Differentiation of COVID-19 conditions in planar chest radiographs using optimized convolutional neural networks

In this study, an attempt has been made to differentiate Novel Coronavirus-2019 (COVID-19) conditions from healthy subjects in Chest radiographs using a simplified end-to-end Convolutional Neural Network (CNN) model and occlusion sensitivity maps. Early detection and faster automated screening of the COVID-19 patients is essential. For this, the images are considered from publicly available datasets. Significant biomarkers representing critical image features are extracted from CNN by experimentally investigating on cross-validation methods and hyperparameter settings. The performance of the network is evaluated using standard metrics. Perturbation based occlusion sensitivity maps are employed on the features obtained from the classification model to visualise the localization of abnormal areas. Results demonstrate that the simplified CNN model with optimised parameters is able to extract significant features with a sensitivity of 97.35% and F-measure of 96.71% to detect COVID-19 images. The algorithm achieves an Area Under the Curve-Receiver Operating Characteristic score of 99.4% with Matthews correlation coefficient of 0.93. High value of Diagnostic odds ratio is also obtained. Occlusion sensitivity maps provide precise localization of abnormal regions by identifying COVID-19 conditions. As early detection through chest radiographic images are useful for automated screening of the disease, this method appears to be clinically relevant in providing a visual diagnostic solution using a simplified and efficient model.

Comparison of the Characteristics of Asymptomatic and Presymptomatic Patients with Coronavirus Disease 2019 in the Republic of Korea

PURPOSE

This retrospective study aimed to evaluate the baseline characteristics of asymptomatic patients with coronavirus disease 2019 at admission and to follow-up their clinical manifestations and radiological findings during hospitalization.

METHODS

Patients with coronavirus disease 2019 who were asymptomatic at admission were divided into two groups-those with no symptoms until discharge (group A) and those who developed symptoms after admission (group B). Patients who could not express their own symptoms were excluded.

RESULTS

Overall, 127 patients were enrolled in the study, of whom 19 and 108 were assigned to groups A and B, respectively. The mean age and median C-reactive protein level were higher in group B than in group A. All patients in group A and one-third of patients in group B had normal initial chest radiographs; 15.8% and 48.1% of patients in groups A and B, respectively, had pneumonia during hospitalization. One patient in group B, whose condition was not severe at the time of admission, deteriorated due to aggravated pneumonia and was transferred to a tertiary hospital.

CONCLUSION

We summarize the clinical characteristics during hospitalization of patients with coronavirus disease 2019 who were purely asymptomatic at the time of admission. The majority of asymptomatic patients with coronavirus disease 2019 were discharged without significant events during hospitalization. However, it may be difficult to predict subsequent events from initial chest radiographs or oxygen saturation at admission.

Effect of deep learning-based assistive technology use on chest radiograph interpretation by emergency department physicians: a prospective interventional simulation-based study

BACKGROUND

Interpretation of chest radiographs (CRs) by emergency department (ED) physicians is inferior to that by radiologists. Recent studies have investigated the effect of deep learning-based assistive technology on CR interpretation (DLCR), although its relevance to ED physicians remains unclear. This study aimed to investigate whether DLCR supports CR interpretation and the clinical decision-making of ED physicians.

METHODS

We conducted a prospective interventional study using a web-based performance assessment system. Study participants were recruited through the official notice targeting board for certified emergency physicians and residents working at the present ED. Of the eight ED physicians who volunteered to participate in the study, seven ED physicians were included, while one participant declared withdrawal during performance assessment. Seven physicians' CR interpretations and clinical decision-making were assessed based on the clinical data from 388 patients, including detecting the target lesion with DLCR. Participant performance was evaluated by area under the receiver operating characteristic curve (AUROC), sensitivity, specificity, and accuracy analyses; decision-making consistency was measured by kappa statistics. ED physicians with < 24 months of experience were defined as 'inexperienced'.

