The role of chest imaging in the diagnosis, management, and monitoring of coronavirus disease 2019 (COVID-19)

Predictive models of severe disease in patients with COVID-19 pneumonia at an early stage on CT images using topological properties

Prediction of severe disease (SVD) in patients with coronavirus disease (COVID-19) pneumonia at an early stage could allow for more appropriate triage and improve patient prognosis. Moreover, the visualization of the topological properties of COVID-19 pneumonia could help clinical physicians describe the reasons for their decisions. We aimed to construct predictive models of SVD in patients with COVID-19 pneumonia at an early stage on computed tomography (CT) images using SVD-specific features that can be visualized on accumulated Betti number (BN) maps. BN maps (b0 and b1 maps) were generated by calculating the BNs within a shifting kernel in a manner similar to a convolution. Accumulated BN maps were constructed by summing BN maps (b0 and b1 maps) derived from a range of multiple-threshold values. Topological features were computed as intrinsic topological properties of COVID-19 pneumonia from the accumulated BN maps. Predictive models of SVD were constructed with two feature selection methods and three machine learning models using nested fivefold cross-validation. The proposed model achieved an area under the receiver-operating characteristic curve of 0.854 and a sensitivity of 0.908 in a test fold. These results suggested that topological image features could characterize COVID-19 pneumonia at an early stage as SVD.

Performance of AI Approaches for COVID-19 Diagnosis Using Chest CT Scans: The Impact of Architecture and Dataset

AI is emerging as a promising tool for diagnosing COVID-19 based on chest CT scans. The aim of this study was the comparison of AI models for COVID-19 diagnosis. Therefore, we: (1) trained three distinct AI models for classifying COVID-19 and non-COVID-19 pneumonia (nCP) using a large, clinically relevant CT dataset, (2) evaluated the models' performance using an independent test set, and (3) compared the models both algorithmically and experimentally.In this multicenter multi-vendor study, we collected n=1591 chest CT scans of COVID-19 (n=762) and nCP (n=829) patients from China and Germany. In Germany, the data was collected from three RACOON sites. We trained and validated three COVID-19 AI models with different architectures: COVNet based on 2D-CNN, DeCoVnet based on 3D-CNN, and AD3D-MIL based on 3D-CNN with attention module. 991 CT scans were used for training the AI models using 5-fold cross-validation. 600 CT scans from 6 different centers were used for independent testing. The models' performance was evaluated using accuracy (Acc), sensitivity (Se), and specificity (Sp).The average validation accuracy of the COVNet, DeCoVnet, and AD3D-MIL models over the 5 folds was 80.9%, 82.0%, and 84.3%, respectively. On the independent test set with n=600 CT scans, COVNet yielded Acc=76.6%, Se=67.8%, Sp=85.7%; DeCoVnet provided Acc=75.1%, Se=61.2%, Sp=89.7%; and AD3D-MIL achieved Acc=73.9%, Se=57.7%, Sp=90.8%.The classification performance of the evaluated AI models is highly dependent on the training data rather than the architecture itself. Our results demonstrate a high specificity and moderate sensitivity. The AI classification models should not be used unsupervised but could potentially assist radiologists in COVID-19 and nCP identification. · This study compares AI approaches for diagnosing COVID-19 in chest CT scans, which is essential for further optimizing the delivery of healthcare and for pandemic preparedness.. · Our experiments using a multicenter, multi-vendor, diverse dataset show that the training data is the key factor in determining the diagnostic performance.. · The AI models should not be used unsupervised but as a tool to assist radiologists.. · Jaiswal A, Fervers P, Meng F et al. Performance of AI Approaches for COVID-19 Diagnosis Using Chest CT Scans: The Impact of Architecture and Dataset. Rofo 2025; DOI 10.1055/a-2577-3928.

