COVID-19 or non-COVID viral pneumonia: How to differentiate based on the radiologic findings?

A Literature Review on the Relative Diagnostic Accuracy of Chest CT Scans versus RT-PCR Testing for COVID-19 Diagnosis

BACKGROUND

Reverse transcription polymerase chain reaction (RT-PCR) is the main technique used to identify COVID-19 from respiratory samples. It has been suggested in several articles that chest CTs could offer a possible alternate diagnostic tool for COVID-19; however, no professional medical body recommends using chest CTs as an early COVID-19 detection modality. This literature review examines the use of CT scans as a diagnostic tool for COVID-19.

METHOD

A comprehensive search of research works published in peer-reviewed journals was carried out utilizing precisely stated criteria. The search was limited to English-language publications, and studies of COVID-19-positive patients diagnosed using both chest CT scans and RT-PCR tests were sought. For this review, four databases were consulted: these were the Cochrane and ScienceDirect catalogs, and the CINAHL and Medline databases made available by EBSCOhost.

FINDINGS

In total, 285 possibly pertinent studies were found during an initial search. After applying inclusion and exclusion criteria, six studies remained for analysis. According to the included studies, chest CT scans were shown to have a 44 to 98% sensitivity and 25 to 96% specificity in terms of COVID-19 diagnosis. However, methodological limitations were identified in all studies included in this review.

CONCLUSION

RT-PCR is still the suggested first-line diagnostic technique for COVID-19; while chest CT is adequate for use in symptomatic patients, it is not a sufficiently robust diagnostic tool for the primary screening of COVID-19.

COVID-19 Detection and Diagnosis Model on CT Scans Based on AI Techniques

The end of 2019 could be mounted in a rudimentary framing of a new medical problem, which globally introduces into the discussion a fulminant outbreak of coronavirus, consequently spreading COVID-19 that conducted long-lived and persistent repercussions. Hence, the theme proposed to be solved arises from the field of medical imaging, where a pulmonary CT-based standardized reporting system could be addressed as a solution. The core of it focuses on certain impediments such as the overworking of doctors, aiming essentially to solve a classification problem using deep learning techniques, namely, if a patient suffers from COVID-19, viral pneumonia, or is healthy from a pulmonary point of view. The methodology's approach was a meticulous one, denoting an empirical character in which the initial stage, given using data processing, performs an extraction of the lung cavity from the CT scans, which is a less explored approach, followed by data augmentation. The next step is comprehended by developing a CNN in two scenarios, one in which there is a binary classification (COVID and non-COVID patients), and the other one is represented by a three-class classification. Moreover, viral pneumonia is addressed. To obtain an efficient version, architectural changes were gradually made, involving four databases during this process. Furthermore, given the availability of pre-trained models, the transfer learning technique was employed by incorporating the linear classifier from our own convolutional network into an existing model, with the result being much more promising. The experimentation encompassed several models including MobileNetV1, ResNet50, DenseNet201, VGG16, and VGG19. Through a more in-depth analysis, using the CAM technique, MobilneNetV1 differentiated itself via the detection accuracy of possible pulmonary anomalies. Interestingly, this model stood out as not being among the most used in the literature. As a result, the following values of evaluation metrics were reached: loss (0.0751), accuracy (0.9744), precision (0.9758), recall (0.9742), AUC (0.9902), and F1 score (0.9750), from 1161 samples allocated for each of the three individual classes.

Digital Recombinase Polymerase Amplification, Digital Loop-Mediated Isothermal Amplification, and Digital CRISPR-Cas Assisted Assay: Current Status, Challenges, and Perspectives

Digital nucleic acid detection based on microfluidics technology can quantify the initial amount of nucleic acid in the sample with low equipment requirements and simple operations, which can be widely used in clinical and in vitro diagnosis. Recently, isothermal amplification technologies such as recombinase polymerase amplification (RPA), loop-mediated isothermal amplification (LAMP), and clustered regularly interspaced short palindromic repeats-CRISPR associated proteins (CRISPR-Cas) assisted technologies have become a hot spot of attention and state-of-the-art digital nucleic acid chips have provided a powerful tool for these technologies. Herein, isothermal amplification technologies including RPA, LAMP, and CRISPR-Cas assisted methods, based on digital nucleic acid microfluidics chips recently, have been reviewed. Moreover, the challenges of digital isothermal amplification and possible strategies to address them are discussed. Finally, future directions of digital isothermal amplification technology, such as microfluidic chip and device manufacturing, multiplex detection, and one-pot detection, are outlined.

Association of subpleural ground-glass opacities with respiratory failure and RNAemia in COVID-19

OBJECTIVES

To examine the radiological patterns specifically associated with hypoxemic respiratory failure in patients with coronavirus disease (COVID-19).

METHODS

We enrolled patients with COVID-19 confirmed by qPCR in this prospective observational cohort study. We explored the association of clinical, radiological, and microbiological data with the development of hypoxemic respiratory failure after COVID-19 onset. Semi-quantitative CT scores and dominant CT patterns were retrospectively determined for each patient. The microbiological evaluation included checking the SARS-CoV-2 viral load by qPCR using nasal swab and serum specimens.