RESULTS

Among the 388 simulated cases, 259 (66.8%) had CR abnormality. Their median value of abnormality score measured by DLCR was 59.3 (31.77, 76.25) compared to a score of 3.35 (1.57, 8.89) for cases of normal CR. There was a difference in performance between ED physicians working with and without DLCR (AUROC: 0.801, P < 0.001). The diagnostic sensitivity and accuracy of CR were higher for all ED physicians working with DLCR than for those working without it. The overall kappa value for decision-making consistency was 0.902 (95% confidence interval [CI] 0.884-0.920); concurrently, the kappa value for the experienced group was 0.956 (95% CI 0.934-0.979), and that for the inexperienced group was 0.862 (95% CI 0.835-0.889).

CONCLUSIONS

This study presents preliminary evidence that ED physicians using DLCR in a clinical setting perform better at CR interpretation than their counterparts who do not use this technology. DLCR use influenced the clinical decision-making of inexperienced physicians more strongly than that of experienced physicians. These findings require prospective validation before DLCR can be recommended for use in routine clinical practice.

Limitations of cardiothoracic ratio derived from chest radiographs to predict real heart size: comparison with magnetic resonance imaging

Background

Cardiothoracic ratio (CTR) in chest radiographs is still widely used to estimate cardiac size despite the advent of newer imaging techniques. We hypothesise that a universal CTR cut-off value of 50% is a poor indicator of cardiac enlargement. Our aim was to compare CTR with volumetric and functional parameters derived from cardiac magnetic resonance imaging (MRI).

Methods

309 patients with a chest radiograph and cardiac MRI acquired within a month were reviewed to assess how CTR correlates with multiple cardiac MRI variables: bi-ventricular EDV (absolute and indexed to body surface area), EF, indexed total heart volume and bi-atrial areas. In addition, we have also determined CTR accuracy by creating multiple ROC curves with the described variables.

Results

All cardiac MRI variables correlate weakly but statistically significantly with CTR. This weak correlation is explained by a substantial overlap of cardiac MRI parameters in patients with normal and increased CTR. For all variables, CTR was only mildly to moderately better than a chance to discriminate cardiac enlargement (AUC 0.6–0.7). Large CTR values (> 55%) are specific but not sensitive, while low CTR values (< 45%) are sensitive but not specific. Values in between are not sensitive nor specific.

Conclusions

CTR correlates weakly with true chamber size assessed by gold standard cardiac MRI and has a weak discriminatory power. Thus, clinical decisions based on intermediate CTRs (45–55%) should be avoided. Large CTRs (> 55%) are likely indicative of true heart chamber enlargement. Low CTRs (< 45%) are likely indicative of normal heart size.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13244-021-01097-0.

ACR Appropriateness Criteria® Chronic Cough

Chronic cough is defined by a duration lasting at least 8 weeks. The most common causes of chronic cough include smoking-related lung disease, upper airway cough syndrome, asthma, gastroesophageal reflux disease, and nonasthmatic eosinophilic bronchitis. The etiology of chronic cough in some patients may be difficult to localize to an isolated source and is often multifactorial. The complex pathophysiology, clinical presentation, and variable manifestations of chronic cough underscore the challenges faced by clinicians in the evaluation and management of these patients. Imaging plays a role in the initial evaluation, although there is a lack of high-quality evidence guiding which modalities are useful and at what point in time the clinical evaluation should be performed. The American College of Radiology Appropriateness Criteria are evidence-based guidelines for specific clinical conditions that are reviewed annually by a multidisciplinary expert panel. The guideline development and revision include an extensive analysis of current medical literature from peer reviewed journals and the application of well-established methodologies (RAND/UCLA Appropriateness Method and Grading of Recommendations Assessment, Development, and Evaluation or GRADE) to rate the appropriateness of imaging and treatment procedures for specific clinical scenarios. In those instances where evidence is lacking or equivocal, expert opinion may supplement the available evidence to recommend imaging or treatment.

Limited generalizability of deep learning algorithm for pediatric pneumonia classification on external data

Purpose

(1) Develop a deep learning system (DLS) to identify pneumonia in pediatric chest radiographs, and (2) evaluate its generalizability by comparing its performance on internal versus external test datasets.

