Quantitative Burden of COVID-19 Pneumonia on Chest CT Predicts Adverse Outcomes: A Post-Hoc Analysis of a Prospective International Registry

Artificial intelligence-based analysis of the spatial distribution of abnormal computed tomography patterns in SARS-CoV-2 pneumonia: association with disease severity

BACKGROUND

The substantial heterogeneity of clinical presentations in patients with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pneumonia still requires robust chest computed tomography analysis to identify high-risk patients. While extension of ground-glass opacity and consolidation from peripheral to central lung fields on chest computed tomography (CT) might be associated with severely ill conditions, quantification of the central-peripheral distribution of ground glass opacity and consolidation in assessments of SARS-CoV-2 pneumonia remains unestablished. This study aimed to examine whether the central-peripheral distributions of ground glass opacity and consolidation were associated with severe outcomes in patients with SARS-CoV-2 pneumonia independent of the whole-lung extents of these abnormal shadows.

METHODS

This multicenter retrospective cohort included hospitalized patients with SARS-CoV-2 pneumonia between January 2020 and August 2021. An artificial intelligence-based image analysis technology was used to segment abnormal shadows, including ground glass opacity and consolidation. The area ratio of ground glass opacity and consolidation to the whole lung (GGO%, CON%) and the ratio of ground glass opacity and consolidation areas in the central lungs to those in the peripheral lungs (GGO(C/P)) and (CON(C/P)) were automatically calculated. Severe outcome was defined as in-hospital death or requirement for endotracheal intubation.

RESULTS

Of 512 enrolled patients, the severe outcome was observed in 77 patients. GGO% and CON% were higher in patients with severe outcomes than in those without. Multivariable logistic models showed that GGO(C/P), but not CON(C/P), was associated with the severe outcome independent of age, sex, comorbidities, GGO%, and CON%.

CONCLUSION

In addition to GGO% and CON% in the whole lung, the higher the ratio of ground glass opacity in the central regions to that in the peripheral regions was, the more severe the outcomes in patients with SARS-CoV-2 pneumonia were. The proposed method might be useful to reproducibly quantify the extension of ground glass opacity from peripheral to central lungs and to estimate prognosis.

Patients Hospitalized for COVID-19 in the Periods of Delta and Omicron Variant Dominance in Greece: Determinants of Severity and Mortality

BACKGROUND

Coronavirus disease 2019 (COVID-19) has been a pandemic since 2020, and depending on the SARS-CoV-2 mutation, different pandemic waves have been observed. The aim of this study was to compare the baseline characteristics of patients in two phases of the pandemic and evaluate possible predictors of mortality.

METHODS

This is a retrospective multicenter observational study that included patients with COVID-19 in 4 different centers in Greece. Patients were divided into two groups depending on the period during which they were infected during the Delta and Omicron variant predominance.

RESULTS

A total of 979 patients (433 Delta, 546 Omicron) were included in the study (median age 67 years (54, 81); 452 [46.2%] female). Compared to the Omicron period, the patients during the Delta period were younger (median age [IQR] 65 [51, 77] vs. 70 [55, 83] years, p < 0.001) and required a longer duration of hospitalization (8 [6, 13] vs. 7 [5, 12] days, p = 0.001), had higher procalcitonin levels (ng/mL): 0.08 [0.05, 0.17] vs. 0.06 [0.02, 0.16], p = 0.005, ferritin levels (ng/mL): 301 [159, 644] vs. 239 [128, 473], p = 0.002, C- reactive protein levels (mg/L): 40.4 [16.7, 98.5] vs. 31.8 [11.9, 81.7], p = 0.003, and lactate dehydrogenase levels (U/L): 277 [221, 375] vs. 255 [205, 329], p < 0.001. The Charlson Comorbidity Index was lower (3 [0, 5] vs. 4 [1, 6], p < 0.001), and the extent of disease on computed tomography (CT) was greater during the Delta wave (p < 0.001). No evidence of a difference in risk of death or admission to the intensive care unit was found between the two groups. Age, cardiovascular events, acute kidney injury during hospitalization, extent of disease on chest CT, D-dimer, and neutrophil/lymphocyte ratio values were identified as independent predictors of mortality for patients in the Delta period. Cardiovascular events and acute liver injury during hospitalization and the PaO2/FiO2 ratio on admission were identified as independent predictors of mortality for patients in the Omicron period.

CONCLUSIONS

In the Omicron wave, patients were older with a higher number of comorbidities, but patients with the Delta variant had more severe disease and a longer duration of hospitalization.

Artificial intelligence-assisted quantification of COVID-19 pneumonia burden from computed tomography improves prediction of adverse outcomes over visual scoring systems

OBJECTIVE

We aimed to evaluate the effectiveness of utilizing artificial intelligence (AI) to quantify the extent of pneumonia from chest CT scans, and to determine its ability to predict clinical deterioration or mortality in patients admitted to the hospital with COVID-19 in comparison to semi-quantitative visual scoring systems.

METHODS

A deep-learning algorithm was utilized to quantify the pneumonia burden, while semi-quantitative pneumonia severity scores were estimated through visual means. The primary outcome was clinical deterioration, the composite end point including admission to the intensive care unit, need for invasive mechanical ventilation, or vasopressor therapy, as well as in-hospital death.