The blind spots on chest computed tomography: what do we miss

Chest computed tomography (CT) is the most frequently performed imaging examination worldwide. Compared with chest radiography, chest CT greatly improves the detection rate and diagnostic accuracy of chest lesions because of the absence of overlapping structures and is the best imaging technique for the observation of chest lesions. However, there are still frequently missed diagnoses during the interpretation process, especially in certain areas or "blind spots", which may possibly be overlooked by radiologists. Awareness of these blind spots is of great significance to avoid false negative results and potential adverse consequences for patients. In this review, we summarize the common blind spots identified in actual clinical practice, encompassing the central areas within the pulmonary parenchyma (including the perihilar regions, paramediastinal regions, and operative area after surgery), trachea and bronchus, pleura, heart, vascular structure, external mediastinal lymph nodes, thyroid, osseous structures, breast, and upper abdomen. In addition to careful review, clinicians can employ several techniques to mitigate or minimize errors arising from these blind spots in film interpretation and reporting. In this review, we also propose technical methods to reduce missed diagnoses, including advanced imaging post-processing techniques such as multiplanar reconstruction (MPR), maximum intensity projection (MIP), artificial intelligence (AI) and structured reporting which can significantly enhance the detection of lesions and improve diagnostic accuracy.

Prior knowledge-guided vision-transformer-based unsupervised domain adaptation for intubation prediction in lung disease at one week

Data-driven approaches have achieved great success in various medical image analysis tasks. However, fully-supervised data-driven approaches require unprecedentedly large amounts of labeled data and often suffer from poor generalization to unseen new data due to domain shifts. Various unsupervised domain adaptation (UDA) methods have been actively explored to solve these problems. Anatomical and spatial priors in medical imaging are common and have been incorporated into data-driven approaches to ease the need for labeled data as well as to achieve better generalization and interpretation. Inspired by the effectiveness of recent transformer-based methods in medical image analysis, the adaptability of transformer-based models has been investigated. How to incorporate prior knowledge for transformer-based UDA models remains under-explored. In this paper, we introduce a prior knowledge-guided and transformer-based unsupervised domain adaptation (PUDA) pipeline. It regularizes the vision transformer attention heads using anatomical and spatial prior information that is shared by both the source and target domain, which provides additional insight into the similarity between the underlying data distribution across domains. Besides the global alignment of class tokens, it assigns local weights to guide the token distribution alignment via adversarial training. We evaluate our proposed method on a clinical outcome prediction task, where Computed Tomography (CT) and Chest X-ray (CXR) data are collected and used to predict the intubation status of patients in a week. Abnormal lesions are regarded as anatomical and spatial prior information for this task and are annotated in the source domain scans. Extensive experiments show the effectiveness of the proposed PUDA method.

Comparison of the Degree of Chest CT Scan Abnormalities in COVID-19 and Influenza Patients

Introduction In the emergency department (ED), COVID-19 and influenza are two common viral diseases. They cause similar symptoms in the respiratory system, and most patients' symptoms are relatively mild. We have reported previously that COVID-19 and influenza infections cause similar abnormalities in chest X-ray readings in the ED. Chest X-ray is a convenient, cost-effective, and useful tool, but it is not as sensitive as computed tomography (CT) scans and does not reveal a high level of detail. To assist physicians in obtaining the most advantageous and specific data to guide the diagnosis and treatment of these diseases, this study aimed to compare the degree of abnormalities on chest CT scans between COVID-19 and influenza patients when they were evaluated in the ED. Methods From a general diagnostic radiologist's teaching files, 87 chest CT scans of COVID-19 patients and 87 chest CT scans of influenza patients were collected. Based on our initial review, four severity categories of lung abnormalities were established. These four categories were normal, mildly abnormal, moderately abnormal, and severely abnormal. Each CT scan was categorized into one of these four categories after being evaluated by two independent raters. The number of CT scans in each category was then counted for the COVID-19 and influenza groups. The resulting number was also divided by the total number of CT scans in each disease group to obtain the percentage within each category. Finally, the results were compared between the COVID-19 and influenza groups. Results In the COVID-19 group, the number and percentage of CT scans in each of the four categories were 10 (11.5%) normal, 44 (50.6%) mildly abnormal, 19 (21.8%) moderately abnormal, and 14 (16.1%) severely abnormal. In the influenza group, there were 13 (14.9%) normal, 48 (55.2%) mildly abnormal, 15 (17.3%) moderately abnormal, and 11 (12.6%) severely abnormal. Chi-square tests revealed no significant difference in these two groups' chest CT abnormalities severity levels. Conclusion Our results indicate that most COVID-19 and influenza patients had mild to moderate abnormalities on their chest CT scans at the time of their ED visits, and the overall severity levels of chest CT abnormalities were similar in both groups of patients.