Methods

Radiographs of patients between 1 and 5 years old from the Guangzhou Women and Children’s Medical Center (Guangzhou dataset) and NIH ChestXray14 dataset were included. We utilized 5232 radiographs from the Guangzhou dataset to train a ResNet-50 deep convolutional neural network (DCNN) to identify pediatric pneumonia. DCNN testing was performed on a holdout set of 624 radiographs from the Guangzhou dataset (internal test set) and 383 radiographs from the NIH ChestXray14 dataset (external test set). Receiver operating characteristic curves were generated, and area under the curve (AUC) was compared via DeLong parametric method. Colored heatmaps were generated using class activation mapping (CAM) to identify important image pixels for DCNN decision-making.

Results

The DCNN achieved AUC of 0.95 and 0.54 for identifying pneumonia on internal and external test sets, respectively (p < 0.0001). Heatmaps generated by the DCNN showed the algorithm focused on clinically relevant features for images from the internal test set, but not for images from the external test set.

Conclusion

Our model had high performance when tested on an internal dataset but significantly lower accuracy when tested on an external dataset. Likewise, marked differences existed in the clinical relevance of features highlighted by heatmaps generated from internal versus external datasets. This study underscores potential limitations in the generalizability of such DLS models.

Gender and Age Differences in Cardiac Size Parameters of Ghanaian Adults: Can One Parameter Fit All? Part Two

Background

The cardiothoracic ratio (CTR) is a radiographic parameter commonly used in assessing the size of the heart. This study evaluated the gender and age-based differences in the average cardiothoracic ratios, and transverse cardiac diameters (TCD) of adults in Ghana.

Method

Plain chest radiography reports of 2004 patients (without known chest related diseases) generated by two radiologists with at least 15 years' experience from July 2016 to June 2020 were retrospectively analyzed for this study. The CTR for each radiograph was calculated using the formula CTR=(TCD÷TTD)×100, where TCD and TTD represent transverse cardiac diameters and transverse thoracic diameters, respectively. Data were analyzed with the statistical package for social sciences version 23. The independent t-test and One-way Analysis of Variance tests were used in the analyses.

Results

A total of 2004 patients' chest x-rays were used in the analyses. The ages of the patients ranged from 20–86 years old with a mean of 39.4±14.04 years. The mean CTR for males was 46.6 ± 3.7% while that of females was 47.7±3.7%. The difference in the overall CTR among the gender groupings was statistically significant (p = 0.001). There were statistically significant differences between the gender categories among patients in the following age groups: 30–39 (p=0.046), 40–49 (p=0.001), 50–59 (p=0.001) and 60–69 (p=0.001).

Conclusion

The study reveals there are significant gender and age-related differences in cardiac size parameters obtained from routine, frontal chest radiographs. These differences, if considered, may result in early and appropriate treatment of cardiac pathology in some age groups.

The routine pre-employment screening chest radiograph: Should it be routine?

BACKGROUND AND OBJECTIVE

A routine chest radiograph is mandatory in many institutions as a part of pre-employment screening. The usefulness of this has been studied over the years keeping in mind the added time, cost, and radiation concerns. Studies conducted outside India have shown different results, some for and some against it. To our knowledge, there is no published data from India on this issue.

MATERIALS AND METHODS

A retrospective review of the reports of 4113 pre-employment chest radiographs done between 2007 and 2009 was conducted.

RESULTS

Out of 4113 radiographs, 24 (0.58%) candidates required further evaluation based on findings from the screening chest radiograph. Out of these, 7 (0.17%) candidates required appropriate further treatment.

INTERPRETATION AND CONCLUSIONS

The percentage of significant abnormalities detected which needed further medical intervention was small (0.17%). Although the individual radiation exposure is very small, the large numbers done nation-wide would significantly add to the community radiation, with added significant cost and time implications. We believe that pre-employment chest radiographs should be restricted to candidates in whom there is relevant history and/or clinical findings suggestive of cardiopulmonary disease.

