Interpretation of lung disease classification with light attention connected module.

Lung diseases lead to complications from obstructive diseases, and the COVID-19 pandemic has increased lung disease-related deaths. Medical practitioners use stethoscopes to diagnose lung disease. However, an artificial intelligence model capable of objective judgment is required since the experience and diagnosis of respiratory sounds differ. Therefore, in this study, we propose a lung disease classification model that uses an attention module and deep learning. Respiratory sounds were extracted using log-Mel spectrogram MFCC. Normal and five types of adventitious sounds were effectively classified by improving VGGish and adding a light attention connected module to which the efficient channel attention module (ECA-Net) was applied. The performance of the model was evaluated for accuracy, precision, sensitivity, specificity, f1-score, and balanced accuracy, which were 92.56%, 92.81%, 92.22%, 98.50%, 92.29%, and 95.4%, respectively. We confirmed high performance according to the attention effect. The classification causes of lung diseases were analyzed using gradient-weighted class activation mapping (Grad-CAM), and the performances of their models were compared using open lung sounds measured using a Littmann 3200 stethoscope. The experts' opinions were also included. Our results will contribute to the early diagnosis and interpretation of diseases in patients with lung disease by utilizing algorithms in smart medical stethoscopes.

Very severe immune aplastic anemia after mRNA vaccination against COVID-19 responds well to immunosuppressive therapy: clinical characteristics and comparison to previous reports.

OBJECTIVE: Various hematologic side effects of the Coronavirus Disease 2019 (COVID-19) vaccination has been reported, and most of them are thought to be related to autoimmune pathways. To the best of our knowledge, only few cases of post-COVID-19 vaccination aplastic anemia (AA) have been reported and there is no reported Korean case of COVID-19 vaccine-induced AA yet. We present a case of severe immune-mediated AA that developed after the administration of a messenger ribonucleic acid (mRNA) gene-based spike protein vaccine against COVID-19, which responded well to immunosuppressive therapy, and discuss the probable pathogenesis of AA and the implication of vaccination along with a comparison of previous cases reported. METHODS: A 53-year-old Korean man developed sudden pancytopenia three months after COVID-19 vaccination. To evaluate the cause of pancytopenia, a bone marrow study was performed. RESULTS: A diagnosis of AA was made through the bone marrow study and he received triple immunosuppressive therapy (IST). After triple IST for five months, his blood cell count was improved and maintained without transfusion and his follow-up bone marrow examination showed improved cellularity. CONCLUSION: COVID-19 vaccine might be associated with the development of immune-mediated AA. Prompt hematologic evaluation should be performed when there are symptoms or signs suggestive of cytopenia after COVID-19 vaccination. Although the clinical outcome of post-vaccination AA varies, a good prognosis can be possible for patients with COVID-19 vaccination-induced AA.

COVID-19 CXR Classification: Applying Domain Extension Transfer Learning and Deep Learning

The infectious coronavirus disease-19 (COVID-19) is a viral disease that affects the lungs, which caused great havoc when the epidemic rapidly spread around the world. Polymerase chain reaction (PCR) tests are conducted to screen for COVID-19 and respond to quarantine measures. However, PCR tests take a considerable amount of time to confirm the test results. Therefore, to supplement the accuracy and quickness of a COVID-19 diagnosis, we proposed an effective deep learning methodology as a quarantine response through COVID-19 chest X-ray image classification based on domain extension transfer learning. As part of the data preprocessing, contrast limited adaptive histogram equalization was applied to chest X-ray images using Medical Information Mart for Intensive Care (MIMIC)-IV obtained from the Beth Israel Deaconess Medical Center. The classification of the COVID-19 X-ray images was conducted using a pretrained ResNet-50. We also visualized and interpreted the classification performance of the model through explainable artificial intelligence and performed statistical tests to validate the reliability of the model. The proposed method correctly classified images with 96.7% accuracy, an improvement of about 9.9% over the reference model. This study is expected to help medical staff make an integrated decision in selecting the first confirmed case and contribute to suppressing the spread of the virus in the community.

Association of Lupus Anticoagulant status with Disease Course in SARS-CoV-2 (COVID-19) Infection.

