COVID-19 automatic diagnosis with CT images using the novel Transformer architecture

Even though vaccines are already in use worldwide, the COVID-19 pandemic is far from over, with some countries re-establishing the lockdown state, the virus has taken over 2 million lives until today, being a serious health issue. Although real-time reverse transcription-polymerase chain reaction (RTPCR) is the first tool for COVID-19 diagnosis, its high false-negative rate and low sensitivity might delay accurate diagnosis. Therefore, fast COVID-19 diagnosis and quarantine, combined with effective vaccination plans, is crucial for the pandemic to be over as soon as possible. To that end, we propose an intelligent system to classify computed tomography (CT) of lung images between a normal, pneumonia caused by something other than the coronavirus or pneumonia caused by the coronavirus. This paper aims to evaluate a complete selfattention mechanism with a Transformer network to capture COVID-19 pattern over CT images. This approach has reached the state-of-the-art in multiple NLP problems and just recently is being applied for computer vision tasks. We combine vision transformer and performer (linear attention transformers), and also a modified vision transformer, reaching 96.00% accuracy.

High false-negative rate of the anterior spinal artery by intercostobronchial trunk arteriography alone compared to CT during arteriography

Objectives: This study aimed to assess the frequencies of the anterior spinal arteries identified by CT during arteriography (CTA) and arteriography alone. Methods: This retrospective study included 137 vessels in 83 patients who performed both bronchial arteriography and CTA and categorized them into three groups based on the catheter tip locations: intercostobronchial trunk (ICBT), bronchial artery (BA), and intercostal artery (ICA). The frequencies of anterior spinal artery identified by CTA and arteriography alone were compared for each group. Results: ICBT, BA, and ICA groups were evaluated by CTA in 46, 79, and 12 vessels, respectively. By CTA evaluation, anterior spinal artery was identified in seven vessels (15.2%) in ICBT group, 0 in BA group, and 2 (16.7%) in ICA group. The frequencies of anterior spinal artery were significantly higher (p < 0.05) in ICBT and ICA groups than in BA group. By arteriography alone evaluation, a faint anterior spinal artery was identified in two vessels (4.3%) in ICBT group, 0 in BA group, and 1 (8.3%) in ICA group. Conclusions: Anterior spinal artery branched only from the ICBT or ICA and not from the BA in both arteriography and CTA assessments. There was high false-negative rate (71%) of the anterior spinal artery by ICBT arteriography alone assessment compared to CTA assessment. This result explains one of the reasons that spinal ischemia occurs in arteriography-negative spinal artery cases. Advances in knowledge: False-negative rate of anterior spinal artery was 71% (5/7) by intercostobronchial trunk arteriography alone assessment.

False Negative Mitigation in Group Testing for COVID-19 Screening

After lifting the COVID-19 lockdown restrictions and opening businesses, screening is essential to prevent the spread of the virus. Group testing could be a promising candidate for screening to save time and resources. However, due to the high false-negative rate (FNR) of the RT-PCR diagnostic test, we should be cautious about using group testing because a group's false-negative result identifies all the individuals in a group as uninfected. Repeating the test is the best solution to reduce the FNR, and repeats should be integrated with the group-testing method to increase the sensitivity of the test. The simplest way is to replicate the test twice for each group (the 2Rgt method). In this paper, we present a new method for group testing (the groupMix method), which integrates two repeats in the test. Then we introduce the 2-stage sequential version of both the groupMix and the 2Rgt methods. We compare these methods analytically regarding the sensitivity and the average number of tests. The tradeoff between the sensitivity and the average number of tests should be considered when choosing the best method for the screening strategy. We applied the groupMix method to screening 263 people and identified 2 infected individuals by performing 98 tests. This method achieved a 63% saving in the number of tests compared to individual testing. Our experimental results show that in COVID-19 screening, the viral load can be low, and the group size should not be more than 6; otherwise, the FNR increases significantly. A web interface of the groupMix method is publicly available for laboratories to implement this method.