RESULTS

The final population comprised 743 patients (mean age 65  ±  17 years, 55% men), of whom 175 (23.5%) experienced clinical deterioration or death. The area under the receiver operating characteristic curve (AUC) for predicting the primary outcome was significantly higher for AI-assisted quantitative pneumonia burden (0.739, p = 0.021) compared with the visual lobar severity score (0.711, p < 0.001) and visual segmental severity score (0.722, p = 0.042). AI-assisted pneumonia assessment exhibited lower performance when applied for calculation of the lobar severity score (AUC of 0.723, p = 0.021). Time taken for AI-assisted quantification of pneumonia burden was lower (38 ± 10 s) compared to that of visual lobar (328 ± 54 s, p < 0.001) and segmental (698 ± 147 s, p < 0.001) severity scores.

CONCLUSION

Utilizing AI-assisted quantification of pneumonia burden from chest CT scans offers a more accurate prediction of clinical deterioration in patients with COVID-19 compared to semi-quantitative severity scores, while requiring only a fraction of the analysis time.

ADVANCES IN KNOWLEDGE

Quantitative pneumonia burden assessed using AI demonstrated higher performance for predicting clinical deterioration compared to current semi-quantitative scoring systems. Such an AI system has the potential to be applied for image-based triage of COVID-19 patients in clinical practice.

Is the generalizability of a developed artificial intelligence algorithm for COVID-19 on chest CT sufficient for clinical use? Results from the International Consortium for COVID-19 Imaging AI (ICOVAI)

OBJECTIVES

Only few published artificial intelligence (AI) studies for COVID-19 imaging have been externally validated. Assessing the generalizability of developed models is essential, especially when considering clinical implementation. We report the development of the International Consortium for COVID-19 Imaging AI (ICOVAI) model and perform independent external validation.

METHODS

The ICOVAI model was developed using multicenter data (n = 1286 CT scans) to quantify disease extent and assess COVID-19 likelihood using the COVID-19 Reporting and Data System (CO-RADS). A ResUNet model was modified to automatically delineate lung contours and infectious lung opacities on CT scans, after which a random forest predicted the CO-RADS score. After internal testing, the model was externally validated on a multicenter dataset (n = 400) by independent researchers. CO-RADS classification performance was calculated using linearly weighted Cohen's kappa and segmentation performance using Dice Similarity Coefficient (DSC).

RESULTS

Regarding internal versus external testing, segmentation performance of lung contours was equally excellent (DSC = 0.97 vs. DSC = 0.97, p = 0.97). Lung opacities segmentation performance was adequate internally (DSC = 0.76), but significantly worse on external validation (DSC = 0.59, p < 0.0001). For CO-RADS classification, agreement with radiologists on the internal set was substantial (kappa = 0.78), but significantly lower on the external set (kappa = 0.62, p < 0.0001).

CONCLUSION

In this multicenter study, a model developed for CO-RADS score prediction and quantification of COVID-19 disease extent was found to have a significant reduction in performance on independent external validation versus internal testing. The limited reproducibility of the model restricted its potential for clinical use. The study demonstrates the importance of independent external validation of AI models.

KEY POINTS

• The ICOVAI model for prediction of CO-RADS and quantification of disease extent on chest CT of COVID-19 patients was developed using a large sample of multicenter data. • There was substantial performance on internal testing; however, performance was significantly reduced on external validation, performed by independent researchers. The limited generalizability of the model restricts its potential for clinical use. • Results of AI models for COVID-19 imaging on internal tests may not generalize well to external data, demonstrating the importance of independent external validation.

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.

A novel evidence-based predictor tool for hospitalization and length of stay: insights from COVID-19 patients in New York city

Predictive models for key outcomes of coronavirus disease 2019 (COVID-19) can optimize resource utilization and patient outcome. We aimed to design and internally validate a web-based calculator predictive of hospitalization and length of stay (LOS) in a large cohort of COVID-19-positive patients presenting to the Emergency Department (ED) in a New York City health system. The study cohort consisted of consecutive adult (> 18 years) patients presenting to the ED of Mount Sinai Health System hospitals between March 2020 and April 2020, diagnosed with COVID-19. Logistic regression was utilized to construct predictive models for hospitalization and prolonged (> 3 days) LOS. Discrimination was evaluated using area under the receiver operating curve (AUC). Internal validation with bootstrapping was performed, and a web-based calculator was implemented. From 5859 patients, 65% were hospitalized. Independent predictors of hospitalization and extended LOS included older age, chronic kidney disease, elevated maximum temperature, and low minimum oxygen saturation (p < 0.001). Additional predictors of hospitalization included male sex, chronic obstructive pulmonary disease, hypertension, and diabetes. AUCs of 0.881 and 0.770 were achieved for hospitalization and LOS, respectively. Elevated levels of CRP, creatinine, and ferritin were key determinants of hospitalization and LOS (p < 0.05). A calculator was made available under the following URL: https://covid19-outcome-prediction.shinyapps.io/COVID19_Hospitalization_Calculator/ . This study yielded internally validated models that predict hospitalization risk in COVID-19-positive patients, which can be used to optimize resource allocation. Predictors of hospitalization and extended LOS included older age, CKD, fever, oxygen desaturation, elevated C-reactive protein, creatinine, and ferritin.

