CMC_v2: Towards More Accurate COVID-19 Detection with Discriminative Video Priors

This paper presents our solution for the 2nd COVID-19 Competition, occurring in the framework of the AIMIA Workshop at the European Conference on Computer Vision (ECCV 2022). In our approach, we employ the winning solution last year which uses a strong 3D Contrastive Mixup Classification network (CMC_v1) as the baseline method, composed of contrastive representation learning and mixup classification. In this paper, we propose CMC_v2 by introducing natural video priors to COVID-19 diagnosis. Specifically, we adapt a pre-trained (on video dataset) video transformer backbone to COVID-19 detection. Moreover, advanced training strategies, including hybrid mixup and cutmix, slice-level augmentation, and small resolution training are also utilized to boost the robustness and the generalization ability of the model. Among 14 participating teams, CMC_v2 ranked 1st in the 2nd COVID-19 Competition with an average Macro F1 Score of 89.11%. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Boosting COVID-19 Severity Detection with Infection-Aware Contrastive Mixup Classification

This paper presents our solution for the 2nd COVID-19 Severity Detection Competition. This task aims to distinguish the Mild, Moderate, Severe, and Critical grades in COVID-19 chest CT images. In our approach, we devise a novel infection-aware 3D Contrastive Mixup Classification network for severity grading. Specifically, we train two segmentation networks to first extract the lung region and then the inner lesion region. The lesion segmentation mask serves as complementary information for the original CT slices. To relieve the issue of imbalanced data distribution, we further improve the advanced Contrastive Mixup Classification network by weighted cross-entropy loss. On the COVID-19 severity detection leaderboard, our approach won the first place with a Macro F1 Score of 51.76%. It significantly outperforms the baseline method by over 11.46%. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Photo validation study using cutaneous dermatomyositis disease area and severity index in dermatomyositis patients

The Coronavirus disease 2019 (COVID-19) pandemic revealed our need for reliable tools to evaluate patients with skin disease virtually. Thus far, there has not been a study that has attempted to score the Cutaneous Dermatomyositis Disease Area and Severity Index (CDASI), a validated outcome measure of skin activity and damage, from photographs. In this study, patients were prospectively recruited during routine clinic visits and skin regions used in scoring the CDASI were photographed by research staff using two iPhone cameras (an iPhone 8 and iPhone 11). Two dermatologists served as the raters. The in-person CDASI assessment was scored by rater 1 at the clinic visit and the photographs were scored at a later date by both rater 1 and rater 2. Of the 34 patients participating in the study, 82.3% were female, 85.3% were Caucasian with a mean age of 54 years (SD=12). For the total activity score, the intraclass correlation coefficient (ICC) between rater 1’s in-person assessment compared to photograph assessment was 0.806 (95% CI 0.649-0.898 p<0.001) and was 0.822 (95% CI 0.675-0.907 p<0.001) between rater 2 and the in-person assessment. For the total damage score, the ICC between rater 1 and the in-person assessment was 0.54 (95% CI 0.254-0.739 p=0.004) and was 0.601 (95% CI 0.338-0.778 p<0.001) between rater 2 and the in-person assessment. The reliability was interpreted as “excellent” for skin activity, an important measure in clinical trials for dermatomyositis. Photographs may be a useful tool for evaluating clinical trial patients in the future. More research is needed to determine innovations for improving our ability to evaluate skin activity through photographs such as the use of a color checker card or color correction algorithm.

AN ANALYSIS FOR THE WORK RESUMPTION UNDER THE COVID-19 EPIDEMIC BASED ON VIIRS-DNB NIGHTTIME LIGHTS IN CHINA

Timely and effective quantitative measurement of enterprises' offline resumption of work after public emergencies is conducive to the formulation and implementation of relevant policies. In this paper, we analyze the level of work resumption after the coronavirus disease 2019 (COVID-19)-influenced Chinese Spring Festival in 2020 with National Polar-orbiting Partnership Visible Infrared Imaging Radiometer Suite (NPP-VIIRS) Day/Night Band (DNB) daily data. The results demonstrate that COVID-19 has seriously affected the resumption of work after the Spring Festival holiday. Since February 10th, work has been resuming in localities. By late March, the work resumption indexes of most cities exceeded 50%, and Shanghai and Nanjing even had achieved complete resumption of work. Our method effectively estimates the resumption of work, which provides a scientific basis for local governments to formulate subsequent resumption policies. © 2021 IEEE

