From Minimal Experience: How People Predict the Duration of an Ongoing Epidemic.

People are known for good predictions in domains they have rich experience with, such as everyday statistics and intuitive physics. But how well can they predict for problems they lack experience with, such as the duration of an ongoing epidemic caused by a new virus? Amid the first wave of COVID-19 in China, we conducted an online diary study, asking each of over 400 participants to predict the remaining duration of the epidemic, once per day for 14 days. Participants' predictions reflected a reasonable use of publicly available information but were meanwhile biased, subject to the influence of negative affect and future time perspectives. Computational modeling revealed that participants neither relied on prior distributions of epidemic durations as in inferring everyday statistics, nor on mechanistic simulations of epidemic dynamics as in computing intuitive physics. Instead, with minimal experience, participants' predictions were best explained by similarity-based generalization of the temporal pattern of epidemic statistics. In two control experiments, we further confirmed that such cognitive algorithm is not specific to the epidemic scenario and that minimal and rich experience do lead to different prediction behaviors for the same observations. We conclude that people generalize patterns in recent history to predict the future under minimal experience.

Integrated safety analysis of baricitinib in adults with severe alopecia areata from two randomized clinical trials.

BACKGROUND: Baricitinib, an oral, selective, reversible Janus kinase (JAK)1/JAK2 inhibitor, is an approved treatment for adults with severe alopecia areata (AA) in the USA, European Union and Japan. OBJECTIVES: To report safety data for baricitinib in patients with severe AA from two clinical trials including long-term extension periods. METHODS: This analysis includes pooled patient-level safety data from two trials, an adaptive phase II/III trial (BRAVE-AA1) and a phase III trial (BRAVE-AA2) (ClinicalTrials.gov, NCT03570749 and NCT03899259). Data are reported in three datasets: (i) the placebo-controlled dataset (up to week 36): baricitinib 2 mg and 4 mg vs. placebo; (ii) the extended dataset (up to the data cutoff): patients remaining on continuous treatment with baricitinib 2 mg or 4 mg from baseline; and (iii) the all-baricitinib dataset (all-BARI, up to the data cutoff): all patients receiving any dose of baricitinib at any time during the trials. Safety outcomes include treatment-emergent adverse events (TEAEs), adverse events of special interest and abnormal laboratory changes. Proportions of patients with events and incidence rates (IR) were calculated. RESULTS: Data were collected for 1303 patients who were given baricitinib, reflecting 1868 patient-years of exposure (median 532 days). The most frequently reported TEAEs during the placebo-controlled period (based on the baricitinib 4-mg group) were upper respiratory tract infection, nasopharyngitis, headache, acne and elevated blood creatine phosphokinase (CPK). During the placebo-controlled period, the frequency of acne was higher with baricitinib than placebo, and elevated CPK was higher with baricitinib 4 mg than placebo and baricitinib 2 mg. In all-BARI, the IR of serious infections was low (n = 16, IR 0.8). There was one opportunistic infection (IR 0.1), and 34 cases of herpes zoster (IR 1.8). There was one positively adjudicated major adverse cardiovascular event (myocardial infarction) (IR 0.1), one pulmonary embolism (IR 0.1), three malignancies other than nonmelanoma skin cancer (IR 0.2) and one gastrointestinal perforation (IR 0.1). No deaths were reported. CONCLUSIONS: This integrated safety analysis in patients with severe AA is consistent with the overall safety profile of baricitinib. Some differences with atopic dermatitis were noted that may be attributable to the disease characteristics of AA.

The associations between cognitive function and depressive symptoms among older Chinese population: A cohort study.

With the rapid development of society, population aging has emerged as a significant global challenge. This study aimed to evaluate the impact of baseline cognitive performance, current cognitive function, and cognitive decline on subsequent depressive symptoms. Data were obtained from participants aged 65 years and older in the Chinese Longitudinal Healthy Longevity Survey (CLHLS), Wave 2014-2018. Of the 7,192 participants in Wave 2014, 1,627 were included in the analysis. Multivariate regressions were conducted to estimate the associations between cognitive measures and depressive symptoms. Our results indicated that baseline cognitive function was not associated with subsequent depressive symptoms, but current cognitive function was. Furthermore, participants who experienced significant cognitive decline were more likely to develop depressive symptoms. Covariates, including marital status, economic status, physical activity, and recreational activity, were also associated with subsequent depressive symptoms. These findings suggest that slowing cognitive decline is an effective strategy for preventing depressive symptoms in older adults, promoting their health and wellbeing.