Rapid quantification of COVID-19 pneumonia burden from computed tomography with convolutional long short-term memory networks

Purpose: Quantitative lung measures derived from computed tomography (CT) have been demonstrated to improve prognostication in coronavirus disease 2019 (COVID-19) patients but are not part of clinical routine because the required manual segmentation of lung lesions is prohibitively time consuming. We aim to automatically segment ground-glass opacities and high opacities (comprising consolidation and pleural effusion). Approach: We propose a new fully automated deep-learning framework for fast multi-class segmentation of lung lesions in COVID-19 pneumonia from both contrast and non-contrast CT images using convolutional long short-term memory (ConvLSTM) networks. Utilizing the expert annotations, model training was performed using five-fold cross-validation to segment COVID-19 lesions. The performance of the method was evaluated on CT datasets from 197 patients with a positive reverse transcription polymerase chain reaction test result for SARS-CoV-2, 68 unseen test cases, and 695 independent controls. Results: Strong agreement between expert manual and automatic segmentation was obtained for lung lesions with a Dice score of 0.89 ± 0.07 ; excellent correlations of 0.93 and 0.98 for ground-glass opacity (GGO) and high opacity volumes, respectively, were obtained. In the external testing set of 68 patients, we observed a Dice score of 0.89 ± 0.06 as well as excellent correlations of 0.99 and 0.98 for GGO and high opacity volumes, respectively. Computations for a CT scan comprising 120 slices were performed under 3 s on a computer equipped with an NVIDIA TITAN RTX GPU. Diagnostically, the automated quantification of the lung burden % discriminate COVID-19 patients from controls with an area under the receiver operating curve of 0.96 (0.95-0.98). Conclusions: Our method allows for the rapid fully automated quantitative measurement of the pneumonia burden from CT, which can be used to rapidly assess the severity of COVID-19 pneumonia on chest CT.

Imaging-based indices combining disease severity and time from disease onset to predict COVID-19 mortality: A cohort study

Background

COVID-19 prognostic factors include age, sex, comorbidities, laboratory and imaging findings, and time from symptom onset to seeking care.

Purpose

The study aim was to evaluate indices combining disease severity measures and time from disease onset to predict mortality of COVID-19 patients admitted to the emergency department (ED).

Materials and methods

All consecutive COVID-19 patients who underwent both computed tomography (CT) and chest X-ray (CXR) at ED presentation between 27/02/2020 and 13/03/2020 were included. CT visual score of disease extension and CXR Radiographic Assessment of Lung Edema (RALE) score were collected. The CT- and CXR-based scores, C-reactive protein (CRP), and oxygen saturation levels (sO₂) were separately combined with time from symptom onset to ED presentation to obtain severity/time indices. Multivariable regression age- and sex-adjusted models without and with severity/time indices were compared. For CXR-RALE, the models were tested in a validation cohort.

Results

Of the 308 included patients, 55 (17.9%) died. In multivariable logistic age- and sex-adjusted models for death at 30 days, severity/time indices showed good discrimination ability, higher for imaging than for laboratory measures (AUC_CT = 0.92, AUC_CXR = 0.90, AUC_CRP = 0.88, AUC_sO2 = 0.88). AUC_CXR was lower in the validation cohort (0.79). The models including severity/time indices performed slightly better than models including measures of disease severity not combined with time and those including the Charlson Comorbidity Index, except for CRP-based models.

Conclusion

Time from symptom onset to ED admission is a strong prognostic factor and provides added value to the interpretation of imaging and laboratory findings at ED presentation.

Impact of Mediastinal Lymphadenopathy on the Severity of COVID-19 Pneumonia: A Nationwide Multicenter Cohort Study

BACKGROUND

We analyzed the differences between clinical characteristics and computed tomography (CT) findings in patients with coronavirus disease 2019 (COVID-19) to establish potential relationships with mediastinal lymphadenopathy and clinical outcomes.

METHODS

We compared the clinical characteristics and CT findings of COVID-19 patients from a nationwide multicenter cohort who were grouped based on the presence or absence of mediastinal lymphadenopathy. Differences between clinical characteristics and CT findings in these groups were analyzed. Univariate and multivariate analyses were performed to determine the impact of mediastinal lymphadenopathy on clinical outcomes.

RESULTS

Of the 344 patients included in this study, 53 (15.4%) presented with mediastinal lymphadenopathy. The rate of diffuse alveolar damage pattern pneumonia and the visual CT scores were significantly higher in patients with mediastinal lymphadenopathy than in those without (P < 0.05). A positive correlation between the number of enlarged mediastinal lymph nodes and visual CT scores was noted in patients with mediastinal lymphadenopathy (Spearman's ρ = 0.334, P < 0.001). Multivariate analysis showed that mediastinal lymphadenopathy was independently associated with a higher risk of intensive care unit (ICU) admission (odds ratio, 95% confidence interval; 3.25, 1.06-9.95) but was not significantly associated with an increased risk of in-hospital death in patients with COVID-19.

CONCLUSION

COVID-19 patients with mediastinal lymphadenopathy had a larger extent of pneumonia than those without. Multivariate analysis adjusted for clinical characteristics and CT findings revealed that the presence of mediastinal lymphadenopathy was significantly associated with ICU admission.

Radiomorphological signs and clinical severity of SARS-CoV-2 lineage B.1.1.7

Objective

We aimed to assess the differences in the severity and chest-CT radiomorphological signs of SARS-CoV-2 B.1.1.7 and non-B.1.1.7 variants.

Methods

We collected clinical data of consecutive patients with laboratory-confirmed COVID-19 and chest-CT imaging who were admitted to the Emergency Department between September 1- November 13, 2020 (non-B.1.1.7 cases) and March 1-March 18, 2021 (B.1.1.7 cases). We also examined the differences in the severity and radiomorphological features associated with COVID-19 pneumonia. Total pneumonia burden (%), mean attenuation of ground-glass opacities and consolidation were quantified using deep-learning research software.