CMC-COV19D: Contrastive Mixup Classification for COVID-19 Diagnosis

Deep learning methods have been extensively investigated for rapid and precise computer-aided diagnosis during the outbreak of the COVID-19 epidemic. However, there are still remaining issues to be addressed, such as distinguishing COVID-19 in the complex scenario of multi-type pneumonia classification. In this paper, we aim to boost the COVID-19 diagnostic performance with more discriminative deep representations of COVID and non-COVID categories. We propose a novel COVID-19 diagnosis approach with contrastive representation learning to effectively capture the intra-class similarity and inter-class difference. Besides, we design an adaptive joint training strategy to integrate the classification loss, mixup loss, and contrastive loss. Through the joint loss function, we obtain the high-level representations which are highly discriminative in COVID-19 screening. Extensive experiments on two chest CT image datasets, i.e., CC-CCII dataset and COV19-CT-DB database, demonstrate the effectiveness of our proposed approach in COVID-19 diagnosis. Our method won the first prize in the ICCV 2021 Covid-19 Diagnosis Competition of AI-enabled Medical Image Analysis Workshop. Our code is publicly available at https://github.com/houjunlin/Team-FDVTS-COVID-Solution.

Multi-DeepNet: A Novel Weakly-Supervised Multi-Task and Multi-View-Oriented Convolution Neural Network for COVID-19 Diagnosis from CT Images

Currently, manual analysis performed by professional radiologists is required for COVID-19 diagnosis given the patient's chest Computed Tomography (CT) images, but this process is inefficient and costly. Deep learning methods can provide computer vision-based solutions to help guide radiologists perform faster and more accurate diagnosis. However, current well performed methods require training on large and balanced datasets with pixel level lung lesion annotations, both of which are not easily accessible. Moreover, visual similarities between COVID-19 and other pneumonia in CT scans make it difficult to learn their distinguishing features. To address these issues, we propose a novel weakly-supervised deep learning model, named Multi-DeepNet, that can be well trained to perform fine-grained classification on small and imbalanced datasets. Specifically, a multi-task pre-training module is introduced to better extract distinguishing features between COVID-19 and other similar pneumonia. Furthermore, a multi-view-oriented classifier is proposed to extract complimentary information from the axial, coronal and sagittal planes. Experimental results demonstrate that our Multi-DeepNet achieves superior sensitivities, specificity, and accuracies compared to state-of-the-art methods. © 2021 IEEE.

Agent-Based Campus Novel Coronavirus Infection and Control Simulation

Corona Virus Disease 2019 (COVID-19), due to its extremely high infectivity, has been spreading rapidly around the world and bringing huge influence to socioeconomic development and people's daily life. Taking for example the virus transmission that may occur after college students return to school, we analyze the quantitative influence of the key factors on the virus spread, including crowd density and self-protection. One Campus Virus Infection and Control Simulation (CVICS) model of the novel coronavirus is proposed in this article, fully considering the characteristics of repeated contact and strong mobility of crowd in the closed environment. Specifically, we build an agent-based infection model, introduce the mean field theory to calculate the probability of virus transmission, and microsimulate the daily prevalence of infection among individuals. The experimental results show that the proposed model in this article efficiently simulates how the virus spreads in the dense crowd in frequent contact under a closed environment. Furthermore, preventive and control measures, such as self-protection, crowd decentralization, and isolation during the epidemic, can effectively delay the arrival of infection peak, reduce the prevalence, and, finally, lower the risk of COVID-19 transmission after the students return to school. IEEE