Organoids to Remodel SARS-CoV-2 Research: Updates, Limitations and Perspectives.

The novel COVID-19 pneumonia caused by the SARS-CoV-2 virus poses a significant threat to human health. Scientists have made significant efforts to control this virus, consequently leading to the development of novel research methods. Traditional animal and 2D cell line models might not be suitable for large-scale applications in SARS-CoV-2 research owing to their limitations. As an emerging modelling method, organoids have been applied in the study of various diseases. Their advantages include their ability to closely mirror human physiology, ease of cultivation, low cost, and high reliability; thus, they are considered to be a suitable choice to further the research on SARS-CoV-2. During the course of various studies, SARS-CoV-2 was shown to infect a variety of organoid models, exhibiting changes similar to those observed in humans. This review summarises the various organoid models used in SARS-CoV-2 research, revealing the molecular mechanisms of viral infection and exploring the drug screening tests and vaccine research that have relied on organoid models, hence illustrating the role of organoids in remodelling SARS-CoV-2 research.

Preschoolers' screen time in China before and during COVID-19 lockdown: a parental survey.

OBJECTIVES: Since COVID-19 first appeared in Wuhan, China, in December 2019, the disease has spread rapidly across the country. To reduce the spread of infection, schools including kindergartens were closed. Prolonged home confinement can affect children's behaviour. Therefore, we investigated the change of preschoolers' total daily screen time during the COVID-19 lockdown in China. DESIGN, SETTING AND PARTICIPANTS: In the parental survey, 1121 preschoolers, whose parents or grandparents completed an online survey from 1 June 2020 to 5 June 2020, were enrolled. PRIMARY OUTCOME: Total daily screen time. Multivariable modelling was performed to identify factors associated with increased screen time. RESULTS: The results showed that preschoolers' total daily screen time during lockdown(median 2.5 hours, IQR 2.5 hours) was significantly longer than it was before lockdown (median 1.5 hours, IQR 1.0 hour). Older age (OR 1.26, 95% CI 1.07 to 1.48), higher annual household income (OR 1.18, 95% CI 1.04 to 1.34),and decrease of moderate-vigorous physical activity (OR 1.41, 95% CI 1.20 to 1.66) were independently associated with increased screen time. CONCLUSIONS: Preschoolers' total daily screen time significantly increased during lockdown.

Anxiety and Depression Among College Students During the COVID-19 Lockdown: A Cross-Sectional Survey in Jiangsu, China.

Background: This study aimed to explore the levels of anxiety and depression in a sample of home-quarantined college students to identify the risk factors for psychological distress during the COVID-19 lockdown. Methods: A total of 1156 college students studying in Jiangsu, China, participated from August 5 to August 14. A structured questionnaire was used to collect data anonymously, including demographic characteristics, the generalized anxiety disorder-7 (GAD-7) questionnaire, the patient health questionnaire-9 (PHQ-9), a physical activity scale, and items related to COVID-19. The chi-square test and Fisher's exact test were employed to establish the differences in levels of anxiety and depression across sociodemographic characteristics. Binary logistic regression was applied to measure the predictors of levels of anxiety and depression, and associations were considered significant at p <0.05. Results: The estimates of anxiety and depression were 48.1% and 57.6%, respectively. The univariate analysis indicated a significant difference in anxiety levels across student grades, whether the student was an only child, the distance from the worst-hit areas, and physical activity/exercise intensity. Physical activity intensity and living in communities with infected people were statistically correlated with the level of depression. Binary logistic regression results indicated that the predictive factors for anxiety were living within a short distance from the worst-hit areas (10~20 km), engaging in higher education (graduate students), and low-intensity daily exercise. Factors statistically predicting depression symptoms were having siblings, a COVID-19 diagnosis in the community and low-intensity daily exercise. Conclusion: During outbreaks, students in an extremely stressful state are more likely to develop anxiety and depression, particularly postgraduates. Psychological interventions to reduce fears and encourage exercise should be available to home-quarantined college students. Students who live in the worst-hit areas and are not the only child in the family should be prioritized.

