CT quantitative analysis in patients with severe coronavirus disease 2019 and its correlation with laboratory examination results

OBJECTIVE: To quantitatively analyze the longitudinal changes of ground-glass opacity (GGO), consolidation and total lesion in patients infected with severe coronavirus disease 2019 (COVID-19), and its correlation with laboratory examination results. MATERIALS AND METHODS: All 76 computed tomography (CT) images and laboratory examination results from the admission to discharge of 15 patients confirmed with severe COVID-19 were reviewed, whereas the GGO volume ratio, consolidation volume ratio, and total lesion volume ratio in different stages were analyzed. The correlations of lesions volume ratio and laboratory examination results were investigated. RESULTS: Four stages were identified based on the degree of lung involvement from day 1 to day 28 after disease onset. GGO was the most common CT manifestation in the four stages. The peak of lung involvement was at around stage 2, and corresponding total lesion volume ratio, GGO volume ratio, and consolidation volume ratio were 17.48 (13.44−24.33), 12.11 (7.34−17.08), and 5.51 (2.58−8.58), respectively. Total lesion volume ratio was positively correlated with neutrophil percentage, C-reactive protein (CRP), high-sensitivity CRP (Hs-CRP), procalcitonin, lactate dehydrogenase (LD), and creatine kinase isoenzyme MB (CK-MB), but negatively correlated with lymphocyte count, lymphocyte percentage, arterial oxygen saturation, and arterial oxygen tension. Consolidation volume ratio was correlated with most above laboratory examination results except Hs-CRP, LD, and CK-MB. GGO, however, was only correlated with lymphocyte count. CONCLUSION: CT quantitative parameters could show longitudinal changes well. Total lesion volume ratio and consolidation volume ratio are well correlated with laboratory examination results, suggesting that CT quantitative parameters may be an effective tool to reflect the changes in the condition.

Clinical Applicable AI System Based on Deep Learning Algorithm for Differentiation of Pulmonary Infectious Disease.

Objective: To assess the performance of a novel deep learning (DL)-based artificial intelligence (AI) system in classifying computed tomography (CT) scans of pneumonia patients into different groups, as well as to present an effective clinically relevant machine learning (ML) system based on medical image identification and clinical feature interpretation to assist radiologists in triage and diagnosis. Methods: The 3,463 CT images of pneumonia used in this multi-center retrospective study were divided into four categories: bacterial pneumonia (n = 507), fungal pneumonia (n = 126), common viral pneumonia (n = 777), and COVID-19 (n = 2,053). We used DL methods based on images to distinguish pulmonary infections. A machine learning (ML) model for risk interpretation was developed using key imaging (learned from the DL methods) and clinical features. The algorithms were evaluated using the areas under the receiver operating characteristic curves (AUCs). Results: The median AUC of DL models for differentiating pulmonary infection was 99.5% (COVID-19), 98.6% (viral pneumonia), 98.4% (bacterial pneumonia), 99.1% (fungal pneumonia), respectively. By combining chest CT results and clinical symptoms, the ML model performed well, with an AUC of 99.7% for SARS-CoV-2, 99.4% for common virus, 98.9% for bacteria, and 99.6% for fungus. Regarding clinical features interpreting, the model revealed distinctive CT characteristics associated with specific pneumonia: in COVID-19, ground-glass opacity (GGO) [92.5%; odds ratio (OR), 1.76; 95% confidence interval (CI): 1.71-1.86]; larger lesions in the right upper lung (75.0%; OR, 1.12; 95% CI: 1.03-1.25) with viral pneumonia; older age (57.0 years ± 14.2, OR, 1.84; 95% CI: 1.73-1.99) with bacterial pneumonia; and consolidation (95.8%, OR, 1.29; 95% CI: 1.05-1.40) with fungal pneumonia. Conclusion: For classifying common types of pneumonia and assessing the influential factors for triage, our AI system has shown promising results. Our ultimate goal is to assist clinicians in making quick and accurate diagnoses, resulting in the potential for early therapeutic intervention.