Results

The final population comprised 500 B.1.1.7 and 500 non-B.1.1.7 cases. Patients with B.1.1.7 infection were younger (58.5 ± 15.6 vs 64.8 ± 17.3; p < .001) and had less comorbidities. Total pneumonia burden was higher in the B.1.1.7 patient group (16.1% [interquartile range (IQR):6.0-34.2%] vs 6.6% [IQR:1.2-18.3%]; p < .001). In the age-specific analysis, in patients <60 years B.1.1.7 pneumonia had increased consolidation burden (0.1% [IQR:0.0-0.7%] vs 0.1% [IQR:0.0-0.2%]; p < .001), and severe COVID-19 was more prevalent (11.5% vs  4.9%; p = .032). Mortality rate was similar in all age groups.

Conclusion

Despite B.1.1.7 patients were younger and had fewer comorbidities, they experienced more severe disease than non-B.1.1.7 patients, however, the risk of death was the same between the two groups.

Advances in knowledge

Our study provides data on deep-learning based quantitative lung lesion burden and clinical outcomes of patients infected by B.1.1.7 VOC. Our findings might serve as a model for later investigations, as new variants are emerging across the globe.

The association of obesity with the progression and outcome of COVID-19: The insight from an artificial-intelligence-based imaging quantitative analysis on computed tomography

AIMS

To explore the association of obesity with the progression and outcome of coronavirus disease 2019 (COVID-19) at the acute period and 5-month follow-up from the perspectives of computed tomography (CT) imaging with artificial intelligence (AI)-based quantitative evaluation, which may help to predict the risk of obese COVID-19 patients progressing to severe and critical disease.

MATERIALS AND METHODS

This retrospective cohort enrolled 213 hospitalized COVID-19 patients. Patients were classified into three groups according to their body mass index (BMI): normal weight (from 18.5 to <24 kg/m² ), overweight (from 24 to <28 kg/m² ) and obesity (≥28 kg/m² ).

RESULTS

Compared with normal-weight patients, patients with higher BMI were associated with more lung involvements in lung CT examination (lung lesions volume [cm³ ], normal weight vs. overweight vs. obesity; 175.5[34.0-414.9] vs. 261.7[73.3-576.2] vs. 395.8[101.6-1135.6]; p = 0.002), and were more inclined to deterioration at the acute period. At the 5-month follow-up, the lung residual lesion was more serious (residual total lung lesions volume [cm³ ], normal weight vs. overweight vs. obesity; 4.8[0.0-27.4] vs. 10.7[0.0-55.5] vs. 30.1[9.5-91.1]; p = 0.015), and the absorption rates were lower for higher BMI patients (absorption rates of total lung lesions volume [%], normal weight vs. overweight vs. obesity; 99.6[94.0-100.0] vs. 98.9[85.2-100.0] vs. 88.5[66.5-95.2]; p = 0.013). The clinical-plus-AI parameter model was superior to the clinical-only parameter model in the prediction of disease deterioration (areas under the ROC curve, 0.884 vs. 0.794, p < 0.05).

CONCLUSIONS

Obesity was associated with severe pneumonia lesions on CT and adverse clinical outcomes. The AI-based model with combinational use of clinical and CT parameters had incremental prognostic value over the clinical parameters alone.

Expiratory CT scanning in COVID-19 patients: can we add useful data?

OBJECTIVE

To evaluate small airway disease in COVID-19 patients using the prevalence of air trapping (AT) and correlating it with clinical outcomes. The relationship between CT-based opacities in small blood vessels and ventilation in patients with SARS-CoV-2 pneumonia was also assessed.

METHODS

We retrospectively included 53 patients with positive RT-PCR results for SARS-CoV-2 between March and April of 2020. All subjects underwent HRCT scanning, including inspiratory and expiratory acquisitions. Subjects were divided into two groups based on visual identification of AT. Small blood vessel volumes were estimated by means of cross-sectional areas < 5 mm2 (BV5) derived from automated segmentation algorithms. Mixed-effect models were obtained to represent the BV5 as a function of CT-based lobar opacities and lobar ventilation.

RESULTS

Of the 53 participants, AT was identified in 23 (43.4%). The presence of AT was associated with increased SpO2 at admission (OR = 1.25; 95% CI, 1.07-1.45; p = 0.004) and reduced D-dimer levels (OR = 0.99; 95% CI, 0.99-0.99; p = 0.039). Patients with AT were less likely to be hospitalized (OR = 0.27; 95% CI, 0.08-0.89; p = 0.032). There was a significant but weak inverse correlation between BV5 and CT-based lobar opacities (R2 = 0.19; p = 0.03), as well as a nonsignificant and weak direct correlation between BV5 and lobar ventilation (R2 = 0.08; p = 0.54).

CONCLUSIONS

AT is a common finding in patients with COVID-19 that undergo expiratory CT scanning. The presence of AT may correlate with higher SpO2 at admission, lower D-dimer levels, and fewer hospitalizations when compared with absence of AT. Also, the volume of small pulmonary vessels may negatively correlate with CT opacities but not with lobar ventilation.

Intubation and mortality prediction in hospitalized COVID-19 patients using a combination of convolutional neural network-based scoring of chest radiographs and clinical data

Objective

To predict short-term outcomes in hospitalized COVID-19 patients using a model incorporating clinical variables with automated convolutional neural network (CNN) chest radiograph analysis.