Robust T cell responses to Pfizer/BioNTech vaccine compared to infection and evidence of attenuated peripheral CD8+ T cell responses due to COVID-19

T cells are a critical component of the response to SARS-CoV-2, but their kinetics after infection and vaccination are insufficiently understood. Using "spheromer” peptide-MHC multimer reagents, we analyzed healthy subjects receiving two doses of the Pfizer/BioNTech BNT162b2 vaccine. Vaccination resulted in robust Spike-specific T cell responses for the dominant CD4+ (HLA-DRB1∗15:01/S191) and CD8+ (HLA-A∗02/S691) T cell epitopes. Antigen-specific CD4+ and CD8+ T cell responses were asynchronous, with the peak CD4+ T cell responses occurring one week post the second vaccination (boost), whereas CD8+ T cells peaked two weeks later. These peripheral T cell responses were elevated compared to COVID-19 patients. We also found that prior SARS-CoV-2 infection resulted in decreased CD8+ T cell activation and expansion, suggesting that prior infection can influence the T cell response to vaccination. Graphical Our understanding of T cell responses in COVID-19 and vaccination is incomplete. Gao et al. examine SARS-CoV-2-specific T cell responses to infection and vaccination, revealing disparate kinetics between CD4+ and CD8+ T cells. Furthermore, compared to vaccination alone, circulating CD8+ T cells are attenuated during infection and in subsequent vaccination.

Therapeutic drug combinations against COVID-19 obtained by employing a collaborative filtering method

The outbreak of coronavirus disease 2019 (COVID-19) has severely harmed human society and health. Because there is currently no specific drug for the treatment and prevention of COVID-19, we used a collaborative filtering algorithm to predict which traditional Chinese medicines (TCMs) would be effective in combination for the prevention and treatment of COVID-19. First, we performed drug screening based on the receptor structure prediction method, molecular docking using q-vina to measure the binding ability of TCMs, TCM formulas, and neo-coronavirus proteins, and then performed synergistic filtering based on Laplace matrix calculations to predict potentially effective TCM formulas. Combining the results of molecular docking and synergistic filtering, the new recommended formulas were analyzed by reviewing data platforms or tools such as PubMed, Herbnet, the Traditional Chinese Medicine Systems Pharmacology (TCMSP) database, the Guide to the Dispensing of Medicines for Clinical Evidence, and the Dictionary of Chinese Medicine Formulas, as well as medical experts' treatment consensus in terms of herbal efficacy, modern pharmacological studies, and clinical identification and typing of COVID-19 pneumonia, to determine the recommended solutions. We found that the therapeutic effect of a combination of six TCM formulas on the COVID-19 virus is the result of the overall effect of the formula rather than that of specific components of the formula. Based on this, we recommend a formula similar to that of Jinhua Qinggan Granules for the treatment of COVID-19 pneumonia. This study may provide new ideas and new methods for future clinical research. Classification Biological Science.

Delivery of spike-RBD by bacterial type three secretion system for SARS-CoV-2 vaccine development.