RESULTS

Of the 214 eligible patients, 75 developed hypoxemic respiratory failure and 139 did not. The CT score was significantly higher in patients who developed hypoxemic respiratory failure than in those did not (median [interquartile range]: 9 [6-14] vs 0 [0-3]; p < 0.001). The dominant CT patterns were subpleural ground-glass opacities (GGOs) extending beyond the segmental area (n = 44); defined as "extended GGOs." Multivariable analysis showed that hypoxemic respiratory failure was significantly associated with extended GGOs (odds ratio [OR] 29.6; 95% confidence interval [CI], 9.3-120; p < 0.001), and a CT score  > 4 (OR 12.7; 95% CI, 5.3-33; p < 0.001). The incidence of RNAemia was significantly higher in patients with extended GGOs (58.3%) than in those without any pulmonary lesion (14.7%; p < 0.001).

CONCLUSIONS

Extended GGOs along the subpleural area were strongly associated with hypoxemia and viremia in patients with COVID-19.

KEY POINTS

• Extended ground-glass opacities (GGOs) along the subpleural area and a CT score  > 4, in the early phase of COVID-19, were independently associated with the development of hypoxemic respiratory failure. • The absence of pulmonary lesions on CT in the early phase of COVID-19 was associated with a lower risk of developing hypoxemic respiratory failure. • Compared to patients with other CT findings, the extended GGOs and a higher CT score were also associated with a higher incidence of RNAemia.

A multi-output network with U-net enhanced class activation map and robust classification performance for medical imaging analysis

Computer vision in medical diagnosis has achieved a high level of success in diagnosing diseases with high accuracy. However, conventional classifiers that produce an image-to-label result provide insufficient information for medical professionals to judge and raise concerns over the trust and reliability of a model with results that cannot be explained. To gain local insight of cancerous regions, separate tasks such as imaging segmentation needs to be implemented to aid the doctors in treating patients which doubles the training time and costs which renders the diagnosis system inefficient and difficult to be accepted by the public. To tackle this issue and drive the AI-first medical solutions further, this paper proposes a multi-output network which follows a U-Net architecture for image segmentation output and features an additional CNN module for auxiliary classification output. Class Activation Maps or CAMs are a method of providing insight into a convolutional neural network’s feature maps that lead to its classification but in the case of lung diseases, the region of interest is enhanced by U-net assisted Class Activation Mapping (CAM) visualization. Therefore, our proposed model combines image segmentation models and classifiers to crop out only the lung region of a chest X-ray’s class activation map to provide a visualization that improves the explainability and can generate classification results simultaneously which builds trust for AI-led diagnosis system. The proposed U-Net model achieves 97.72% accuracy and a dice coefficient of 0.9691 on a testing data from the COVID-QU-Ex Dataset which includes both diseased and healthy lungs.

Black Fungus and beyond: COVID-19 associated infections

Globally, many hospitalized COVID-19 patients can experience an unexpected acute change in status, prompting rapid and expert clinical assessment. Superimposed infections can be a significant cause of clinical and radiologic deviations in this patient population, further worsening clinical outcome and muddling the differential diagnosis. As thrombotic, inflammatory, and medication-induced complications can also trigger an acute change in COVID-19 patient status, imaging early and often plays a vital role in distinguishing the cause of patient decline and monitoring patient outcome. While the common radiologic findings of COVID-19 infection are now widely reported, little is known about the clinical manifestations and imaging findings of superimposed infection. By discussing case studies of patients who developed bacterial, fungal, parasitic, and viral co-infections and identifying the most frequently reported imaging findings of superimposed infections, physicians will be more familiar with common infectious presentations and initiate a directed workup sooner. Ultimately, any abrupt changes in the expected COVID-19 imaging presentation, such as the presence of new consolidations or cavitation, should prompt further workup to exclude superimposed opportunistic infection.

FDG-PET/CT of COVID-19 and Other Lung Infections

While not conventionally used as the first-line modality, [¹⁸F]-2-fluoro-2-deoxy-D-glucose (FDG) - positron emission tomography/computed tomography (PET/CT) can identify infection and inflammation both earlier and with higher sensitivity than anatomic imaging modalities [including chest X-ray (CXR), computed tomography (CT), and magnetic resonance imaging (MRI)]. The extent of inflammation and, conversely, recovery within the lungs, can be roughly quantified on FDG-PET/CT using maximum standardized uptake value (SUVmax) values. The Coronavirus disease 2019 (COVID-19) pandemic has highlighted the value of FDG-PET/CT in diagnosis, elucidation of acute pulmonary and extrapulmonary manifestations, and long-term follow up. Similarly, many other pulmonary infections such as previously documented coronaviruses, aspergillosis, blastomycosis, candidiasis, coccidioidomycosis, cryptococcosis, histoplasmosis, mucormycosis, and typical/atypical mycobacterial infections have all been identified and characterized using FDG-PET/CT imaging. The goal of this review is to summarize the actual and potential benefits of FDG-PET/CT in the imaging of COVID-19 and other lung infections. Further research is necessary to determine the best indications and clinical applications of FDG-PET/CT, improve its specificity, and ultimately ascertain how this modality can best be utilized in the diagnostic work up of infectious pathologies.