Methods

A retrospective single center study was performed on patients consecutively admitted with COVID-19 between March 14 and April 21 2020. Demographic, clinical and laboratory data were collected, and automated CNN scoring of the admission chest radiograph was performed. The two outcomes of disease progression were intubation or death within 7 days and death within 14 days following admission. Multiple imputation was performed for missing predictor variables and, for each imputed data set, a penalized logistic regression model was constructed to identify predictors and their functional relationship to each outcome. Cross-validated area under the characteristic (AUC) curves were estimated to quantify the discriminative ability of each model.

Results

801 patients (median age 59; interquartile range 46-73 years, 469 men) were evaluated. 36 patients were deceased and 207 were intubated at 7 days and 65 were deceased at 14 days. Cross-validated AUC values for predictive models were 0.82 (95% CI, 0.79-0.86) for death or intubation within 7 days and 0.82 (0.78-0.87) for death within 14 days. Automated CNN chest radiograph score was an important variable in predicting both outcomes.

Conclusion

Automated CNN chest radiograph analysis, in combination with clinical variables, predicts short-term intubation and death in patients hospitalized for COVID-19 infection. Chest radiograph scoring of more severe disease was associated with a greater probability of adverse short-term outcome.

Advances in knowledge

Model-based predictions of intubation and death in COVID-19 can be performed with high discriminative performance using admission clinical data and convolutional neural network-based scoring of chest radiograph severity.

Combined Model of Quantitative Evaluation of Chest Computed Tomography and Laboratory Values for Assessing the Prognosis of Coronavirus Disease 2019

PURPOSE

 To assess the prognostic power of quantitative analysis of chest CT, laboratory values, and their combination in COVID-19 pneumonia.

MATERIALS AND METHODS

 Retrospective analysis of patients with PCR-confirmed COVID-19 pneumonia and chest CT performed between March 07 and November 13, 2020. Volume and percentage (PO) of lung opacifications and mean HU of the whole lung were quantified using prototype software. 13 laboratory values were collected. Negative outcome was defined as death, ICU admittance, mechanical ventilation, or extracorporeal membrane oxygenation. Positive outcome was defined as care in the regular ward or discharge. Logistic regression was performed to evaluate the prognostic value of CT parameters and laboratory values. Independent predictors were combined to establish a scoring system for prediction of prognosis. This score was validated on a separate validation cohort.

RESULTS

 89 patients were included for model development between March 07 and April 27, 2020 (mean age: 60.3 years). 38 patients experienced a negative outcome. In univariate regression analysis, all quantitative CT parameters as well as C-reactive protein (CRP), relative lymphocyte count (RLC), troponin, and LDH were associated with a negative outcome. In a multivariate regression analysis, PO, CRP, and RLC were independent predictors of a negative outcome. Combination of these three values showed a strong predictive value with a C-index of 0.87. A scoring system was established which categorized patients into 4 groups with a risk of 7 %, 30 %, 67 %, or 100 % for a negative outcome. The validation cohort consisted of 28 patients between May 5 and November 13, 2020. A negative outcome occurred in 6 % of patients with a score of 0, 50 % with a score of 1, and 100 % with a score of 2 or 3.

CONCLUSION

 The combination of PO, CRP, and RLC showed a high predictive value for a negative outcome. A 4-point scoring system based on these findings allows easy risk stratification in the clinical routine and performed exceptionally in the validation cohort.

KEY POINTS

  · A high PO is associated with an unfavorable outcome in COVID-19. · PO, CRP, and RLC are independent predictors of an unfavorable outcome, and their combination has strong predictive power. · A 4-point scoring system based on these values allows quick risk stratification in a clinical setting.

CITATION FORMAT

· Scharf G, Meiler S, Zeman F et al. Combined Model of Quantitative Evaluation of Chest Computed Tomography and Laboratory Values for Assessing the Prognosis of Coronavirus Disease 2019. Fortschr Röntgenstr 2022; DOI: 10.1055/a-1731-7905.

COVID-rate: an automated framework for segmentation of COVID-19 lesions from chest CT images

Novel Coronavirus disease (COVID-19) is a highly contagious respiratory infection that has had devastating effects on the world. Recently, new COVID-19 variants are emerging making the situation more challenging and threatening. Evaluation and quantification of COVID-19 lung abnormalities based on chest Computed Tomography (CT) images can help determining the disease stage, efficiently allocating limited healthcare resources, and making informed treatment decisions. During pandemic era, however, visual assessment and quantification of COVID-19 lung lesions by expert radiologists become expensive and prone to error, which raises an urgent quest to develop practical autonomous solutions. In this context, first, the paper introduces an open-access COVID-19 CT segmentation dataset containing 433 CT images from 82 patients that have been annotated by an expert radiologist. Second, a Deep Neural Network (DNN)-based framework is proposed, referred to as the [Formula: see text], that autonomously segments lung abnormalities associated with COVID-19 from chest CT images. Performance of the proposed [Formula: see text] framework is evaluated through several experiments based on the introduced and external datasets. Third, an unsupervised enhancement approach is introduced that can reduce the gap between the training set and test set and improve the model generalization. The enhanced results show a dice score of 0.8069 and specificity and sensitivity of 0.9969 and 0.8354, respectively. Furthermore, the results indicate that the [Formula: see text] model can efficiently segment COVID-19 lesions in both 2D CT images and whole lung volumes. Results on the external dataset illustrate generalization capabilities of the [Formula: see text] model to CT images obtained from a different scanner.