Background: We investigated the importance of lupus anticoagulant (LA) in patients with SARS-CoV-2. Methods: Medical records of 41 SARS-CoV-2 infected patients were reviewed. Patients were classified into two groups according to the frequency of positive LA test results: "LA (-)" and "LA (+) ≥1" (LA positive at least once). Statistical analysis was performed to determine the association between LA presence and change in LA test results and disease course according to both hospital days (HD) and days after diagnosis (DD). Results: The prevalence of LA was 51.2%. Averagely, the first change in LA test result occurred during DD 12-13 and between HD 9-10. The second change occurred on DD 15-16 and HD 13-14. The presence of LA was associated with severe disease (P = .004) but was not associated with thrombotic complications or mortality. The change of results from negative to positive or vice versa or the frequency of the changes was not associated with disease severity, thrombotic complications, or mortality. Conclusions: LA positivity can be regarded as one of the findings suggesting more serious SARS-CoV-2 infection.

Age-Specific Characteristics of Adult and Pediatric Respiratory Viral Infections: A Retrospective Single-Center Study.

This study aimed to identify age-specific characteristics of respiratory viral infections. Hospitalized patients with confirmed viral respiratory infections were included in the sample. The patients were divided into the pediatric group (<19 years old) and the adult group (≥19 years old). The groups were then subdivided based on age: 0-6, 7-12, 13-18, 19-49, 50-64, and ≥65 years old. These groups were compared to evaluate the differences in the pattern of respiratory viral infections. Among a total of 4058 pediatric patients (mean age 3.0 ± 2.9 years, n = 1793 females), 2829 (48.9%) had mono-infections, while 1229 (51.1%) had co-infections. Co-infections were the most common in the 0-6-year-old group (31.6%). Among 1550 adult patients (mean age 70.2 ± 15.3 years, n = 710 females), 1307 (85.6%) had mono-infections and 243 (14.4%) had co-infections. Co-infections were most common in the ≥65-year-old group (16.8%). Viral infection and co-infection rates decreased with age in pediatric patients but increased with increasing age in adults. In pediatric patients, the rates of viral infections and co-infections were high; the rate of co-infections was higher in younger patients. In adult patients, the rates of viral infections and co-infections were lower than those in pediatric patients; the rate of co-infections was higher in older patients.

Analysis of the Association Among Air Pollutants, Allergenic Pollen, and Respiratory Virus Infection of Children in Guri, Korea During Recent 5 Years.

PURPOSE: Concerns about the spread of infectious diseases have increased due to the coronavirus disease pandemic. Knowing the factors that exacerbate or increase the contagiousness of a virus could be a key to pandemic prevention. Therefore, we investigated whether the pandemic potential of infectious diseases correlates with the concentration of atmospheric substances. We also investigated whether environmental deterioration causes an increase in viral infections. METHODS: Pediatric patients (0-18 years old; n = 6,223) were recruited from those hospitalized for aggravated respiratory symptoms at Hanyang University Guri Hospital between January 1, 2015 and December 31, 2019. The number of viral infections was defined as the total number of virus-infected patients hospitalized for respiratory symptoms. We analyzed the association between the number of viral infections/week and the average concentrations of atmospheric substances including particulate matter (PM)10, PM2.5, O3, NO2, CO, SO2, and allergenic pollen) for that week. The cross-correlation coefficient between the weekly measures of pollens and viral infections was checked to determine which time point had the most influence. The association of atmospheric substances in that time, with the number of viral infections/week was investigated using multiple linear regression analysis to identify factors with the greatest influence. RESULTS: In spring the tree pollen average concentration one week earlier (t-1) had the greatest correlation with the average virus infection of a given week (t) (ρXY (h) =0.5210). The number of viral infections showed a statistically significant correlation with especially tree pollen concentration of 1 week prior (adj R²=0.2280). O3 concentration was correlated to the number of viral infections within that week (adj R²=0.2552) in spring, and weed pollen and CO concentration correlated (adj R²=0.1327) in autumn. CONCLUSIONS: Seasonal co-exposure to air pollutants and allergenic pollens may enhance respiratory viral infection susceptibility in children. Therefore, reducing the concentrations of air pollutants and pollens may help prevent future epidemics.