A comparison of COVID-19 and imaging radiation risk in clinical patient populations

Deep learning classification of COVID-19 in chest radiographs: performance and influence of supplemental training

Purpose: Accurate classification of COVID-19 in chest radiographs is invaluable to hard-hit pandemic hot spots. Transfer learning techniques for images using well-known convolutional neural networks show promise in addressing this problem. These methods can significantly benefit from supplemental training on similar conditions, considering that there currently exists no widely available chest x-ray dataset on COVID-19. We evaluate whether targeted pretraining for similar tasks in radiography labeling improves classification performance in a sample radiograph dataset containing COVID-19 cases. Approach: We train a DenseNet121 to classify chest radiographs through six training schemes. Each training scheme is designed to incorporate cases from established datasets for general findings in chest radiography (CXR) and pneumonia, with a control scheme with no pretraining. The resulting six permutations are then trained and evaluated on a dataset of 1060 radiographs collected from 475 patients after March 2020, containing 801 images of laboratory-confirmed COVID-19 cases. Results: Sequential training phases yielded substantial improvement in classification accuracy compared to a baseline of standard transfer learning with ImageNet parameters. The test set area under the receiver operating characteristic curve for COVID-19 classification improved from 0.757 in the control to 0.857 for the optimal training scheme in the available images. Conclusions: We achieve COVID-19 classification accuracies comparable to previous benchmarks of pneumonia classification. Deliberate sequential training, rather than pooling datasets, is critical in training effective COVID-19 classifiers within the limitations of early datasets. These findings bring clinical-grade classification through CXR within reach for more regions impacted by COVID-19.

Ultrasensitive Detection of SARS-CoV-2 Spike Proteins Using the Thio-NAD Cycling Reaction: A Preliminary Study before Clinical Trials

To help control the global pandemic of coronavirus disease 2019 (COVID-19), we developed a diagnostic method targeting the spike protein of the virus that causes the infection, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). We applied an ultrasensitive method by combining a sandwich enzyme-linked immunosorbent assay (ELISA) and the thio-nicotinamide adenine dinucleotide (thio-NAD) cycling reaction to quantify spike S1 proteins. The limit of detection (LOD) was 2.62 × 10⁻¹⁹ moles/assay for recombinant S1 proteins and 2.6 × 10⁶ RNA copies/assay for ultraviolet B-inactivated viruses. We have already shown that the ultrasensitive ELISA for nucleocapsid proteins can detect ultraviolet B-inactivated viruses at the 10⁴ RNA copies/assay level, whereas the nucleocapsid proteins of SARS-CoV-2 are difficult to distinguish from those in conventional coronaviruses and SARS-CoV. Thus, an antigen test for only the nucleocapsid proteins is insufficient for virus specificity. Therefore, the use of a combination of tests against both spike and nucleocapsid proteins is recommended to increase both the detection sensitivity and testing accuracy of the COVID-19 antigen test. Taken together, our present study, in which we incorporate S1 detection by combining the ultrasensitive ELISA for nucleocapsid proteins, offers an ultrasensitive, antigen-specific test for COVID-19.

Antigen tests for COVID-19

PCR diagnosis has been considered as the gold standard for coronavirus disease 2019 (COVID-19) and other many diseases. However, there are many problems in using PCR, such as non-specific (i.e., false-positive) and false-negative amplifications, the limits of a target sample volume, deactivation of the enzymes used, complicated techniques, difficulty in designing probe sequences, and the expense. We, thus, need an alternative to PCR, for example an ultrasensitive antigen test. In the present review, we summarize the following three topics. (1) The problems of PCR are outlined. (2) The antigen tests are surveyed in the literature that was published in 2020, and their pros and cons are discussed for commercially available antigen tests. (3) Our own antigen test on the basis of an ultrasensitive enzyme-linked immunosorbent assay (ELISA) is introduced. Finally, we discuss the possibility that our antigen test by an ultrasensitive ELISA technique will become the gold standard for diagnosis of COVID-19 and other diseases.