Pixel-based analysis of pulmonary changes on CT lung images due to COVID-19 pneumonia

Objectives: Computed tomography (CT) plays a complementary role in the diagnosis of the pneumonia-burden of COVID-19 disease. However, the low contrast of areas of inflammation on CT images, areas of infection are difficult to identify. The purpose of this study is to develop a post-image-processing method for quantitative analysis of COVID-19 pneumonia-related changes in CT attenuation values using a pixel-based analysis rather than more commonly used clustered focal pneumonia volumes. The COVID-19 pneumonia burden is determined by experienced radiologists in the clinic. Previous AI software was developed for the measurement of COVID-19 lesions based on the extraction of local pneumonia features. In this respect, changes in the pixel levels beyond the clusters may be overlooked by deep learning algorithms. The proposed technique focuses on the quantitative measurement of COVID-19 related pneumonia over the entire lung in pixel-by-pixel fashion rather than only clustered focal pneumonia volumes.

Material and Methods: Fifty COVID-19 and 50 age-matched negative control patients were analyzed using the proposed technique and commercially available artificial intelligence (AI) software. The %pneumonia was calculated using the relative volume of parenchymal pixels within an empirically defined CT density range, excluding pulmonary airways, vessels, and fissures. One-way ANOVA analysis was used to investigate the statistical difference between lobar and whole lung %pneumonia in the negative control and COVID-19 cohorts.

Results: The threshold of high-and-low CT attenuation values related to pneumonia caused by COVID-19 were found to be between ₋642.4 HU and 143 HU. The %pneumonia of the whole lung, left upper, and lower lobes were 8.1 ± 4.4%, 6.1 ± 4.5, and 11.3 ± 7.3% for the COVID-19 cohort, respectively, and statistically different (P < 0.01). Additionally, the pixel-based methods correlate well with existing AI methods and are approximately four times more sensitive to pneumonia particularly at the upper lobes compared with commercial software in COVID-19 patients (P < 0.01).

Conclusion: Pixel-by-pixel analysis can accurately assess pneumonia in COVID-19 patients with CT. Pixel-based techniques produce more sensitive results than AI techniques. Using the proposed novel technique, %pneumonia could be quantitatively calculated not only in the clusters but also in the whole lung with an improved sensitivity by a factor of four compared to AI-based analysis. More significantly, pixel-by-pixel analysis was more sensitive to the upper lobe pneumonia, while AI-based analysis overlooked the upper lung pneumonia region. In the future, this technique can be used to investigate the efficiency of vaccines and drugs and post COVID-19 effects.

Biomarker evaluation for prognostic stratification of patients with COVID-19: the added value of quantitative chest CT

Aim: Pulmonary disease burden and biomarkers are possible predictors of outcomes in patients with COVID-19 and provide complementary information. In this study, the prognostic value of adding quantitative chest computed tomography (CT) to a multiple biomarker approach was evaluated among 148 hospitalized patients with confirmed COVID-19. Materials & methods: Patients admitted between March and July 2020 who were submitted to chest CT and biomarker measurement (troponin I, D-dimer and C-reactive protein) were retrospectively analyzed. Biomarker and tomographic data were compared and associated with death and intensive care unit admission. Results: The number of elevated biomarkers was significantly associated with greater opacification percentages, lower lung volumes and higher death and intensive care unit admission rates. Total lung volume <3.0 l provided further stratification for mortality when combined with biomarker evaluation. Conclusion: Adding automated CT data to a multiple biomarker approach may provide a simple strategy for enhancing risk stratification of patients with COVID-19.

CT Examinations for COVID-19: A Systematic Review of Protocols, Radiation Dose, and Numbers Needed to Diagnose and Predict

Purpose

Although chest CT has been discussed as a first-line test for coronavirus disease 2019 (COVID-19), little research has explored the implications of CT exposure in the population. To review chest CT protocols and radiation doses in COVID-19 publications and explore the number needed to diagnose (NND) and the number needed to predict (NNP) if CT is used as a first-line test.

Materials and Methods

We searched nine highly cited radiology journals to identify studies discussing the CT-based diagnosis of COVID-19 pneumonia. Study-level information on the CT protocol and radiation dose was collected, and the doses were compared with each national diagnostic reference level (DRL). The NND and NNP, which depends on the test positive rate (TPR), were calculated, given a CT sensitivity of 94% (95% confidence interval [CI]: 91%-96%) and specificity of 37% (95% CI: 26%-50%), and applied to the early outbreak in Wuhan, New York, and Italy.

Results

From 86 studies, the CT protocol and radiation dose were reported in 81 (94.2%) and 17 studies (19.8%), respectively. Low-dose chest CT was used more than twice as often as standard-dose chest CT (39.5% vs.18.6%), while the remaining studies (44.2%) did not provide relevant information. The radiation doses were lower than the national DRLs in 15 of the 17 studies (88.2%) that reported doses. The NND was 3.2 scans (95% CI: 2.2-6.0). The NNPs at TPRs of 50%, 25%, 10%, and 5% were 2.2, 3.6, 8.0, 15.5 scans, respectively. In Wuhan, 35418 (TPR, 58%; 95% CI: 27710-56755) to 44840 (TPR, 38%; 95% CI: 35161-68164) individuals were estimated to have undergone CT examinations to diagnose 17365 patients. During the early surge in New York and Italy, daily NNDs changed up to 5.4 and 10.9 times, respectively, within 10 weeks.

Conclusion

Low-dose CT protocols were described in less than half of COVID-19 publications, and radiation doses were frequently lacking. The number of populations involved in a first-line diagnostic CT test could vary dynamically according to daily TPR; therefore, caution is required in future planning.