Follow-up chest radiographic findings in patients with MERS-CoV after recovery

PURPOSE

To evaluate the follow-up chest radiographic findings in patients with Middle East respiratory syndrome coronavirus (MERS-CoV) who were discharged from the hospital following improved clinical symptoms.

MATERIALS AND METHODS

Thirty-six consecutive patients (9 men, 27 women; age range 21-73 years, mean ± SD 42.5 ± 14.5 years) with confirmed MERS-CoV underwent follow-up chest radiographs after recovery from MERS-CoV. The 36 chest radiographs were obtained at 32 to 230 days with a median follow-up of 43 days. The reviewers systemically evaluated the follow-up chest radiographs from 36 patients for lung parenchymal, airway, pleural, hilar and mediastinal abnormalities. Lung parenchyma and airways were assessed for consolidation, ground-glass opacity (GGO), nodular opacity and reticular opacity (i.e., fibrosis). Follow-up chest radiographs were also evaluated for pleural thickening, pleural effusion, pneumothorax and lymphadenopathy. Patients were categorized into two groups: group 1 (no evidence of lung fibrosis) and group 2 (chest radiographic evidence of lung fibrosis) for comparative analysis. Patient demographics, length of ventilations days, number of intensive care unit (ICU) admission days, chest radiographic score, chest radiographic deterioration pattern (Types 1-4) and peak lactate dehydrogenase level were compared between the two groups using the student t-test, Mann-Whitney U test and Fisher's exact test.

RESULTS

Follow-up chest radiographs were normal in 23 out of 36 (64%) patients. Among the patients with abnormal chest radiographs (13/36, 36%), the following were found: lung fibrosis in 12 (33%) patients GGO in 2 (5.5%) patients, and pleural thickening in 2 (5.5%) patients. Patients with lung fibrosis had significantly greater number of ICU admission days (19 ± 8.7 days; P value = 0.001), older age (50.6 ± 12.6 years; P value = 0.02), higher chest radiographic scores [10 (0-15.3); P value = 0.04] and higher peak lactate dehydrogenase levels (315-370 U/L; P value = 0.001) when compared to patients without lung fibrosis.

CONCLUSION

Lung fibrosis may develop in a substantial number of patients who have recovered from Middle East respiratory syndrome coronavirus (MERS-CoV). Significantly greater number of ICU admission days, older age, higher chest radiographic scores, chest radiographic deterioration patterns and peak lactate dehydrogenase levels were noted in the patients with lung fibrosis on follow-up chest radiographs after recovery from MERS-CoV.

Systematic review and meta-analysis of chest radiograph (CXR) findings in COVID-19

Chest radiography (CXR) is most likely to be the utilized modality for diagnosing COVID-19 and following up on any lung-associated abnormalities. This review provides a meta-analysis of the current literature on CXR imaging findings to determine the most common appearances of lung abnormalities in COVID-19 patients in order to equip medical researchers and healthcare professionals in their efforts to combat this pandemic. Twelve studies met the inclusion criteria and were analyzed. The inclusion criteria consisted of: (1) published in English literature; (2) original research study; (3) sample size of at least 5 patients; (4) reporting clinical characteristics of COVID-19 patients as well as CXR imaging features; and (5) noting the number of patients with each corresponding imaging feature. A total of 1948 patients were included in this study. To perform the meta-analysis, a random-effects model calculated the pooled prevalence and 95% confidence intervals of abnormal CXR imaging findings. Seventy-four percent (74%) (95% CI: 51–92%) of patients with COVID-19 had an abnormal CXR at the initial time of diagnosis or sometime during the disease course. While there was no single feature on CXR that was diagnostic of COVID-19 viral pneumonia, a characteristic set of findings were obvious. The most common abnormalities were consolidation (28%, 95% CI: 8–54%) and ground-glass opacities (29%, 95% CI: 10–53%). The distribution was most frequently bilateral (43%, 95% CI: 27–60%), peripheral (51%, 95% CI: 36–66%), and basal zone (56%, 95% CI: 37–74%) predominant. Contrary to parenchymal abnormalities, pneumothorax (1%, 95% CI: 0–3%) and pleural effusions (6%, 95% CI: 1–16%) were rare.