Trends in CT examination utilization in the emergency department during and after the COVID-19 pandemic

BACKGROUND

The increasing use of CT imaging in emergency departments, despite efforts of reducing low-value imaging, is not fully understood, especially during and after the COVID-19 pandemic. The aim of this study was to investigate the impact of COVID-19 pandemic related measures on trends and volume in CT examinations requested in the emergency department.

METHODS

CT examinations of the head, chest, and/or abdomen-pelvis (n = 161,008), and chest radiographs (n = 113,240) performed at our tertiary care hospital between 01/2014 and 12/2023 were retrospectively analyzed. CT examinations (head, chest, abdomen, dual-region and polytrauma) and chest radiographs requested by the emergency department during (03/2020-03/2022) and after the COVID-19 pandemic (04/2022-12/2023) were compared to a pre-pandemic control period (02/2018-02/2020). Analyses included CT examinations per emergency department visit, and prediction models based on pre-pandemic trends and inpatient data. A regular expressions text search algorithm determined the most common clinical questions.

RESULTS

The usage of dual-region and chest CT examinations were higher during (+ 116,4% and + 115.8%, respectively; p < .001) and after the COVID-19 pandemic (+ 88,4% and + 70.7%, respectively; p < .001), compared to the control period. Chest radiograph usage decreased (-54.1% and - 36.4%, respectively; p < .001). The post-pandemic overall CT examination rate per emergency department visit increased by 4.7%. The prediction model underestimated (p < .001) the growth (dual-region CT: 22.3%, chest CT: 26.7%, chest radiographs: -30.4%), and the rise (p < .001) was higher compared to inpatient data (dual-region CT: 54.8%, chest CT: 52.0%, CR: -32.3%). Post-pandemic, the number of clinical questions to rule out "pulmonary infiltrates", "abdominal pain" and "infection focus" increased up to 235.7% compared to the control period.

CONCLUSIONS

Following the COVID-19 pandemic, chest CT and dual-region CT usage in the emergency department experienced a disproportionate and sustained surge compared to pre-pandemic growth.

Trends in Coronary Artery Anomalies Detection by Coronary Computed Tomography Angiography (CCTA): A Real-Life Comparative Study before and during the COVID-19 Pandemic

BACKGROUND

In the wake of the coronavirus disease 19 (COVID-19) pandemic, affecting healthcare systems globally, urgent research is needed to understand its potential repercussions on the diagnosis and management of cardiovascular disorders. This emphasises the importance of detecting coronary artery anomalies (CAAs), rare conditions that can range from benign to potentially life-threatening manifestations. We aimed to retrospectively assess the impact of the COVID-19 pandemic on the detection of various coronary anomalies using Coronary Computed Tomography Angiography (CCTA) within a regional tertiary cardiology unit in north-eastern Romania, focusing on perceived occurrence in the population under study, types, and related demographic and clinical factors.

METHODS

We analysed CCTA scans and investigated the trends in CAA detection among cardiology patients over a decade. We compared pre-COVID-19 and pandemic-era data to assess the impact of healthcare utilisation, patient behaviour, and diagnostic approaches on anomaly detection.

RESULTS

Our analysis revealed a higher detection rate of CAAs during the pandemic (3.9% versus 2.2%), possibly highlighting differences in patient clinical profile and addressability changes presentation compared to the previous period. Origination and course anomalies, often linked to severe symptoms, were significantly higher pre-COVID-19 (64.1% versus 51.3%). Conversely, intrinsic CAAs, typically asymptomatic or manifesting later in life, notably increased during the pandemic (49.0% versus 61.4%; p = 0.020).

CONCLUSIONS

Our study underscores a significant rise in CAA detection during the COVID-19 era, potentially linked to changes in cardiovascular and respiratory clinical patterns, with advanced imaging modalities like CCTA offering accuracy in identification.

Radiation Overuse in Intensive Care Units

Radiological imaging is essential for acute patient management in Intensive Care Units (ICUs); however, it introduces the risk of exposure to ionizing radiation. This review synthesizes research on radiation exposure in ICU settings, highlighting its rise during the COVID-19 pandemic and the rise in repetitive imaging. Our analysis extends to radiation safety thresholds, revealing that some ICU patients exceed the diagnostic reference level, emphasizing the need to balance diagnostic utility against potential long-term risks, such as cancer. Prospective studies have demonstrated an increase in the median cumulative effective dose in patients with trauma over time, prompting calls for improved awareness and monitoring. Moreover, innovative dose-reduction strategies and optimized protocols, especially in neuro-ICUs, promise to mitigate these risks. This review highlights the essential but risky role of radiological imaging in critical care. It advocates for rigorous radiation management protocols to safeguard patient health while ensuring the continuity of high-quality medical care.