An ensemble learning method based on ordinal regression for COVID-19 diagnosis from chest CT.

Coronavirus disease 2019 (COVID-19) has brought huge losses to the world, and it remains a great threat to public health. X-ray computed tomography (CT) plays a central role in the management of COVID-19. Traditional diagnosis with pulmonary CT images is time-consuming and error-prone, which could not meet the need for precise and rapid COVID-19 screening. Nowadays, deep learning (DL) has been successfully applied to CT image analysis, which assists radiologists in workflow scheduling and treatment planning for patients with COVID-19. Traditional methods use cross-entropy as the loss function with a Softmax classifier following a fully-connected layer. Most DL-based classification methods target intraclass relationships in a certain class (early, progressive, severe, or dissipative phases), ignoring the natural order of different phases of the disease progression,i.e.,from an early stage and progress to a late stage. To learn both intraclass and interclass relationships among different stages and improve the accuracy of classification, this paper proposes an ensemble learning method based on ordinal regression, which leverages the ordinal information on COVID-19 phases. The proposed method uses multi-binary, neuron stick-breaking (NSB), and soft labels (SL) techniques, and ensembles the ordinal outputs through a median selection. To evaluate our method, we collected 172 confirmed cases. In a 2-fold cross-validation experiment, the accuracy is increased by 22% compared with traditional methods when we use modified ResNet-18 as the backbone. And precision, recall, andF1-score are also improved. The experimental results show that our proposed method achieves a better classification performance than the traditional methods, which helps establish guidelines for the classification of COVID-19 chest CT images.

Elaboration of a Radiomics Strategy for the Prediction of the Re-positive Cases in the Discharged Patients With COVID-19.

Objective: A considerable part of COVID-19 patients were found to be re-positive in the SARS-CoV-2 RT-PCR test after discharge. Early prediction of re-positive COVID-19 cases is of critical importance in determining the isolation period and developing clinical protocols. Materials and Methods: Ninety-one patients discharged from Wanzhou Three Gorges Central Hospital, Chongqing, China, from February 10, 2020 to March 3, 2020 were administered nasopharyngeal swab SARS-CoV-2 tests within 12-14 days, and 50 eligible patients (32 male and 18 female) with completed data were enrolled. Average age was 48 ± 11.5 years. All patients underwent non-enhanced chest CT on admission. A total of 568 radiomics features were extracted from the CT images, and 17 clinical factors were collected based on the medical record. Student's t-test and support vector machine-based recursive feature elimination (SVM-RFE) method were used to determine an optimal subset of features for the discriminative model development. Results: After Student's t-test, 62 radiomics features showed significant inter-group differences (p < 0.05) between the re-positive and negative cases, and none of the clinical features showed significant differences. These significant features were further selected by SVM-RFE algorithm, and a more compact feature subset containing only two radiomics features was finally determined, achieving the best predictive performance with the accuracy and area under the curve of 72.6% and 0.773 for the identification of the re-positive case. Conclusion: The proposed radiomics method has preliminarily shown potential in identifying the re-positive cases among the recovered COVID-19 patients after discharge. More strategies are to be integrated into the current pipeline to improve its precision, and a larger database with multi-clinical enrollment is required to extensively verify its performance.

PAM-DenseNet: A Deep Convolutional Neural Network for Computer-Aided COVID-19 Diagnosis.

