Deep learning attention-guided radiomics for COVID-19 chest radiograph classification.

Background: Accurate assessment of coronavirus disease 2019 (COVID-19) lung involvement through chest radiograph plays an important role in effective management of the infection. This study aims to develop a two-step feature merging method to integrate image features from deep learning and radiomics to differentiate COVID-19, non-COVID-19 pneumonia and normal chest radiographs (CXR). Methods: In this study, a deformable convolutional neural network (deformable CNN) was developed and used as a feature extractor to obtain 1,024-dimensional deep learning latent representation (DLR) features. Then 1,069-dimensional radiomics features were extracted from the region of interest (ROI) guided by deformable CNN's attention. The two feature sets were concatenated to generate a merged feature set for classification. For comparative experiments, the same process has been applied to the DLR-only feature set for verifying the effectiveness of feature concatenation. Results: Using the merged feature set resulted in an overall average accuracy of 91.0% for three-class classification, representing a statistically significant improvement of 0.6% compared to the DLR-only classification. The recall and precision of classification into the COVID-19 class were 0.926 and 0.976, respectively. The feature merging method was shown to significantly improve the classification performance as compared to using only deep learning features, regardless of choice of classifier (P value <0.0001). Three classes' F1-score were 0.892, 0.890, and 0.950 correspondingly (i.e., normal, non-COVID-19 pneumonia, COVID-19). Conclusions: A two-step COVID-19 classification framework integrating information from both DLR and radiomics features (guided by deep learning attention mechanism) has been developed. The proposed feature merging method has been shown to improve the performance of chest radiograph classification as compared to the case of using only deep learning features.

Development and validation of bone-suppressed deep learning classification of COVID-19 presentation in chest radiographs.

Background: Coronavirus disease 2019 (COVID-19) is a pandemic disease. Fast and accurate diagnosis of COVID-19 from chest radiography may enable more efficient allocation of scarce medical resources and hence improved patient outcomes. Deep learning classification of chest radiographs may be a plausible step towards this. We hypothesize that bone suppression of chest radiographs may improve the performance of deep learning classification of COVID-19 phenomena in chest radiographs. Methods: Two bone suppression methods (Gusarev et al. and Rajaraman et al.) were implemented. The Gusarev and Rajaraman methods were trained on 217 pairs of normal and bone-suppressed chest radiographs from the X-ray Bone Shadow Suppression dataset (https://www.kaggle.com/hmchuong/xray-bone-shadow-supression). Two classifier methods with different network architectures were implemented. Binary classifier models were trained on the public RICORD-1c and RSNA Pneumonia Challenge datasets. An external test dataset was created retrospectively from a set of 320 COVID-19 positive patients from Queen Elizabeth Hospital (Hong Kong, China) and a set of 518 non-COVID-19 patients from Pamela Youde Nethersole Eastern Hospital (Hong Kong, China), and used to evaluate the effect of bone suppression on classifier performance. Classification performance, quantified by sensitivity, specificity, negative predictive value (NPV), accuracy and area under the receiver operating curve (AUC), for non-suppressed radiographs was compared to that for bone suppressed radiographs. Some of the pre-trained models used in this study are published at (https://github.com/danielnflam). Results: Bone suppression of external test data was found to significantly (P<0.05) improve AUC for one classifier architecture [from 0.698 (non-suppressed) to 0.732 (Rajaraman-suppressed)]. For the other classifier architecture, suppression did not significantly (P>0.05) improve or worsen classifier performance. Conclusions: Rajaraman suppression significantly improved classification performance in one classification architecture, and did not significantly worsen classifier performance in the other classifier architecture. This research could be extended to explore the impact of bone suppression on classification of different lung pathologies, and the effect of other image enhancement techniques on classifier performance.

Performance comparison of the Cobas® Liat® and Cepheid® GeneXpert® systems on SARS-CoV-2 detection in nasopharyngeal swab and posterior oropharyngeal saliva.

