Deep Neural Network Augments Performance of Junior Residents in Diagnosing COVID-19 Pneumonia on Chest Radiographs.

Chest X-rays (CXRs) are essential in the preliminary radiographic assessment of patients affected by COVID-19. Junior residents, as the first point-of-contact in the diagnostic process, are expected to interpret these CXRs accurately. We aimed to assess the effectiveness of a deep neural network in distinguishing COVID-19 from other types of pneumonia, and to determine its potential contribution to improving the diagnostic precision of less experienced residents. A total of 5051 CXRs were utilized to develop and assess an artificial intelligence (AI) model capable of performing three-class classification, namely non-pneumonia, non-COVID-19 pneumonia, and COVID-19 pneumonia. Additionally, an external dataset comprising 500 distinct CXRs was examined by three junior residents with differing levels of training. The CXRs were evaluated both with and without AI assistance. The AI model demonstrated impressive performance, with an Area under the ROC Curve (AUC) of 0.9518 on the internal test set and 0.8594 on the external test set, which improves the AUC score of the current state-of-the-art algorithms by 1.25% and 4.26%, respectively. When assisted by the AI model, the performance of the junior residents improved in a manner that was inversely proportional to their level of training. Among the three junior residents, two showed significant improvement with the assistance of AI. This research highlights the novel development of an AI model for three-class CXR classification and its potential to augment junior residents' diagnostic accuracy, with validation on external data to demonstrate real-world applicability. In practical use, the AI model effectively supported junior residents in interpreting CXRs, boosting their confidence in diagnosis. While the AI model improved junior residents' performance, a decline in performance was observed on the external test compared to the internal test set. This suggests a domain shift between the patient dataset and the external dataset, highlighting the need for future research on test-time training domain adaptation to address this issue.

Specific lateral flow detection of isothermal nucleic acid amplicons for accurate point-of-care testing.

For point-of-care testing (POCT), coupling isothermal nucleic acid amplification schemes (e.g., recombinase polymerase amplification, RPA) with lateral flow assay (LFA) readout is an ideal platform, since such integration offers both high sensitivity and deployability. However, isothermal schemes typically suffers from non-specific amplification, which is difficult to be differentiated by LFA and thus results in false-positives. Here, we proposed an accurate POCT platform by specific recognition of target amplicons with peptide nucleic acid (PNA, assisted by T7 Exonuclease), which could be directly plugged into the existing RPA kits and commercial LFA test strips. With SARS-CoV-2 as the model, the proposed method (RPA-TeaPNA-LFA) efficiently eliminated the false-positives, exhibiting a lowest detection concentration of 6.7 copies/µL of RNA and 90 copies/µL of virus. Using dual-gene (orf1ab and N genes of SARS-CoV-2) as the targets, RPA-TeaPNA-LFA offered a high specificity (100%) and sensitivity (RT-PCR Ct < 31, 100%; Ct < 40, 71.4%), and is valuable for on-site screening or self-testing during isolation. In addition, the dual test lines in the test strips were successfully explored for simultaneous detection of SARS-CoV-2 and H1N1, showing great potential in response to future pathogen-based pandemics.

Contrastive domain adaptation with consistency match for automated pneumonia diagnosis.

Pneumonia can be difficult to diagnose since its symptoms are too variable, and the radiographic signs are often very similar to those seen in other illnesses such as a cold or influenza. Deep neural networks have shown promising performance in automated pneumonia diagnosis using chest X-ray radiography, allowing mass screening and early intervention to reduce the severe cases and death toll. However, they usually require many well-labelled chest X-ray images for training to achieve high diagnostic accuracy. To reduce the need for training data and annotation resources, we propose a novel method called Contrastive Domain Adaptation with Consistency Match (CDACM). It transfers the knowledge from different but relevant datasets to the unlabelled small-size target dataset and improves the semantic quality of the learnt representations. Specifically, we design a conditional domain adversarial network to exploit discriminative information conveyed in the predictions to mitigate the domain gap between the source and target datasets. Furthermore, due to the small scale of the target dataset, we construct a feature cloud for each target sample and leverage contrastive learning to extract more discriminative features. Lastly, we propose adaptive feature cloud expansion to push the decision boundary to a low-density area. Unlike most existing transfer learning methods that aim only to mitigate the domain gap, our method instead simultaneously considers the domain gap and the data deficiency problem of the target dataset. The conditional domain adaptation and the feature cloud generation of our method are learning jointly to extract discriminative features in an end-to-end manner. Besides, the adaptive feature cloud expansion improves the model's generalisation ability in the target domain. Extensive experiments on pneumonia and COVID-19 diagnosis tasks demonstrate that our method outperforms several state-of-the-art unsupervised domain adaptation approaches, which verifies the effectiveness of CDACM for automated pneumonia diagnosis using chest X-ray imaging.

Correlation Analysis of Positive Therapy Based on High Content Image Analysis Technology on Posttraumatic Nerve Growth in Patients with COVID-19 in the Context of Intelligent Medical Treatment.