COVID-19 pandemic continues to spread throughout the world with an urgent demand for a safe and protective vaccine to effectuate herd protection and control the spread of SARS-CoV-2. Here, we report the development of a bacterial vector COVID-19 vaccine (aPA-RBD) that carries the gene for the receptor-binding domain (RBD) of the SARS-CoV-2 spike protein. Live-attenuated strains of Pseudomonas aeruginosa (aPA) were constructed which express the recombinant RBD and effectively deliver RBD protein into various antigen presenting cells through bacterial type 3 secretion system (T3SS) in vitro. In mice, two-dose of intranasal aPA-RBD vaccinations elicited the development of RBD-specific serum IgG and IgM. Importantly, the sera from the immunized mice were able to neutralize host cell infections by SARS-CoV-2 pseudovirus as well as the authentic virus variants potently. T-cell responses of immunized mice were assessed by enzyme-linked immunospot (ELISPOT) and intracellular cytokine staining (ICS) assays. aPA-RBD vaccinations can elicit RBD-specific CD4+and CD8+T cell responses. T3SS-based RBD intracellular delivery heightens the efficiency of antigen presentation and enables the aPA-RBD vaccine to elicit CD8+T cell response. Thus, aPA vector has the potential as an inexpensive, readily manufactured, and respiratory tract vaccination route vaccine platform for other pathogens.

Therapeutic drug combinations against COVID-19 obtained by employing a collaborative filtering method.

The outbreak of coronavirus disease 2019 (COVID-19) has severely harmed human society and health. Because there is currently no specific drug for the treatment and prevention of COVID-19, we used a collaborative filtering algorithm to predict which traditional Chinese medicines (TCMs) would be effective in combination for the prevention and treatment of COVID-19. First, we performed drug screening based on the receptor structure prediction method, molecular docking using q-vina to measure the binding ability of TCMs, TCM formulas, and neo-coronavirus proteins, and then performed synergistic filtering based on Laplace matrix calculations to predict potentially effective TCM formulas. Combining the results of molecular docking and synergistic filtering, the new recommended formulas were analyzed by reviewing data platforms or tools such as PubMed, Herbnet, the Traditional Chinese Medicine Systems Pharmacology (TCMSP) database, the Guide to the Dispensing of Medicines for Clinical Evidence, and the Dictionary of Chinese Medicine Formulas, as well as medical experts' treatment consensus in terms of herbal efficacy, modern pharmacological studies, and clinical identification and typing of COVID-19 pneumonia, to determine the recommended solutions. We found that the therapeutic effect of a combination of six TCM formulas on the COVID-19 virus is the result of the overall effect of the formula rather than that of specific components of the formula. Based on this, we recommend a formula similar to that of Jinhua Qinggan Granules for the treatment of COVID-19 pneumonia. This study may provide new ideas and new methods for future clinical research. Classification: Biological Science.

Network analysis of COVID-19-related PTSD and depressive symptoms among Chinese general population and comorbidity subgroup.

OBJECTIVE: Posttraumatic stress disorder (PTSD) and depressive symptoms are prevalent and highly correlated mental health outcomes of traumatic events, but their comorbidity in the COVID-19 pandemic has not been examined with evidence from Chinese culture. METHOD: With information from 2,858 Chinese adults, this study used a network analysis to investigate the relationship between PTSD and depressive symptoms along with their symptoms-structure associations. RESULTS: Results indicated similar positive connections among similar cluster symptoms (i.e., positive affect) in both general and comorbidity-reported populations. Self-destructive/reckless behaviors were core symptoms in the general population, and interpersonal difficulties were core symptoms in the comorbidity subgroup. Finally, the very strong communication seen between "arousal and reactivity alterations" and "depressed affect" deserves more attention. CONCLUSION: Our results indicated interpersonal symptoms can be important targets when intervening or treating PTSD and depressive symptoms related to COVID-19 in the clinical population. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Spheromers reveal robust T cell responses to the Pfizer/BioNTech vaccine and attenuated peripheral CD8+ T cell responses post SARS-CoV-2 infection.

T cells are a critical component of the response to SARS-CoV-2, but their kinetics after infection and vaccination are insufficiently understood. Using "spheromer" peptide-MHC multimer reagents, we analyzed healthy subjects receiving two doses of the Pfizer/BioNTech BNT162b2 vaccine. Vaccination resulted in robust spike-specific T cell responses for the dominant CD4+ (HLA-DRB1∗15:01/S191) and CD8+ (HLA-A∗02/S691) T cell epitopes. Antigen-specific CD4+ and CD8+ T cell responses were asynchronous, with the peak CD4+ T cell responses occurring 1 week post the second vaccination (boost), whereas CD8+ T cells peaked 2 weeks later. These peripheral T cell responses were elevated compared with COVID-19 patients. We also found that previous SARS-CoV-2 infection resulted in decreased CD8+ T cell activation and expansion, suggesting that previous infection can influence the T cell response to vaccination.