Avoiding COVID-19 Hospital Outbreaks: RT-PCR Swab and Clinical Assessment

Nasopharyngeal RT-PCR swab test for COVID-19 diagnosis has a high specificity but also a low sensitivity. The high false-negative rate and the overconfidence in negative results sometimes lead to hospital outbreaks. Therefore, we recommend always integrating the clinical assessment in the diagnostic process, mostly after the test, to determine what degree of confidence can be attributed to a negative result.

A sequence-based method for predicting extant fold switchers that undergo α-helix β-strand transitions

Extant fold-switching proteins remodel their secondary structures and change their functions in response to cellular stimuli, regulating biological processes and affecting human health. In spite of their biological importance, these proteins remain understudied. Few representative examples of fold switchers are available in the Protein Data Bank, and they are difficult to predict. In fact, all 96 experimentally validated examples of extant fold switchers were stumbled upon by chance. Thus, predictive methods are needed to expedite the process of discovering and characterizing more of these shapeshifting proteins. Previous approaches require a solved structure or all-atom simulations, greatly constraining their use. Here, we propose a high-throughput sequence-based method for predicting extant fold switchers that transition from α-helix in one conformation to β-strand in the other. This method leverages two previous observations: (1) α-helix <-> β-strand prediction discrepancies from JPred4 are a robust predictor of fold switching, and (2) the fold-switching regions (FSRs) of some extant fold switchers have different secondary structure propensities when expressed in isolation (isolated FSRs) than when expressed within the context of their parent protein (contextualized FSRs). Combining these two observations, we ran JPred4 on the sequences of isolated and contextualized FSRs from 14 known extant fold switchers and found α-helix <->β-strand prediction discrepancies in every case. To test the overall robustness of this finding, we randomly selected regions of proteins not expected to switch folds (single-fold proteins) and found significantly fewer α-helix <-> β-strand prediction discrepancies (p < 4.2*10−20, Kolmogorov-Smirnov test). Combining these discrepancies with the overall percentage of predicted secondary structure, we developed a classifier that often robustly identifies extant fold switchers (Matthews Correlation Coefficient of 0.70). Although this classifier had a high false negative rate (6/14), its false positive rate was very low (1/211), suggesting that it can be used to predict a subset of extant fold switchers from billions of available genomic sequences.

Routine Chest X-Rays Are Inaccurate in Detecting Relevant Intrapulmonary Anomalies During Medical Assessments of Fitness to Dive

Introduction: Intrapulmonary pathology, such as bullae or blebs, can cause pulmonary barotrauma when diving. Many diving courses require chest X-rays (CXR) or high-resolution computed tomography (HRCT) to exclude asymptomatic healthy individuals with these lesions. The ability of routine CXRs and HRCT to assess fitness to dive has never been evaluated.Methods: Military divers who underwent yearly medical assessments at the Royal Netherlands Navy Diving Medical Center, including CXR at initial assessment, and who received a HRCT between January and June 2018, were included. The correlations of CXR and HRCT results with fitness to dive assessments were analyzed using Fisher's exact tests.Results: This study included 101 military divers. CXR identified bullae or blebs in seven divers, but HRCT found that these anomalies were not present in three subjects and were something else in four. CXR showed no anomalies in 94 subjects, but HRCT identified coincidental findings in 23 and bullae or blebs in seven. The differences between CXR and HRCT results were statistically significant (p = 0.023). Of the 34 subjects with anomalies on HRCT, 18 (53%) were disqualified for diving.Discussion: Routine CXR in asymptomatic military divers does not contribute to the identification of relevant pathology in fitness to dive assessments and has a high false negative rate (32%). HRCT is more diagnostic than CXR but yields unclear results, leading to disqualification for diving. Fitness to dive tests should exclude routine CXR; rather, HRCT should be performed only in subjects with clinical indications.