Application of artificial intelligence in chest imaging for COVID-19

Background: The coronavirus disease 2019 (COVID-19) pandemic has threatened public health. Medical imaging tools such as chest X-ray and computed tomography (CT) play an essential role in the global fight against COVID-19. Recently emerging artificial intelligence (AI) technologies further strengthen the power of imaging tools and help medical professionals. We reviewed the current progress in the development of AI technologies for the diagnostic imaging of COVID-19.Current Concepts: The rapid development of AI, including deep learning, has led to the development of technologies that may assist in the diagnosis and treatment of diseases, prediction of disease risk and prognosis, health index monitoring, and drug development. In the era of the COVID-19 pandemic, AI can improve work efficiency through accurate delineation of infections on chest X-ray and CT images, differentiation of COVID-19 from other diseases, and facilitation of subsequent disease quantification. Moreover, computer-aided platforms help radiologists make clinical decisions for disease diagnosis, tracking, and prognosis.Discussion and Conclusion: We reviewed the current progress in AI technology for chest imaging for COVID-19. However, it is necessary to combine clinical experts’ observations, medical image data, and clinical and laboratory findings for reliable and efficient diagnosis and management of COVID-19. Future AI research should focus on multimodality-based models and how to select the best model architecture for COVID-19 diagnosis and management.

Automated AI-Driven CT Quantification of Lung Disease Predicts Adverse Outcomes in Patients Hospitalized for COVID-19 Pneumonia

The purpose of our work was to assess the independent and incremental value of AI-derived quantitative determination of lung lesions extent on initial CT scan for the prediction of clinical deterioration or death in patients hospitalized with COVID-19 pneumonia. 323 consecutive patients (mean age 65 ± 15 years, 192 men), with laboratory-confirmed COVID-19 and an abnormal chest CT scan, were admitted to the hospital between March and December 2020. The extent of consolidation and all lung opacities were quantified on an initial CT scan using a 3D automatic AI-based software. The outcome was known for all these patients. 85 (26.3%) patients died or experienced clinical deterioration, defined as intensive care unit admission. In multivariate regression based on clinical, biological and CT parameters, the extent of all opacities, and extent of consolidation were independent predictors of adverse outcomes, as were diabetes, heart disease, C-reactive protein, and neutrophils/lymphocytes ratio. The association of CT-derived measures with clinical and biological parameters significantly improved the risk prediction (p = 0.049). Automated quantification of lung disease at CT in COVID-19 pneumonia is useful to predict clinical deterioration or in-hospital death. Its combination with clinical and biological data improves risk prediction.

Quantitative evaluation of COVID-19 pneumonia severity by CT pneumonia analysis algorithm using deep learning technology and blood test results

PURPOSE

To evaluate whether early chest computed tomography (CT) lesions quantified by an artificial intelligence (AI)-based commercial software and blood test values at the initial presentation can differentiate the severity of COVID-19 pneumonia.

MATERIALS AND METHODS

This retrospective study included 100 SARS-CoV-2-positive patients with mild (n = 23), moderate (n = 37) or severe (n = 40) pneumonia classified according to the Japanese guidelines. Univariate Kruskal-Wallis and multivariate ordinal logistic analyses were used to examine whether CT parameters (opacity score, volume of opacity, % opacity, volume of high opacity, % high opacity and mean HU total on CT) as well as blood test parameters [procalcitonin, estimated glomerular filtration rate (eGFR), C-reactive protein, % lymphocyte, ferritin, aspartate aminotransferase, lactate dehydrogenase, alanine aminotransferase, creatine kinase, hemoglobin A1c, prothrombin time, activated partial prothrombin time (APTT), white blood cell count and creatinine] differed by disease severity.

RESULTS

All CT parameters and all blood test parameters except procalcitonin and APPT were significantly different among mild, moderate and severe groups. By multivariate analysis, mean HU total and eGFR were two independent factors associated with severity (p < 0.0001). Cutoff values for mean HU total and eGFR were, respectively, - 801 HU and 77 ml/min/1.73 m² between mild and moderate pneumonia and - 704 HU and 53 ml/min/1.73 m² between moderate and severe pneumonia.

CONCLUSION

The mean HU total of the whole lung, determined by the AI algorithm, and eGFR reflect the severity of COVID-19 pneumonia.

Artificial intelligence-based approaches for COVID-19 patient management

During the highly infectious pandemic of coronavirus disease 2019 (COVID-19), artificial intelligence (AI) has provided support in addressing challenges and accelerating achievements in controlling this public health crisis. It has been applied in fields varying from outbreak forecasting to patient management and drug/vaccine development. In this paper, we specifically review the current status of AI-based approaches for patient management. Limitations and challenges still exist, and further needs are highlighted.

Quantitative analysis based on chest CT classifies common and severe patients with coronavirus disease 2019 pneumonia in Wuhan, China

Objective

This study aimed to compare quantifiable radiologic findings and their dynamic change throughout the clinical course of common and severe coronavirus disease 2019 (COVID-19), and to provide valuable evidence for radiologic classification of the two types of this disease.

Methods

112 patients with laboratory-confirmed COVID-19 were retrospectively analyzed. Volumetric percentage of infection and density of the lung were measured by a computer-aided software. Clinical parameters were recorded to reflect disease progression. Baseline data and dynamic change were compared between two groups and a decision-tree algorithm was developed to determine the cut-off value for classification.