Background: Nucleic acid amplification tests (NAATs) based methods such as real-time reverse transcription polymerase-chain reaction (real-time RT-PCR) are the gold standard for diagnosis of current infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The cobas® Liat® and cepheid® GeneXpert® systems are two rapid real-time RT-PCR platforms offering rapid, specimen-to-answer detection of SARS-CoV-2.Research design and methods: In this study, we compared the performance of these two systems on SARS-CoV-2 detection in 9 nasopharyngeal swab (NPS) and 70 posterior oropharyngeal saliva specimens collected from 79 patients suspected of SARS-CoV-2 infection between August 2020 and March 2021.Results: The Positive Percent Agreement (PPA), Negative Percent Agreement (NPA) and overall Percent Agreement (OPA) between cepheid® Xpress SARS-CoV-2 assay and cobas® Liat® SARS-CoV-2 & Influenza A/B assay were found to be 100%. We demonstrated an excellent overall test concordance of the Liat® SARS-CoV-2 & Influenza A/B assay and Xpress SARS-CoV-2 assay. The small sample size of SARS-CoV-2 positive and weak-positive specimens is the inherent limitation of this study.Conclusions: The performance of the cobas® Liat® SARS-CoV-2 & Influenza A/B assay is equivalent to the cepheid® Xpress SARS-CoV-2 assay for SARS-CoV-2 detection using NPS and posterior oropharyngeal saliva.

Spatial analysis of the impact of urban geometry and socio-demographic characteristics on COVID-19, a study in Hong Kong.

The World Health Organization considered the wide spread of COVID-19 over the world as a pandemic. There is still a lack of understanding of its origin, transmission, and treatment methods. Understanding the influencing factors of COVID-19 can help mitigate its spread, but little research on the spatial factors has been conducted. Therefore, this study explores the effects of urban geometry and socio-demographic factors on the COVID-19 cases in Hong Kong. For each patient, the places they visited during the incubation period before going to hospital were identified, and matched with corresponding attributes of urban geometry (i.e., building geometry, road network and greenspace) and socio-demographic factors (i.e., demographic, educational, economic, household and housing characteristics) based on the coordinates. The local cases were then compared with the imported cases using stepwise logistic regression, logistic regression with case-control of time, and least absolute shrinkage and selection operator regression to identify factors influencing local disease transmission. Results show that the building geometry, road network and certain socio-economic characteristics are significantly associated with COVID-19 cases. In addition, the results indicate that urban geometry is playing a more important role than socio-demographic characteristics in affecting COVID-19 incidence. These findings provide a useful reference to the government and the general public as to the spatial vulnerability of COVID-19 transmission and to take appropriate preventive measures in high-risk areas.

An update on COVID-19 pandemic: the epidemiology, pathogenesis, prevention and treatment strategies.

INTRODUCTION: To date, the transmission of Coronavirus Disease-2019 (COVID-19) is still uncontrollable with the fact that the numbers of confirmed and death cases are still increasing. Up to 1st October 2020, 33,842,281 confirmed cases and 1,010,634 confirmed deaths have been reported to the World Health Organization from 216 different countries, areas and territories. Despite the urgent demand for effective treatment strategies, there is still no specific antiviral treatment for COVID-19 and the treatment guidelines for COVID-19 vary between countries. AREA COVERED: In this article, we summarized the current knowledge on COVID-19 and the pandemic worldwide. Moreover, the epidemiology, pathogenesis, prevention and different treatment options will be discussed so that we shall prepare ourselves better to fight with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). EXPERT OPINION: The situation of the COVID-19 pandemic is still unpredictable. There is no effective vaccine or specific anti-viral drug to treat serve COVID-19 patients. Combination therapies have shown promising clinical improvement. Repurposing FDA-approved drugs might be one of possible treatment options. Without specific treatment and vaccines for COVID-19, the most effective way to prevent from being infected is to generate an ecosystem with effective protection, precautions and preventive measures.

The diagnostic methods in the COVID-19 pandemic, today and in the future.

INTRODUCTION: The emergence of anovel coronavirus identified in patients with unknown cause of acute respiratory disease in Wuhan, China at the end of 2019 has caused aglobal outbreak. The causative coronavirus was later named as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and the disease caused by SARS-CoV-2 was named as Coronavirus Disease-2019 (COVID-19). As of 10 August 2020, more than 19,718,030 confirmed cases and 728,013 deaths have been reported. COVID-19 is spread via respiratory droplets which are inhaled into the lungs. AREAS COVERED: In this article, we summarized the knowledge about the causative pathogen of COVID-19 and various diagnostic methods in this pandemic for better understanding of the limitations and the nuances of virus testing for COVID-19. EXPERT OPINION: In this pandemic, rapid and accurate identification of COVID-19 patients are critical to break the chain of infection in the community. RT-PCR provides a rapid and reliable identification of SARS-CoV-2 infection. In the future, molecular diagnostics will still be the gold standard and next-generation sequencing can help us to understand more on the pathogenesis and detect novel mutations. It is believed that more sophisticated detection methods will be introduced to detect SARS-CoV-2 as earliest as possible.