Currently, several convolutional neural network (CNN)-based methods have been proposed for computer-aided COVID-19 diagnosis based on lung computed tomography (CT) scans. However, the lesions of pneumonia in CT scans have wide variations in appearances, sizes, and locations in the lung regions, and the manifestations of COVID-19 in CT scans are also similar to other types of viral pneumonia, which hinders the further improvement of CNN-based methods. Delineating infection regions manually is a solution to this issue, while excessive workload of physicians during the epidemic makes it difficult for manual delineation. In this article, we propose a CNN called dense connectivity network with parallel attention module (PAM-DenseNet), which can perform well on coarse labels without manually delineated infection regions. The parallel attention module automatically learns to strengthen informative features from both channelwise and spatialwise simultaneously, which can make the network pay more attention to the infection regions without any manual delineation. The dense connectivity structure performs feature maps reuse by introducing direct connections from previous layers to all subsequent layers, which can extract representative features from fewer CT slices. The proposed network is first trained on 3530 lung CT slices selected from 382 COVID-19 lung CT scans, 372 lung CT scans infected by other pneumonia, and 200 normal lung CT scans to obtain a pretrained model for slicewise prediction. We then apply this pretrained model to a CT scans dataset containing 94 COVID-19 CT scans, 93 other pneumonia CT scans, and 93 normal lung scans, and achieve patientwise prediction through a voting mechanism. The experimental results show that the proposed network achieves promising results with an accuracy of 94.29%, a precision of 93.75%, a sensitivity of 95.74%, and a specificity of 96.77%, which is comparable to the methods that are based on manually delineated infection regions.

Chest CT Severity Score: An Imaging Tool for Assessing Severe COVID-19.

PURPOSE: To evaluate the value of chest CT severity score (CT-SS) in differentiating clinical forms of coronavirus disease 2019 (COVID-19). MATERIALS AND METHODS: A total of 102 patients with COVID-19 confirmed by a positive result from real-time reverse transcription polymerase chain reaction on throat swabs who underwent chest CT (53 men and 49 women, 15-79 years old, 84 cases with mild and 18 cases with severe disease) were included in the study. The CT-SS was defined by summing up individual scores from 20 lung regions; scores of 0, 1, and 2 were respectively assigned for each region if parenchymal opacification involved 0%, less than 50%, or equal to or more than 50% of each region (theoretic range of CT-SS from 0 to 40). The clinical and laboratory data were collected, and patients were clinically subdivided according to disease severity according to the Chinese National Health Commission guidelines. RESULTS: The posterior segment of upper lobe (left, 68 of 102; right, 68 of 102), superior segment of lower lobe (left, 79 of 102; right, 79 of 102), lateral basal segment (left, 79 of 102; right, 70 of 102), and posterior basal segment of lower lobe (left, 81 of 102; right, 83 of 102) were the most frequently involved sites in COVID-19. Lung opacification mainly involved the lower lobes, in comparison with middle-upper lobes. No significant differences in distribution of the disease were seen between right and left lungs. The individual scores in each lung and the total CT-SS were higher in severe COVID-19 when compared with mild cases (P < .05). The optimal CT-SS threshold for identifying severe COVID-19 was 19.5 (area under curve = 0.892), with 83.3% sensitivity and 94% specificity. CONCLUSION: The CT-SS could be used to evaluate the severity of pulmonary involvement quickly and objectively in patients with COVID-19.© RSNA, 2020.

CT imaging features of different clinical types of COVID-19 calculated by AI system: a Chinese multicenter study.

BACKGROUND: The study is designed to explore the chest CT features of different clinical types of coronavirus disease 2019 (COVID-19) pneumonia based on a Chinese multicenter dataset using an artificial intelligence (AI) system. METHODS: A total of 164 patients confirmed COVID-19 were retrospectively enrolled from 6 hospitals. All patients were divided into the mild type (136 cases) and the severe type (28 cases) according to their clinical manifestations. The total CT severity score and quantitative CT features were calculated by AI pneumonia detection and evaluation system with correction by radiologists. The clinical and CT imaging features of different types were analyzed. RESULTS: It was observed that patients in the severe type group were older than the mild type group. Round lesions, Fan-shaped lesions, crazy-paving pattern, fibrosis, "white lung", pleural thickening, pleural indentation, mediastinal lymphadenectasis were more common in the CT images of severe patients than in the mild ones. A higher total lung severity score and scores of each lobe were observed in the severe group, with higher scores in bilateral lower lobes of both groups. Further analysis showed that the volume and number of pneumonia lesions and consolidation lesions in overall lung were higher in the severe group, and showed a wider distribution in the lower lobes of bilateral lung in both groups. CONCLUSIONS: Chest CT of patients with severe COVID-19 pneumonia showed more consolidative and progressive lesions. With the assistance of AI, CT could evaluate the clinical severity of COVID-19 pneumonia more precisely and help the early diagnosis and surveillance of the patients.