Tolerance of nasal and oral povidone-iodine antisepsis amid COVID-19 pandemic

Introduction: Because of the high risk to the otolaryngology and dental communities during the COVID-19 pandemic, protocols involving topical antisepsis of the nasal and oral cavities were developed for use. Studies have demonstrated efficy of low-dose povidone-iodine (PVP-I) to inactivate SARSCoV- 2 in 15 seconds as well as safety for use in the nasal and oral cavities. We evaluated patient tolerability of this solution when used in the outpatient office setting prior to nasal and oral procedures. Method: A prospective study was conducted. Data were collected between October 2020 and January 2021. Via cotton pledgets, 0.5% PVP-I was applied to the nasal cavities of patients undergoing transnasal procedures in the otolaryngology office. Dental patients rinsed with 0.5% PVP-I prior to prosthodontic exam. A survey was administered regarding tolerability of the solution, including nasal irritation, oral irritation, sneezing, coughing, changes in taste, changes in smell, and agreeability to use the solution again on a scale of 0 to 3. The initial survey was administered on the day of visit and the follow-up survey 7 to 32 days later. In total, 56 patients were recruited. Follow-up data were collected from 19 patients. Prevalence of symptoms was calculated. Results: On the date of the visit, 59% of patients reported 1 of the symptoms. Most of these symptoms were mild, with 1 nasal patient reporting a moderate amount of nasal irritation and 1 oral patient reporting a moderate taste disturbance;16% of patients noted a symptom on follow-up survey conducted a mean of 13 days later. Initially, 88% of patients reported that they would definitely or probably agree to future use of the solution. On follow-up survey;95% of patients reported that they would definitely or probably agree to future use of the solution. Conclusion: Most patients tolerate commercial low-dose nasal and oral PVP-I solutions well with mainly mild symptoms reported. Few symptoms were noted on follow-up. Importantly, a large majority of patients would agree to repeat application of the solution, suggesting that diluted PVP-I has potential for large-scale routine use in otolaryngology and dental settings.

Early platelet counts distinguish viral and bacterial pulmonary sepsis

Rationale: Respiratory viruses are commonly detected pathogens in pulmonary sepsis. Prior studies have demonstrated that patients with respiratory viral infections may have transient lymphocytopenia and thrombocytopenia. Leukocyte parameters including lymphocyte to monocyte ratio (LMR) and neutrophil to lymphocyte ratio (NLR) have been reported as screening tools for viral infections. Platelet counts and dynamics have been described as quantitative traits for ARDS risk and mortality. Therefore, we hypothesized that early hematologic parameters including lymphocyte count, monocyte count, platelet count, NLR, and LMR may distinguish viral from bacterial pulmonary sepsis. Methods: We enrolled 1,158 critically ill patients with pulmonary sepsis from 2009 to 2020 and measured lymphocyte count, monocyte count, platelet count, NLR, and LMR on ICU admission and at 24-hrs. Respiratory viruses were detected via PCR panel on nasopharyngeal swabs. Pulmonary sepsis was adjudicated by a physician panel. APACHE III scores were collected during the first 24-hrs. Shock was assessed by vasopressor use or mean arterial pressure <65mmHg despite 30cc/kg fluid resuscitation. ARDS was defined per Berlin criteria. We assessed mortality at 30 days. We used multivariable linear regression to test the association between each of the laboratory studies and a positive respiratory pathogen panel (RVP) adjusting for APACHE III score, age, sex, malignancy, and race. We used multivariable logistic regression to assess for associations between a positive RVP and outcomes. Results: The incidence of respiratory virus detection was 33.9%. The incidence of ARDS and mortality were 52.7% and 49.0%, respectively. The most commonly detected pathogens were SARS-CoV-2 and rhinovirus (Table 1). Lower platelet counts at 24-hrs were significantly associated with respiratory virus detection (β-41.59 × 109/L [95%CI-79.03,-4.15], p=0.03), whereas admission platelet counts were not significantly associated (β-22.38 × 109/L [95%CI-63.26, 20.49], p=0.32). The significant association at 24-hrs was also present on sensitivity analyses excluding patients with SARS-CoV-2. There were no statistically significant differences between the populations with respect to lymphocyte count, monocyte count, NLR, LMR, ARDS, shock, and mortality. Conclusion: Lower early platelet counts were identified in patients with viral pulmonary sepsis. Although LMR and NLR have been reported as screening tools for viral infections in non-critically ill populations, we did not detect significant associations between lymphocyte count, monocyte count, NLR or LMR and viral detection in pulmonary sepsis. Our findings suggest that platelet counts in combination with other validated parameters may warrant further investigation for the early discrimination of viral versus bacterial pulmonary sepsis.

