ABSTRACT
Nucleic acid testing is currently the golden reference for coronaviruses (SARS-CoV-2) detection, while the SARS-CoV-2 antigen-detection rapid diagnostic tests (RDT) is an important adjunct. RDT can be widely used in the community or regional screening management as self-test tools and may need to be verified by healthcare authorities. However, manual verification of RDT results is a time-consuming task, and existing object detection algorithms usually suffer from high model complexity and computational effort, making them difficult to deploy. We propose LightR-YOLOv5, a compact rotating SARS-CoV-2 antigen-detection RDT results detector. Firstly, we employ an extremely light-weight L-ShuffleNetV2 network as a feature extraction network with a slight reduction in recognition accuracy. Secondly, we combine semantic and texture features in different layers by judiciously combining and employing GSConv, depth-wise convolution, and other modules, and further employ the NAM attention to locate the RDT result detection region. Furthermore, we propose a new data augmentation approach, Single-Copy-Paste, for increasing data samples for the specific task of RDT result detection while achieving a small improvement in model accuracy. Compared with some mainstream rotating object detection networks, the model size of our LightR-YOLOv5 is only 2.03MB, and it is 12.6%, 6.4%, and 7.3% higher in mAP@.5:.95 metrics compared to RetianNet, FCOS, and R3Det, respectively.
ABSTRACT
Nucleic acid testing is currently the golden reference for coronaviruses (SARS-CoV-2) detection, while the SARS-CoV-2 antigen-detection rapid diagnostic tests (RDT) is an important adjunct. RDT can be widely used in the community or regional screening management as self-test tools and may need to be verified by healthcare authorities. However, manual verification of RDT results is a time-consuming task, and existing object detection algorithms usually suffer from high model complexity and computational effort, making them difficult to deploy. We propose LightR-YOLOv5, a compact rotating SARS-CoV-2 antigen-detection RDT results detector. Firstly, we employ an extremely light-weight L-ShuffleNetV2 network as a feature extraction network with a slight reduction in recognition accuracy. Secondly, we combine semantic and texture features in different layers by judiciously combining and employing GSConv, depth-wise convolution, and other modules, and further employ the NAM attention to locate the RDT result detection region. Furthermore, we propose a new data augmentation approach, Single-Copy-Paste, for increasing data samples for the specific task of RDT result detection while achieving a small improvement in model accuracy. Compared with some mainstream rotating object detection networks, the model size of our LightR-YOLOv5 is only 2.03MB, and it is 12.6%, 6.4%, and 7.3% higher in mAP@.5:.95 metrics compared to RetianNet, FCOS, and R3Det, respectively.
Subject(s)
COVID-19 , Olfaction Disorders , Humans , COVID-19/complications , Olfaction Disorders/etiologyABSTRACT
The front-line imaging modalities computed tomography (CT) and X-ray play important roles for triaging COVID patients. Thoracic CT has been accepted to have higher sensitivity than a chest X-ray for COVID diagnosis. Considering the limited access to resources (both hardware and trained personnel) and issues related to decontamination, CT may not be ideal for triaging suspected subjects. Artificial intelligence (AI) assisted X-ray based application for triaging and monitoring require experienced radiologists to identify COVID patients in a timely manner with the additional ability to delineate and quantify the disease region is seen as a promising solution for widespread clinical use. Our proposed solution differs from existing solutions presented by industry and academic communities. We demonstrate a functional AI model to triage by classifying and segmenting a single chest X-ray image, while the AI model is trained using both X-ray and CT data. We report on how such a multi-modal training process improves the solution compared to single modality (X-ray only) training. The multi-modal solution increases the AUC (area under the receiver operating characteristic curve) from 0.89 to 0.93 for a binary classification between COVID-19 and non-COVID-19 cases. It also positively impacts the Dice coefficient (0.59 to 0.62) for localizing the COVID-19 pathology. To compare the performance of experienced readers to the AI model, a reader study is also conducted. The AI model showed good consistency with respect to radiologists. The DICE score between two radiologists on the COVID group was 0.53 while the AI had a DICE value of 0.52 and 0.55 when compared to the segmentation done by the two radiologists separately. From a classification perspective, the AUCs of two readers was 0.87 and 0.81 while the AUC of the AI is 0.93 based on the reader study dataset. We also conducted a generalization study by comparing our method to the-state-art methods on independent datasets. The results show better performance from the proposed method. Leveraging multi-modal information for the development benefits the single-modal inferencing.