Chest CT Findings in Patients With Coronavirus Disease 2019 and Its Relationship With Clinical Features.

OBJECTIVES: The aim of this study was to investigate the chest computed tomography (CT) findings in patients with confirmed coronavirus disease 2019 (COVID-19) and to evaluate its relationship with clinical features. MATERIALS AND METHODS: Study sample consisted of 80 patients diagnosed as COVID-19 from January to February 2020. The chest CT images and clinical data were reviewed, and the relationship between them was analyzed. RESULTS: Totally, 80 patients diagnosed with COVID-19 were included. With regards to the clinical manifestations, 58 (73%) of the 80 patients had cough, and 61 (76%) of the 80 patients had high temperature levels. The most frequent CT abnormalities observed were ground glass opacity (73/80 cases, 91%), consolidation (50/80 cases, 63%), and interlobular septal thickening (47/80, 59%). Most of the lesions were multiple, with an average of 12 ± 6 lung segments involved. The most common involved lung segments were the dorsal segment of the right lower lobe (69/80, 86%), the posterior basal segment of the right lower lobe (68/80, 85%), the lateral basal segment of the right lower lobe (64/80, 80%), the dorsal segment of the left lower lobe (61/80, 76%), and the posterior basal segment of the left lower lobe (65/80, 81%). The average pulmonary inflammation index value was (34% ± 20%) for all the patients. Correlation analysis showed that the pulmonary inflammation index value was significantly correlated with the values of lymphocyte count, monocyte count, C-reactive protein, procalcitonin, days from illness onset, and body temperature (P < 0.05). CONCLUSIONS: The common chest CT findings of COVID-19 are multiple ground glass opacity, consolidation, and interlobular septal thickening in both lungs, which are mostly distributed under the pleura. There are significant correlations between the degree of pulmonary inflammation and the main clinical symptoms and laboratory results. Computed tomography plays an important role in the diagnosis and evaluation of this emerging global health emergency.

CT Quantitative Analysis and Its Relationship with Clinical Features for Assessing the Severity of Patients with COVID-19.

OBJECTIVE: To investigate the value of initial CT quantitative analysis of ground-glass opacity (GGO), consolidation, and total lesion volume and its relationship with clinical features for assessing the severity of coronavirus disease 2019 (COVID-19). MATERIALS AND METHODS: A total of 84 patients with COVID-19 were retrospectively reviewed from January 23, 2020 to February 19, 2020. Patients were divided into two groups: severe group (n = 23) and non-severe group (n = 61). Clinical symptoms, laboratory data, and CT findings on admission were analyzed. CT quantitative parameters, including GGO, consolidation, total lesion score, percentage GGO, and percentage consolidation (both relative to total lesion volume) were calculated. Relationships between the CT findings and laboratory data were estimated. Finally, a discrimination model was established to assess the severity of COVID-19. RESULTS: Patients in the severe group had higher baseline neutrophil percentage, increased high-sensitivity C-reactive protein (hs-CRP) and procalcitonin levels, and lower baseline lymphocyte count and lymphocyte percentage (p < 0.001). The severe group also had higher GGO score (p < 0.001), consolidation score (p < 0.001), total lesion score (p < 0.001), and percentage consolidation (p = 0.002), but had a lower percentage GGO (p = 0.008). These CT quantitative parameters were significantly correlated with laboratory inflammatory marker levels, including neutrophil percentage, lymphocyte count, lymphocyte percentage, hs-CRP level, and procalcitonin level (p < 0.05). The total lesion score demonstrated the best performance when the data cut-off was 8.2%. Furthermore, the area under the curve, sensitivity, and specificity were 93.8% (confidence interval [CI]: 86.8-100%), 91.3% (CI: 69.6-100%), and 91.8% (CI: 23.0-98.4%), respectively. CONCLUSION: CT quantitative parameters showed strong correlations with laboratory inflammatory markers, suggesting that CT quantitative analysis might be an effective and important method for assessing the severity of COVID-19, and may provide additional guidance for planning clinical treatment strategies.