Reinfection with SARS-CoV-2 and Waning Humoral Immunity: A Case Report.

Recovery from COVID-19 is associated with production of anti-SARS-CoV-2 antibodies, but it is uncertain whether these confer immunity. We describe viral RNA shedding duration in hospitalized patients and identify patients with recurrent shedding. We sequenced viruses from two distinct episodes of symptomatic COVID-19 separated by 144 days in a single patient, to conclusively describe reinfection with a different strain harboring the spike variant D614G. This case of reinfection was one of the first cases of reinfection reported in 2020. With antibody, B cell and T cell analytics, we show correlates of adaptive immunity at reinfection, including a differential response in neutralizing antibodies to a D614G pseudovirus. Finally, we discuss implications for vaccine programs and begin to define benchmarks for protection against reinfection from SARS-CoV-2.

An atlas of the blood virome in healthy individuals.

Emerging evidence indicates that gut virome plays a role in human health and disease, however, much less is known about the viral communities in blood. Here we conducted a direct metatranscriptomic sequencing of virus-like-particles in blood from 1200 healthy individuals, without prior amplification to avoid potential amplification bias and with a strictly bioinformatic and manual check for candidate viral reads to reduce false-positive matches. We identified 55 different viruses from 36 viral families, including 24 human DNA, RNA and retroviruses in 70% of the studied pools. The study showed that anelloviruses are widely distributed and dominate the blood virome in healthy individuals. Human herpesviruses and pegivirus-1 are commonly prevalent in asymptomatic humans. We identified the prevalence of RNA viruses often causing acute infection, like HEV, HPIV, RSV and HCoV-HKU1, revealing of a transmissible risk of asymptomatic infection. Several viruses possible related to transfusion safety were identified, including human Merkel cell polyomavirus, papillomavirus, parvovirus B19 and herpesvirus 8 in addition to HBV. In addition, phages in Caudovirales and Microviridae, were commonly found in pools of samples with a very low abundance; a few sequences for invertebrate, plant and giant viruses were found in some of individuals; however, the remaining 31 viruses mostly reflect extensive contamination from commercial reagents and the work environments. In conclusion, this study is the first comprehensive investigation of blood virome in healthy individuals by metatranscriptomic sequencing of VLP in China. Further investigation of potential false positives representing a major challenge for the identification of novel viruses in mNGS, will offer a systemic idea and means to reveal true viral infections of human.

COVID-19 Related Early Google Search Behavior and Health Communication in the United States: Panel Data Analysis on Health Measures.

The COVID-19 outbreak at the end of December 2019 spread rapidly all around the world. The objective of this study is to investigate and understand the relationship between public health measures and the development of the pandemic through Google search behaviors in the United States. Our collected data includes Google search queries related to COVID-19 from 1 January to 4 April 2020. After using unit root tests (ADF test and PP test) to examine the stationary and a Hausman test to choose a random effect model, a panel data analysis is conducted to investigate the key query terms with the newly added cases. In addition, a full sample regression and two sub-sample regressions are proposed to explain: (1) The changes in COVID-19 cases number are partly related to search variables related to treatments and medical resources, such as ventilators, hospitals, and masks, which correlate positively with the number of new cases. In contrast, regarding public health measures, social distancing, lockdown, stay-at-home, and self-isolation measures were negatively associated with the number of new cases in the US. (2) In mild states, which ranked one to twenty by the average daily new cases from least to most in 50 states, the query terms about public health measures (quarantine, lockdown, and self-isolation) have a significant negative correlation with the number of new cases. However, only the query terms about lockdown and self-isolation are also negatively associated with the number of new cases in serious states (states ranking 31 to 50). Furthermore, public health measures taken by the government during the COVID-19 outbreak are closely related to the situation of controlling the pandemic.