Results

93 patients were finally included and were divided into common group (n = 76) and severe group (n = 17) based on current criteria. Compared with common patients, severe patients experienced shorter advanced stage, peak time and plateau, but longer absorption stage. The dynamic change of volume and density coincided with the clinical course. The interquartile range of volumetric percentage of the two groups were 1.0–7.2% and 11.4–31.2%, respectively. Baseline volumetric percentage of infection was significantly higher in severe group, and the cut-off value of it was 10.10%.

Conclusions

Volumetric percentage between severe and common patients was significantly different. Because serial CT scans are systemically performed in patients with COVID-19 pneumonia, this quantitative analysis can simultaneously provide valuable information for physicians to evaluate their clinical course and classify common and severe patients accurately.

Rapid quantification of COVID-19 pneumonia burden from computed tomography with convolutional LSTM networks

Quantitative lung measures derived from computed tomography (CT) have been demonstrated to improve prognostication in Coronavirus disease 2019 (COVID-19) patients, but are not part of the clinical routine since required manual segmentation of lung lesions is prohibitively time-consuming. We propose a new fully automated deep learning framework for quantification and differentiation between lung lesions in COVID-19 pneumonia from both contrast and non-contrast CT images using convolutional Long Short-Term Memory (LSTM) networks. Utilizing the expert annotations, model training was performed using 5-fold cross-validation to segment ground-glass opacity and high opacity (including consolidation and pleural effusion). The performance of the method was evaluated on CT data sets from 197 patients with positive reverse transcription polymerase chain reaction test result for SARS-CoV-2. Strong agreement between expert manual and automatic segmentation was obtained for lung lesions with a Dice score coefficient of 0.876 ± 0.005; excellent correlations of 0.978 and 0.981 for ground-glass opacity and high opacity volumes. In the external validation set of 67 patients, there was dice score coefficient of 0.767 ± 0.009 as well as excellent correlations of 0.989 and 0.996 for ground-glass opacity and high opacity volumes. Computations for a CT scan comprising 120 slices were performed under 2 seconds on a personal computer equipped with NVIDIA Titan RTX graphics processing unit. Therefore, our deep learning-based method allows rapid fully-automated quantitative measurement of pneumonia burden from CT and may generate the big data with an accuracy similar to the expert readers.

Predictors of Intensive Care Unit Admission or Death in Patients with Coronavirus Disease 2019 Pneumonia in Istanbul, Turkey

We aimed to determine the predictors of intensive care unit (ICU) admission or death in patients with Coronavirus Disease 2019 (COVID-19) pneumonia. This retrospective and single-center study includes patients aged ≥18 years who were diagnosed with COVID-19 pneumonia (laboratory and radiologically confirmed) between March 9 and April 8, 2020. Our composite endpoint was ICU admission or in-hospital death. To evaluate the factors in the composite endpoint, univariate and multivariate logistic regression analyses were performed. A total of 336 patients with COVID-19 pneumonia were recorded. The median age was 54 years [interquartile range (IQR): 21] and 187 (55.7%) were male. Fifty-one (15.2%) patients were admitted to the ICU. In-hospital death occurred in 33 (9.8%) patients. In univariate analysis, 17 parameters were associated with the composite endpoint and procalcitonin had the highest ODDs ratio (OR=36.568 CI=5.145-259.915). Our results revealed that body temperature (OR=1.489 CI=1.023-2.167, p=0.037), peripheral capillary oxygen saturation (SpO2) (OR=0.835 CI=0.773-0.901, p<0.001), and consolidation (>25%) in chest computed tomography (OR=3.170 CI=1.218-8.252, p=0.018) at admission were independent predictors. As a result, increased body temperature, decreased SpO2, a high level of procalcitonin, and degree of consolidation in chest computed tomography may predict a poor prognosis and have utility in the management of patients.

Pulmonary Vascular Thrombosis in COVID-19 Pneumonia

OBJECTIVES

During severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, dramatic endothelial cell damage with pulmonary microvascular thrombosis have been was hypothesized to occur. The aim was to assess whether pulmonary vascular thrombosis (PVT) is due to recurrent thromboembolism from peripheral deep vein thrombosis or to local inflammatory endothelial damage, with a superimposed thrombotic late complication.

DESIGN

Observational study.

SETTING

Medical and intensive care unit wards of a teaching hospital.

PARTICIPANTS

The authors report a subset of patients included in a prospective institutional study (CovidBiob study) with clinical suspicion of pulmonary vascular thromboembolism.

INTERVENTIONS

Computed tomography pulmonary angiography and evaluation of laboratory markers and coagulation profile.

MEASUREMENTS AND MAIN RESULTS

Twenty-eight of 55 (50.9%) patients showed PVT, with a median time interval from symptom onset of 17.5 days. Simultaneous multiple PVTs were identified in 22 patients, with bilateral involvement in 16, mostly affecting segmental/subsegmental pulmonary artery branches (67.8% and 96.4%). Patients with PVT had significantly higher ground glass opacity areas (31.7% [22.9-41] v 17.8% [10.8-22.1], p < 0.001) compared with those without PVT. Remarkably, in all 28 patients, ground glass opacities areas and PVT had an almost perfect spatial overlap. D-dimer level at hospital admission was predictive of PVT.

CONCLUSIONS

The findings identified a specific radiologic pattern of coronavirus disease 2019 (COVID-19) pneumonia with a unique spatial distribution of PVT overlapping areas of ground-glass opacities. These findings supported the hypothesis of a pathogenetic relationship between COVID-19 lung inflammation and PVT and challenged the previous definition of pulmonary embolism associated with COVID-19 pneumonia.