CT imaging changes of corona virus disease 2019(COVID-19): a multi-center study in Southwest China.

BACKGROUND: Since the first case of a coronavirus disease 2019 (COVID-19) infection pneumonia was detected in Wuhan, China, a series of confirmed cases of the COVID-19 were found in Southwest China. The aim of this study was to describe the imaging manifestations of hospitalized patients with confirmed COVID-19 infection in southwest China. METHODS: In this retrospective study, data were collected from 131 patients with confirmed coronavirus disease 2019 (COVID-19) from 3 Chinese hospitals. Their common clinical manifestations, as well as characteristics and evolvement features of chest CT images, were analyzed. RESULTS: A total of 100 (76%) patients had a history of close contact with people living in Wuhan, Hubei. The clinical manifestations of COVID-19 included cough, fever. Most of the lesions identified in chest CT images were multiple lesions of bilateral lungs, lesions were more localized in the peripheral lung, 109 (83%) patients had more than two lobes involved, 20 (15%) patients presented with patchy ground glass opacities, patchy ground glass opacities and consolidation of lesions co-existing in 61 (47%) cases. Complications such as pleural thickening, hydrothorax, pericardial effusion, and enlarged mediastinal lymph nodes were detected but only in rare cases. For the follow-up chest CT examinations (91 cases), We found 66 (73%) cases changed very quickly, with an average of 3.5 days, 25 cases (27%) presented absorbed lesions, progression was observed in 41 cases (46%), 25 (27%) cases showed no significant changes. CONCLUSION: Chest CT plays an important role in diagnosing COVID-19. The imaging pattern of multifocal peripheral ground glass or mixed consolidation is highly suspicious of COVID-19, that can quickly change over a short period of time.

Coronavirus disease 2019: initial chest CT findings.

OBJECTIVES: To systematically analyze CT findings during the early and progressive stages of natural course of coronavirus disease 2019 and also to explore possible changes in pulmonary parenchymal abnormalities during these two stages. METHODS: We retrospectively reviewed the initial chest CT data of 62 confirmed coronavirus disease 2019 patients (34 men, 28 women; age range 20-91 years old) who did not receive any antiviral treatment between January 21 and February 4, 2020, in Chongqing, China. Patients were assigned to the early-stage group (onset of symptoms within 4 days) or progressive-stage group (onset of symptoms within 4-7 days) for analysis. CT characteristics and the distribution, size, and CT score of pulmonary parenchymal abnormalities were assessed. RESULTS: In our study, the major characteristic of coronavirus disease 2019 was ground-glass opacity (61.3%), followed by ground-glass opacity with consolidation (35.5%), rounded opacities (25.8%), a crazy-paving pattern (25.8%), and an air bronchogram (22.6%). No patient presented cavitation, a reticular pattern, or bronchial wall thickening. The CT scores of the progressive-stage group were significantly greater than those of the early-stage group (p = 0.004). CONCLUSIONS: Multiple ground-glass opacities with consolidations in the periphery of the lungs were the primary CT characteristic of coronavirus disease 2019. CT score can be used to evaluate the severity of the disease. If these typical alterations are found, then the differential diagnosis of coronavirus disease 2019 must be considered. KEY POINTS: • Multiple GGOs with consolidations in the periphery of the lungs were the primary CT characteristic of COVID-19. • The halo sign may be a special CT feature in the early-stage COVID-19 patients. • Significantly increased CT score may indicate the aggravation of COVID-19 in the progressive stage.