ABSTRACT
Background: COVID-19 has rapidly become a major health emergency worldwide. The characteristic, outcome, and risk factor of COVID-19 in patients with decompensated cirrhosis remain unclear.Methods: Medical records were collected from 23 Chinese hospitals. Patients with decompensated cirrhosis and age- and sex-matched non-liver disease patients were enrolled with 1:4 ratio using stratified sampling.Results: There were more comorbidities with higher Chalson Complication Index (p < 0.001), higher proportion of patients having gastrointestinal bleeding, jaundice, ascites, and diarrhea among those patients (p < 0.05) and in decompensated cirrhosis patients. Mortality (p < 0.05) and the proportion of severe ill (p < 0.001) were significantly high among those patients. Patients in severe ill subgroup had higher mortality (p < 0.001), MELD, and CRUB65 score but lower lymphocytes count. Besides, this subgroup had larger proportion of patients with abnormal (PT), activated partial thromboplatin time (APTT), D-Dimer, alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin (TBL) and Creatinine (Cr) (p < 0.05). Multivariate logistic regression for severity shown that MELD and CRUB65 score reached significance. Higher Child-Pugh and CRUB65 scores were found among non-survival cases and multivariate logistic regression further inferred risk factors for adverse outcome. Receiver Operating Characteristic (ROC) curves also provided remarkable demonstrations for the predictive ability of Child-Pugh and CRUB65 scores.Conclusions: COVID-19 patients with cirrhosis had larger proportion of more severely disease and higher mortality. MELD and CRUB65 score at hospital admission may predict COVID-19 severity while Child-Pugh and CRUB65 score were highly associated with non-survival among those patients.
Subject(s)
COVID-19/mortality , Liver Cirrhosis/complications , SARS-CoV-2 , Severity of Illness Index , Adult , Aged , COVID-19/etiology , Female , Humans , Male , Middle Aged , Retrospective Studies , Risk FactorsABSTRACT
BACKGROUND: A novel coronavirus (COVID-19) caused pneumonia broke out at the end of 2019 in Wuhan, China. Many cases were subsequently reported in other cities, which has aroused strong reverberations on the Internet and social media around the world. OBJECTIVE: The aim of this study was to investigate the reaction of global Internet users to the outbreak of COVID-19 by evaluating the possibility of using Internet monitoring as an instrument in handling communicable diseases and responding to public health emergencies. METHODS: The disease-related data were retrieved from China's National Health Commission (CNHC) and World Health Organization (WHO) from January 10 to February 29, 2020. Daily Google Trends (GT) and daily Baidu Attention Index (BAI) for the keyword "Coronavirus" were collected from their official websites. Rumors which occurred in the course of this outbreak were mined from Chinese National Platform to Refute Rumors (CNPRR) and Tencent Platform to Refute Rumors (TPRR). Kendall's Tau-B rank test was applied to check the bivariate correlation among the two indexes mentioned above, epidemic trends, and rumors. RESULTS: After the outbreak of COVID-19, both daily BAI and daily GT increased rapidly and remained at a high level, this process lasted about 10 days. When major events occurred, daily BAI, daily GT, and the number of rumors simultaneously reached new peaks. Our study indicates that these indexes and rumors are statistically related to disease-related indicators. Information symmetry was also found to help significantly eliminate the false news and to prevent rumors from spreading across social media through the epidemic outbreak. CONCLUSION: Compared to traditional methods, Internet monitoring could be particularly efficient and economical in the prevention and control of epidemic and rumors by reflecting public attention and attitude, especially in the early period of an outbreak.