Investigation of the mental health of patients with COVID-19

Objective: To evaluate the status and influencing factors of the mental health of patients with COVID-19 during isolation treatment. Methods: From February 2nd to 16th, 2020, 106 COVID-19 patients were anonymously investigated for their mental health status using online questionnaires (including Patient Health Questionnaire-9, Generalized Anxiety Disorder Scale-7, Patient Health Questionnaire-15 scales). The patients were from Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology. After data processing, SPSS19.0 was used for statistical analysis. Results: Of the 106 COVID-19 patients, 46 were male and 60 were female, with an age of (35.90±11.92) years. The detection rates of depression, anxiety and somatic symptoms in those patients were 49.06% (52/106), 56.60% (60/106) and 69.81% (74/106) respectively. The severe cases of depression (scale score greater than 19), anxiety (scale score greater than 14) and somatic symptoms (scale score greater than 14) accounted for 9.43% (10/106), 15.09% (16/106) and 20.75% (22/106) respectively. In addition, 67.92% (72/106) of the patients had sleep problems, 24.53% (26/106) had self-mutilating or suicidal thoughts, and 28.30% (30/106) required psychological counseling-all of which were at significantly higher percentages than those of the general population. Only 39.62% (42/106) of the examined patients had neither deression nor anxiety. By using the non-parametric test of rank conversion for analysis, the results showed that both married patients and nucleic-acid-positive patients had more severe depressions, and both married patients and anoxemic patients had more severe somatic symptoms. Conclusion: This mental health assessment showed that depression, anxiety, and various somatic symptoms exist among the COVID-19 patients, and therefore early identification and intervention should be conducted to avoid extreme events such as self-mutilating or suicidal impulsivity, with a greater focus on both married patients and patients with severe symptoms.

CT characteristics of patients infected with 2019 novel coronavirus: association with clinical type.

AIM: To summarise the features of chest computed tomography (CT) of a series of patients infected with 2019 novel coronavirus (2019-nCov) to speed up recognition and have a better understanding of COVID-19 disease. MATERIALS AND METHODS: The clinical information and chest CT images of 93 patients infected with 2019-nCov from multiple centres were reviewed. RESULTS: Of the 93 cases, abnormalities in 91 cases were located at the subpleural level, presenting with ground-glass opacity (GGO; n=69, 74.2%) and consolidation (n=56, 60.2%) in multiple lobes. Other CT features included vascular dilatation (n=83, 89.2%), interlobular septal thickening (n=29, 31.2%), bronchodilatation (n=44, 47.3%), the crazy-paving sign (n=34, 36.6%), the sieve-hole sign (n=12, 12.9%), pleural thickening (n=21, 22.6%), and pleural effusion (n=8, 8.6%). Multiple lobe involvement, including the presence of consolidation, the crazy-paving sign, interlobular septal thickening, pleural thickening and pleural effusion, was more common in critical patients with heavy/critical infection (p<0.05), whereas the presence of GGO, involvement of one or two lobes, and the halo sign were more common in patients with mild/common-type infections (p<0.05). Moreover, older age, higher body temperature, complaints of chest tightness and breathlessness, and lymphopenia was associated with heavy/critical infections. CONCLUSION: The CT and clinical appearances of COVID-19 are variable and reflect the severity of COVID-19 to some extent.

[Comparison of clinical and pathological features between severe acute respiratory syndrome and coronavirus disease 2019].

Severe acute respiratory syndrome (SARS) and coronavirus disease 2019 (COVID-19) shared similar pathogenetic, clinical and pathological features. Fever and cough were the most common symptoms of both diseases, while myalgia and diarrhea were less common in patients with COVID-19. Acute respiratory distress syndrome (ARDS) was the most severe pulmonary complication that caused high mortality rate. Histologically, diffuse alveolar damage (DAD) was the most characteristic finding in non-survivors with either SARS or COVID-19. Cases of patients died less than 10-14 days of disease duration demonstrated acute-phase DAD, while cases beyond 10-14 days of disease duration exhibited organizing-phase DAD in SARS. Meanwhile, organization and fibrosis were usually accompanied by exudation. Coronavirus was mostly detected in pneumocytes, but less in macrophages and bronchiolar epithelial cells. Hemorrhagic necrosis and lymphocyte depletion were found in lymph nodes and spleen in both SARS and COVID-19, indicating a pathological basis of lymphocytopenia. Thrombosis was commonly observed in small vessels and microvasculaturr in lungs accompanying DAD. Microthrombosis was also found in extrapulmonary organs in COVID-19, that was less reported in SARS. Damages in multiple extrapulmonary organs were observed, but coronavirus was not detected in some of those organs, indicating an alternative mechanism beyond viral infection, such as hypoxemia, ischemia and cytokine storm induced immunological injury. DAD due to viral infection and immunological injury, as well as multi-organ dysfunction and extensive microthrombus formation, brought huge challenge to the management of patients with severe SARS or COVID-19.