Transfer Detection of YOLO to Focus CNN’s Attention on Nude Regions for Adult Content Detection

Video pornography and nudity detection aim to detect and classify people in videos into nude or normal for censorship purposes. Recent literature has demonstrated pornography detection utilising the convolutional neural network (CNN) to extract features directly from the whole frames and support vector machine (SVM) to classify the extracted features into two categories. However, existing methods were not able to detect the small-scale content of pornography and nudity in frames with diverse backgrounds. This limitation has led to a high false-negative rate (FNR) and misclassification of nude frames as normal ones. In order to address this matter, this paper explores the limitation of the existing convolutional-only approaches focusing the visual attention of CNN on the expected nude regions inside the frames to reduce the FNR. The You Only Look Once (YOLO) object detector was transferred to the pornography and nudity detection application to detect persons as regions of interest (ROIs), which were applied to CNN and SVM for nude/normal classification. Several experiments were conducted to compare the performance of various CNNs and classifiers using our proposed dataset. It was found that ResNet101 with random forest outperformed other models concerning the F1-score of 90.03% and accuracy of 87.75%. Furthermore, an ablation study was performed to demonstrate the impact of adding the YOLO before the CNN. YOLO–CNN was shown to outperform CNN-only in terms of accuracy, which was increased from 85.5% to 89.5%. Additionally, a new benchmark dataset with challenging content, including various human sizes and backgrounds, was proposed.

Improving Detection Efficiency of SARS-CoV-2 Nucleic Acid Testing

BackgroundSARS-CoV-2 nucleic acid testing (NAT) has been routinely used for COVID-19 diagnosis during this pandemic; however, there have been concerns about its high false negative rate. We dissected its detection efficiency with a large COVID-19 cohort study.MethodsWe analyzed SARS-CoV-2 NAT positive rates of 4,275 specimens from 532 COVID-19 patients in Sichuan Province with different disease severities, statuses, and stages, as well as different types and numbers of specimens.ResultsThe total positive rate of the 4,275 specimens was 37.5%. Among seven specimen types, BALF generated a 77.8% positive rate, followed by URT specimens (38.5%), sputum (39.8%), and feces/rectal swabs (34.1%). Specimens from critical cases generated a 43.4% positive rate, which was significantly higher than that of other severities. With specimens from patients at stable status, the SARS-CoV-2 positive rate was 40.6%, which was significantly higher than that of improved status (17.1%), but lower than that of aggravated status (61.5%). Notably, the positive rate of specimens from COVID-19 patients varied significantly from 85 to 95% during 3 days before and after symptom onset, to 20% at around 18 days after symptom onset. In addition, the detection rate increased from 72.1% after testing one throat swab, to 93.2% after testing three consecutive respiratory specimens from each patient.ConclusionsSARS-CoV-2 NAT detection rates vary with patient disease severity and status, specimen type, number of specimens, and especially disease progression. Sampling as close to symptom onset as possible, and consecutively collecting more than one respiratory specimen could effectively improve SARS-CoV-2 NAT detection efficiency.

Prednis-OH NO! – A Case of Anaphylaxis Induced by Prednisone

Background: Anaphylaxis is a potentially life-threatening condition caused by the sudden release of inflammatory mediators into the systemic circulation. Among this condition’s etiologies, corticosteroid-induced anaphylaxis, despite being uncommon, should receive due consideration given the frequency of steroid use in various settings. Any patient that presents with shortness of breath, wheezing, hypotension, urticaria, or other characteristic signs of anaphylaxis following the administration of steroids should be promptly evaluated. Because of the potentially fatal nature of anaphylaxis, clinicians must be familiar with the presentation, diagnosis, and management of the reaction. Case Report: The primary objective of this case report is to discuss an example of such a reaction in a 21-year-old female with a past medical history of anxiety, depression, and alcoholism who presented with anaphylaxis following prednisone use, as well as the proposed pathophysiology and management thereafter. She was managed with intravenous epinephrine and diphenhydramine with complete resolution of her symptoms. She was subsequently discharged with an EpiPen, cetirizine, and advised to establish care with an allergist for follow up and additional allergy testing. To complete this case report, we performed a review of current primary literature on the subject. Conclusions: Though uncertain, many potential mechanisms of sensitization to corticosteroids were identified, including haptenization, preservatives, excipients, and conjugated esters. Various means exist to aid in diagnosis, such as skin testing, immunoCAP assays, lymphocyte transformation tests, basophil activation tests, and graded drug challenges, though these tests are associated with a high false negative rate. Accurate identification of the causative agent is crucial in facilitating avoidance or rapid desensitization prior to future corticosteroid use.