Frequency of Developing COVID-19 Pneumonia in Patients Who Were Vaccinated Double-Dose CoronaVac: Data of the Pandemic Authorized Hospital in Northern Cyprus

Background

RT-PCR is the leading method used in the diagnosis of COVID-19, caused by 2019-nCoV. CT applications also provide a fast and easy diagnosis for detecting pneumonia caused by the SARS-CoV-2 virus. The current study, aimed to compare the lung involvement of vaccinated (two-dose CoronaVac) and unvaccinated patients in the early stage of COVID-19 disease.

Methods

In the current retrospective study, which included patients diagnosed with RT-PCR COVID-19 positivity (n=651) between 01 July 2021-15 September 2021, patient information was obtained from the authorized hospital of the pandemic. Data included patients' chest CT scans and whether patients had been vaccinated (two-dose CoronaVac) information.

Results

The ratio of vaccination with double-dose CoronaVac in positive patients was 74.3%. The ratio of patients with normal lung appearance was 61.8%. It was determined that the ratio of involvement in both lungs of patients who were vaccinated with a double dose was significantly lower than the ratio of involvement in patients who were never vaccinated (p <0.001).

Conclusion

In this study, it was determined that pneumonia cases were less common in individuals vaccinated with double-dose CoronaVac. In this study, it was also determined that the protection of the vaccine was higher in females than in males and that the protection of the double-dose CoronaVac vaccine was higher in the 50-60 age group compared to 60 older patients.

COVID-19-Related Pleural Diseases

Coronavirus disease 2019 (COVID-19)-related pleural diseases are now well recognized. Since the beginning of the pandemic, increasing cases of pleural diseases including pneumothorax, pneumomediastinum, and pleural effusion with severe COVID-19 infection have attracted the attention of physicians and are not incidental or due to barotrauma. The complicated course of COVID-19 illness highlights the complex pathophysiological underpinnings of pleural complications. The management of patients with pneumothorax and pneumomediastinum is challenging as the majority require assisted ventilation; physicians therefore appear to have a low threshold to intervene. Conversely, pleural effusion cases, although sharing some similar patient characteristics with pneumothorax and pneumomediastinum, are in general managed more conservatively. The evidence suggests that patients with COVID-19-related pleural diseases, either due to air leak or effusion, have more severe disease with a worse prognosis. This implies that prompt recognition of these complications and targeted management are key to improve outcomes.

A Pictorial Essay Describing the CT Imaging Features of COVID-19 Cases throughout the Pandemic with a Special Focus on Lung Manifestations and Extrapulmonary Vascular Abdominal Complications

With the Omicron wave, SARS-CoV-2 infections improved, with less lung involvement and few cases of severe manifestations. In this pictorial review, there is a summary of the pathogenesis with particular focus on the interaction of the immune system and gut and lung axis in both pulmonary and extrapulmonary manifestations of COVID-19 and the computed tomography (CT) imaging features of COVID-19 pneumonia from the beginning of the pandemic, describing the typical features of COVID-19 pneumonia following the Delta variant and the atypical features appearing during the Omicron wave. There is also an outline of the typical features of COVID-19 pneumonia in cases of breakthrough infection, including secondary lung complications such as acute respiratory distress disease (ARDS), pneumomediastinum, pneumothorax, and lung pulmonary thromboembolism, which were more frequent during the first waves of the pandemic. Finally, there is a description of vascular extrapulmonary complications, including both ischemic and hemorrhagic abdominal complications.

Predicting omicron pneumonia severity and outcome: a single-center study in Hangzhou, China

Background

In December 2022, there was a large Omicron epidemic in Hangzhou, China. Many people were diagnosed with Omicron pneumonia with variable symptom severity and outcome. Computed tomography (CT) imaging has been proven to be an important tool for COVID-19 pneumonia screening and quantification. We hypothesized that CT-based machine learning algorithms can predict disease severity and outcome in Omicron pneumonia, and we compared its performance with the pneumonia severity index (PSI)-related clinical and biological features.