Objective: To investigate the correlation between posttraumatic stress disorder (PTSD) and the incidence of anxiety, depression, and mental disorders in patients with novel coronavirus pneumonia. Methods: Novel coronavirus pneumonia patients in Wuhan from 2020 to April were selected for treatment from hospitals and isolation wards from 1 to April. 70 rehabilitated patients were randomly divided into the control group (35 patients) and the observation group (35 patients) who were treated with conventional therapy. Positive therapy and full perfusion therapy were introduced on the basis of conventional therapy, and the related performances of different patients were observed and evaluated. Results: The anxiety, depression, and incidence rate of related psychotic patients in the observation group after treatment were significantly reduced. Patients could maintain a good mood, increase their confidence in conquering diseases, and promote their early recovery. Conclusion: Active treatment of novel coronavirus pneumonia has positive effects on posttraumatic growth of new crown pneumonia patients, relieving anxiety and negative emotions, improving emotional control, eliminating bad emotions, actively guiding patients, and promoting psychological rehabilitation of patients.

Deep Supervised Domain Adaptation for Pneumonia Diagnosis From Chest X-Ray Images.

Pneumonia is one of the most common treatable causes of death, and early diagnosis allows for early intervention. Automated diagnosis of pneumonia can therefore improve outcomes. However, it is challenging to develop high-performance deep learning models due to the lack of well-annotated data for training. This paper proposes a novel method, called Deep Supervised Domain Adaptation (DSDA), to automatically diagnose pneumonia from chest X-ray images. Specifically, we propose to transfer the knowledge from a publicly available large-scale source dataset (ChestX-ray14) to a well-annotated but small-scale target dataset (the TTSH dataset). DSDA aligns the distributions of the source domain and the target domain according to the underlying semantics of the training samples. It includes two task-specific sub-networks for the source domain and the target domain, respectively. These two sub-networks share the feature extraction layers and are trained in an end-to-end manner. Unlike most existing domain adaptation approaches that perform the same tasks in the source domain and the target domain, we attempt to transfer the knowledge from a multi-label classification task in the source domain to a binary classification task in the target domain. To evaluate the effectiveness of our method, we compare it with several existing peer methods. The experimental results show that our method can achieve promising performance for automated pneumonia diagnosis.

Association between clinical characteristics and CT findings in patients with coronavirus disease-2019.

ABSTRACT: This retrospective study was to investigate the association between clinical characteristics and computerized tomography (CT) findings in patients with coronavirus disease-2019 (COVID-19). The clinical data of COVID-19 patients were retrospectively analyzed. Spearman correlation analysis was used to identify the correlation. Totally 209 consecutive COVID-19 patients were eligible for the study, with the mean age of 47.53 ±â€Š13.52 years. At onset of the disease, the most common symptoms were fever (85.65%) and cough (61.24%). The CT features of COVID-19 included pulmonary, bronchial, and pleural changes, with the significant pulmonary presentation of ground-glass opacification (93.30%), consolidation (48.80%), ground-glass opacification plus a reticular pattern (54.07%), telangiectasia (84.21%), and pulmonary fibrotic streaks (49.76%). Spearman analysis showed that the CT findings had significantly inverse associations with the platelets, lymphocyte counts, and sodium levels, but were positively related to the age, erythrocyte sedimentation rate, D-dimer, lactic dehydrogenase, α-hydroxybutyrate dehydrogenase, and C-reactive protein levels (P < .05). In conclusion, the severity of lung abnormalities on CT in COVID-19 patients is inversely associated with the platelets, lymphocyte count, and sodium levels, whereas positively with the age, erythrocyte sedimentation rate, D-dimer, lactic dehydrogenase, hydroxybutyrate dehydrogenase, and C-reactive protein levels.

Laboratory Predictors of COVID-19 Mortality: A Retrospective Analysis from Tongji Hospital in Wuhan.

BACKGROUND: Novel coronavirus disease 2019 (COVID-19), an acute respiratory disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), rapidly progressed to a global pandemic. Currently, there are limited effective medications approved for this deadly disease. OBJECTIVE: To investigate the potential predictors of COVID-19 mortality and risk factors for hyperinflammation in COVID-19. METHODS: Retrospective analysis was carried out in 1,149 patients diagnosed with COVID-19 in Tongji Hospital, Wuhan, China, from 1/13/2020 to 3/15/2020. RESULTS: We found significant differences in the rates of hyperuricemia (OR: 3.17, 95% CI: 2.13-4.70; p < 0.001) and hypoalbuminemia (OR: 5.68, 95% CI: 3.97-8.32; p < 0.001) between deceased and recovered patients. The percentages of hyperuricemia in deceased patients and recovered patients were 23.6% and 8.9%, respectively, which were higher than the reported age-standardized prevalence of 6.2% in Chinese population. Of note, the percentages of both IL-6 and uric acid levels in survived COVID-19 patients were above 90%, suggesting that they might be good specificity for indicators of mortality in COVID-19 patients. The serum level of uric acid (UA) was positively associated with ferritin, TNF-α, and IL-6 but not with anti-inflammatory cytokine IL-10. In addition, the levels of these proinflammatory cytokines in COVID-19 patients showed a trend of reduction after uric acid lowering therapy. CONCLUSIONS: Our results suggest that uric acid, the end product of purine metabolism, was increased in deceased patients with COVID-19. In addition, the serum level of uric acid was positively associated with inflammatory markers. Uric acid lowering therapy in COVID-19 patients with hyperuricemia may be beneficial.