Methods

Our study included 238 patients with the Omicron variant who have been admitted to our hospital in China from 15 December 2022 to 16 January 2023 (the first wave after the dynamic zero-COVID strategy stopped). All patients had a positive real-time polymerase chain reaction (PCR) or lateral flow antigen test for SARS-CoV-2 after vaccination and no previous SARS-CoV-2 infections. We recorded patient baseline information pertaining to demographics, comorbid conditions, vital signs, and available laboratory data. All CT images were processed with a commercial artificial intelligence (AI) algorithm to obtain the volume and percentage of consolidation and infiltration related to Omicron pneumonia. The support vector machine (SVM) model was used to predict the disease severity and outcome.

Results

The receiver operating characteristic (ROC) area under the curve (AUC) of the machine learning classifier using PSI-related features was 0.85 (accuracy = 87.40%, p < 0.001) for predicting severity while that using CT-based features was only 0.70 (accuracy = 76.47%, p = 0.014). If combined, the AUC was not increased, showing 0.84 (accuracy = 84.03%, p < 0.001). Trained on outcome prediction, the classifier reached the AUC of 0.85 using PSI-related features (accuracy = 85.29%, p < 0.001), which was higher than using CT-based features (AUC = 0.67, accuracy = 75.21%, p < 0.001). If combined, the integrated model showed a slightly higher AUC of 0.86 (accuracy = 86.13%, p < 0.001). Oxygen saturation, IL-6, and CT infiltration showed great importance in both predicting severity and outcome.

Conclusion

Our study provided a comprehensive analysis and comparison between baseline chest CT and clinical assessment in disease severity and outcome prediction in Omicron pneumonia. The predictive model accurately predicts the severity and outcome of Omicron infection. Oxygen saturation, IL-6, and infiltration in chest CT were found to be important biomarkers. This approach has the potential to provide frontline physicians with an objective tool to manage Omicron patients more effectively in time-sensitive, stressful, and potentially resource-constrained environments.

Robust framework for COVID-19 identication from a multicenter dataset of chest CT scans

The main objective of this study is to develop a robust deep learning-based framework to distinguish COVID-19, Community-Acquired Pneumonia (CAP), and Normal cases based on volumetric chest CT scans, which are acquired in different imaging centers using different scanners and technical settings. We demonstrated that while our proposed model is trained on a relatively small dataset acquired from only one imaging center using a specific scanning protocol, it performs well on heterogeneous test sets obtained by multiple scanners using different technical parameters. We also showed that the model can be updated via an unsupervised approach to cope with the data shift between the train and test sets and enhance the robustness of the model upon receiving a new external dataset from a different center. More specifically, we extracted the subset of the test images for which the model generated a confident prediction and used the extracted subset along with the training set to retrain and update the benchmark model (the model trained on the initial train set). Finally, we adopted an ensemble architecture to aggregate the predictions from multiple versions of the model. For initial training and development purposes, an in-house dataset of 171 COVID-19, 60 CAP, and 76 Normal cases was used, which contained volumetric CT scans acquired from one imaging center using a single scanning protocol and standard radiation dose. To evaluate the model, we collected four different test sets retrospectively to investigate the effects of the shifts in the data characteristics on the model's performance. Among the test cases, there were CT scans with similar characteristics as the train set as well as noisy low-dose and ultra-low-dose CT scans. In addition, some test CT scans were obtained from patients with a history of cardiovascular diseases or surgeries. This dataset is referred to as the "SPGC-COVID" dataset. The entire test dataset used in this study contains 51 COVID-19, 28 CAP, and 51 Normal cases. Experimental results indicate that our proposed framework performs well on all test sets achieving total accuracy of 96.15% (95%CI: [91.25-98.74]), COVID-19 sensitivity of 96.08% (95%CI: [86.54-99.5]), CAP sensitivity of 92.86% (95%CI: [76.50-99.19]), Normal sensitivity of 98.04% (95%CI: [89.55-99.95]) while the confidence intervals are obtained using the significance level of 0.05. The obtained AUC values (One class vs Others) are 0.993 (95%CI: [0.977-1]), 0.989 (95%CI: [0.962-1]), and 0.990 (95%CI: [0.971-1]) for COVID-19, CAP, and Normal classes, respectively. The experimental results also demonstrate the capability of the proposed unsupervised enhancement approach in improving the performance and robustness of the model when being evaluated on varied external test sets.

Impact of COVID-19 Pandemic on the Workload of Diagnostic Radiology: A 2-Year Observational Study in a Tertiary Referral Hospital

RATIONALE AND OBJECTIVES

To evaluate the impact of COVID-19 pandemic on diagnostic imaging workload in a tertiary referral hospital.

MATERIALS AND METHODS

Radiological examinations performed in pre-pandemic period (2015-2019) and in pandemic period (2020-2021) were retrospectively included. Based on epidemiological data and restriction measures, four pandemic waves were identified. For each of them, the relative change (RC) in workload was calculated and compared to the 5-year averaged workload in the corresponding pre-COVID-19 periods. Workload variations were also assessed according to technique (radiographs, CT, MRI, ultrasounds), body district (chest, abdomen, breast, musculoskeletal, head/neck, brain/spine, cardiovascular) and care setting (inpatient, outpatient, emergency imaging, pre-admission imaging).

RESULTS

A total of 1384380 examinations were included. In 2020 imaging workload decreased (RC = -11%) compared to the average of the previous 5 years, while in 2021 only a minimal variation (RC = +1%) was observed. During first wave, workload was reduced for all modalities, body regions and types of care setting (RC from -86% to -10%), except for CT (RC = +3%). In subsequent waves, workload increased only for CT (mean RC = +18%) and, regarding body districts, for breast (mean RC = +23%) and cardiovascular imaging (mean RC = +23%). For all other categories, a workload comparable to pre-pandemic period was almost only restored in the fourth wave. In all pandemics periods workload decrease was mainly due to reduced outpatient activity (p < 0.001), while inpatient and emergency imaging was increased (p < 0.001).

CONCLUSION

Evaluating imaging workload changes throughout COVID-19 pandemic helps to understand the response dynamics of radiological services and to improve institutional preparedness to face extreme contingency.

Deep learning attention-guided radiomics for COVID-19 chest radiograph classification

Background

Accurate assessment of coronavirus disease 2019 (COVID-19) lung involvement through chest radiograph plays an important role in effective management of the infection. This study aims to develop a two-step feature merging method to integrate image features from deep learning and radiomics to differentiate COVID-19, non-COVID-19 pneumonia and normal chest radiographs (CXR).

Methods

In this study, a deformable convolutional neural network (deformable CNN) was developed and used as a feature extractor to obtain 1,024-dimensional deep learning latent representation (DLR) features. Then 1,069-dimensional radiomics features were extracted from the region of interest (ROI) guided by deformable CNN's attention. The two feature sets were concatenated to generate a merged feature set for classification. For comparative experiments, the same process has been applied to the DLR-only feature set for verifying the effectiveness of feature concatenation.

Results

Using the merged feature set resulted in an overall average accuracy of 91.0% for three-class classification, representing a statistically significant improvement of 0.6% compared to the DLR-only classification. The recall and precision of classification into the COVID-19 class were 0.926 and 0.976, respectively. The feature merging method was shown to significantly improve the classification performance as compared to using only deep learning features, regardless of choice of classifier (P value <0.0001). Three classes' F1-score were 0.892, 0.890, and 0.950 correspondingly (i.e., normal, non-COVID-19 pneumonia, COVID-19).

Conclusions

A two-step COVID-19 classification framework integrating information from both DLR and radiomics features (guided by deep learning attention mechanism) has been developed. The proposed feature merging method has been shown to improve the performance of chest radiograph classification as compared to the case of using only deep learning features.

A deep learning-based application for COVID-19 diagnosis on CT: The Imaging COVID-19 AI initiative

BACKGROUND

Recently, artificial intelligence (AI)-based applications for chest imaging have emerged as potential tools to assist clinicians in the diagnosis and management of patients with coronavirus disease 2019 (COVID-19).

OBJECTIVES

To develop a deep learning-based clinical decision support system for automatic diagnosis of COVID-19 on chest CT scans. Secondarily, to develop a complementary segmentation tool to assess the extent of lung involvement and measure disease severity.