Peripheral CD4⁺ T cell subsets and antibody response in COVID-19 convalescent individuals

BACKGROUND. Marked progress is achieved in understanding the physiopathology of coronavirus disease 2019 (COVID-19), which caused a global pandemic. However, the CD4· T cell population critical for antibody response in COVID-19 is poorly understood. METHODS. In this study, we provided a comprehensive analysis of peripheral CD4· T cells from 13 COVID-19 convalescent patients, defined as confirmed free of SARS-CoV-2 for 2 to 4 weeks, using flow cytometry and magnetic chemiluminescence enzyme antibody immunoassay. The data were correlated with clinical characteristics. RESULTS. We observed that, relative to healthy individuals, convalescent patients displayed an altered peripheral CD4· T cell spectrum. Specifically, consistent with other viral infections, cTfh1 cells associated with SARS-CoV-2-targeting antibodies were found in COVID-19 covalescent patients. Individuals with severe disease showed higher frequencies of Tem and Tfh-em cells but lower frequencies of Tcm, Tfh-cm, Tfr, and Tnaive cells, compared with healthy individuals and patients with mild and moderate disease. Interestingly, a higher frequency of cTfh-em cells correlated with a lower blood oxygen level, recorded at the time of admission, in convalescent patients. These observations might constitute residual effects by which COVID-19 can impact the homeostasis of CD4· T cells in the long-term and explain the highest ratio of class-switched virus-specific antibody producing individuals found in our severe COVID-19 cohort. CONCLUSION. Our study demonstrated a close connection between CD4· T cells and antibody production in COVID-19 convalescent patients. FUNDING. Six Talent Peaks Project in Jiangsu Province and the National Natural Science Foundation of China (NSFC).

Peripheral CD4+ T cell subsets and antibody response in COVID-19 convalescent individuals.

BACKGROUNDMarked progress is achieved in understanding the physiopathology of coronavirus disease 2019 (COVID-19), which caused a global pandemic. However, the CD4+ T cell population critical for antibody response in COVID-19 is poorly understood.METHODSIn this study, we provided a comprehensive analysis of peripheral CD4+ T cells from 13 COVID-19 convalescent patients, defined as confirmed free of SARS-CoV-2 for 2 to 4 weeks, using flow cytometry and magnetic chemiluminescence enzyme antibody immunoassay. The data were correlated with clinical characteristics.RESULTSWe observed that, relative to healthy individuals, convalescent patients displayed an altered peripheral CD4+ T cell spectrum. Specifically, consistent with other viral infections, cTfh1 cells associated with SARS-CoV-2-targeting antibodies were found in COVID-19 covalescent patients. Individuals with severe disease showed higher frequencies of Tem and Tfh-em cells but lower frequencies of Tcm, Tfh-cm, Tfr, and Tnaive cells, compared with healthy individuals and patients with mild and moderate disease. Interestingly, a higher frequency of cTfh-em cells correlated with a lower blood oxygen level, recorded at the time of admission, in convalescent patients. These observations might constitute residual effects by which COVID-19 can impact the homeostasis of CD4+ T cells in the long-term and explain the highest ratio of class-switched virus-specific antibody producing individuals found in our severe COVID-19 cohort.CONCLUSIONOur study demonstrated a close connection between CD4+ T cells and antibody production in COVID-19 convalescent patients.FUNDINGSix Talent Peaks Project in Jiangsu Province and the National Natural Science Foundation of China (NSFC).

Clinical characteristics and outcomes of COVID-19 patients with gastrointestinal symptoms admitted to Jianghan Fangcang Shelter Hospital in Wuhan, China.

Coronavirus disease 2019 (COVID-19) is a health emergency worldwide, and gastrointestinal (GI) symptoms are increasingly reported in COVID-19 patients. However, sample size was small and the incidence of GI symptoms in patients was variable across studies, and the correlation between these symptoms and clinical outcomes remains incompletely understood. The objective of this study is to compare clinical characteristics and outcomes between patients with and without GI symptoms admitted to Jianghan Fangcang Shelter Hospital in Wuhan. This retrospective study recruited 1320 COVID-19 patients admitted to hospital from 5 February 2020 to 9 March 2020. On the basis of the presence of GI symptoms, the sample was divided into a GI group (n = 192) and a non-GI group (n = 1128). The three most common GI symptoms were diarrhea (8.1%), anorexia (4.7%), and nausea and vomiting (4.3%). The rate of clinical deterioration was significantly higher in the GI group than in the non-GI group (15.6% vs. 10.1%, P = .032). GI symptoms (P = .045), male gender P < .001), and increased C-reactive protein (P = .008) were independent risk factors for clinical worsening. This study demonstrated that the rate of clinical deterioration was significantly higher in the GI group. Furthermore, potential risk factors for developing GI symptoms, male gender, and increased C-reactive protein can help clinicians predict clinical outcomes in COVID-19 patients.