METHODS

The Imaging COVID-19 AI initiative was formed to conduct a retrospective multicentre cohort study including 20 institutions from seven different European countries. Patients with suspected or known COVID-19 who underwent a chest CT were included. The dataset was split on the institution-level to allow external evaluation. Data annotation was performed by 34 radiologists/radiology residents and included quality control measures. A multi-class classification model was created using a custom 3D convolutional neural network. For the segmentation task, a UNET-like architecture with a backbone Residual Network (ResNet-34) was selected.

RESULTS

A total of 2,802 CT scans were included (2,667 unique patients, mean [standard deviation] age = 64.6 [16.2] years, male/female ratio 1.3:1). The distribution of classes (COVID-19/Other type of pulmonary infection/No imaging signs of infection) was 1,490 (53.2%), 402 (14.3%), and 910 (32.5%), respectively. On the external test dataset, the diagnostic multiclassification model yielded high micro-average and macro-average AUC values (0.93 and 0.91, respectively). The model provided the likelihood of COVID-19 vs other cases with a sensitivity of 87% and a specificity of 94%. The segmentation performance was moderate with Dice similarity coefficient (DSC) of 0.59. An imaging analysis pipeline was developed that returned a quantitative report to the user.

CONCLUSION

We developed a deep learning-based clinical decision support system that could become an efficient concurrent reading tool to assist clinicians, utilising a newly created European dataset including more than 2,800 CT scans.

Computed Tomographic Imaging Features of COVID-19 Pneumonia Caused by the Delta (B.1.617.2) and Omicron (B.1.1.529) Variant in a German Nested Cohort Pilot Study Group

Background: The aim of this study was to evaluate CT (computed tomography) imaging differences for the Delta and the Omicron variant in COVID-19 infection. Methods: The study population was derived from a retrospective study cohort investigating chest CT imaging patterns in vaccinated and nonvaccinated COVID-19 patients. CT imaging patterns of COVID-19 infection were evaluated by qualitative and semiquantitative scoring systems, as well as imaging pattern analysis. Results: A total of 60 patients (70.00% male, 62.53 ± 17.3 years, Delta: 43 patients, Omicron: 17 patients) were included. Qualitative scoring systems showed a significant correlation with virus variants; “typical appearance” and “very high” degrees of suspicion were detected more often in patients with Delta (RSNA: p = 0.003; CO-RADS: p = 0.002; COV-RADS: p = 0.001). Semiquantitative assessment of lung changes revealed a significant association with virus variants in univariate (Delta: 6.3 ± 3.5; Omicron: 3.12 ± 3.2; p = 0.002) and multivariate analysis. The vacuolar sign was significantly associated with the Delta variant (OR: 14.74, 95% CI: [2.32; 2094.7], p = 0.017). Conclusion: The Delta variant had significantly more extensive lung involvement and showed changes classified as “typical” more often than the Omicron variant, while the Omicron variant was more likely associated with CT findings such as “absence of pulmonary changes”. A significant correlation between the Delta variant and the vacuolar sign was observed.

Applications of artificial intelligence in lung ultrasound: Review of deep learning methods for COVID-19 fighting

BACKGROUND

The pandemic outbreak of the novel coronavirus disease (COVID-19) has highlighted the need to combine rapid, non-invasive and widely accessible techniques with the least risk of patient’s cross-infection to achieve a successful early detection and surveillance of the disease. In this regard, the lung ultrasound (LUS) technique has been proved invaluable in both the differential diagnosis and the follow-up of COVID-19 patients, and its potential may be destined to evolve. Recently, indeed, LUS has been empowered through the development of automated image processing techniques.

AIM

To provide a systematic review of the application of artificial intelligence (AI) technology in medical LUS analysis of COVID-19 patients using the preferred reporting items of systematic reviews and meta-analysis (PRISMA) guidelines.

METHODS

A literature search was performed for relevant studies published from March 2020 - outbreak of the pandemic - to 30 September 2021. Seventeen articles were included in the result synthesis of this paper.

RESULTS

As part of the review, we presented the main characteristics related to AI techniques, in particular deep learning (DL), adopted in the selected articles. A survey was carried out on the type of architectures used, availability of the source code, network weights and open access datasets, use of data augmentation, use of the transfer learning strategy, type of input data and training/test datasets, and explainability.

CONCLUSION

Finally, this review highlighted the existing challenges, including the lack of large datasets of reliable COVID-19-based LUS images to test the effectiveness of DL methods and the ethical/regulatory issues associated with the adoption of automated systems in real clinical scenarios.