Rare late-presentation congenital diaphragmatic hernia mimicking a tension pneumothorax

Congenital Diaphragmatic Hernia (CDH) is due to a defect in the diaphragm and is usually detected soon after birth. However, in rare cases, asymptomatic CDHs can be missed and present later in life. Late-presentation CDH can be misdiagnosed as tension pneumothorax leading to iatrogenic complications. We report a case of a 10-year-old boy who presented with non-specific symptoms of vomiting and occasional breathlessness, but was subsequently diagnosed as late-presentation CDH. This case highlights the role of imaging in the diagnosis and management of late-presenting CDH. The role of CT imaging as an invaluable tool to further evaluate equivocal radiographic findings in CDH is discussed.

Classification of cardioembolic stroke based on a deep neural network using chest radiographs

BACKGROUND

Although chest radiographs have not been utilised well for classifying stroke subtypes, they could provide a plethora of information on cardioembolic stroke. This study aimed to develop a deep convolutional neural network that could diagnose cardioembolic stroke based on chest radiographs.

METHODS

Overall, 4,064 chest radiographs of consecutive patients with acute ischaemic stroke were collected from a prospectively maintained stroke registry. Chest radiographs were randomly partitioned into training/validation (n = 3,255) and internal test (n = 809) datasets in an 8:2 ratio. A densely connected convolutional network (ASTRO-X) was trained to diagnose cardioembolic stroke based on chest radiographs. The performance of ASTRO-X was evaluated using the area under the receiver operating characteristic curve. Gradient-weighted class activation mapping was used to evaluate the region of focus of ASTRO-X. External testing was performed with 750 chest radiographs of patients with acute ischaemic stroke from 7 hospitals.

FINDINGS

The areas under the receiver operating characteristic curve of ASTRO-X were 0.86 (95% confidence interval [CI], 0.83-0.89) and 0.82 (95% CI, 0.79-0.85) during the internal and multicentre external testing, respectively. The gradient-weighted class activation map demonstrated that ASTRO-X was focused on the area where the left atrium was located. Compared with cases predicted as non-cardioembolism by ASTRO-X, cases predicted as cardioembolism by ASTRO-X had higher left atrial volume index and lower left ventricular ejection fraction in echocardiography.

INTERPRETATION

ASTRO-X, a deep neural network developed to diagnose cardioembolic stroke based on chest radiographs, demonstrated good classification performance and biological plausibility.

FUNDING

Grant No. 14-2020-046 and 08-2016-051 from the Seoul National University Bundang Research Fund and NRF-2020M3E5D9079768 from the National Research Foundation of Korea.

Lung cancer screening by single-shot dual-energy subtraction using flat-panel detector

Purpose

The purpose of this study was to evaluate the usefulness of single-shot dual-energy subtraction (DES) method using a flat-panel detector for lung cancer screening

Materials and methods

The subjects were 13,315 residents (5801 males and 7514 females) aged 50 years or older (50–97 years, with an intermediate value of 68 years) who underwent lung cancer screening for a period of 1 year and 6 months from January 2019 to June 2020. We investigated whether the number of lung cancers detected, the detection rate, and the rate of required scrutiny changed, when DES images were added to the judgment based on conventional chest radiography.

Results

When DES images were added, the number and percentage of cancer detection increased from 16 (0.12%) to 23 (0.17%) (P < 0.05). Five of the newly detected 7 lung cancers were in the early stages of resectable cancer. The rate of participants requiring scrutiny increased slightly from 1.1 to 1.3%.

Conclusion

DES method improved the detection of lung cancer in screening. The increase in the percentage of participants requiring scrutiny was negligible.

Radiology "forensics": determination of age and sex from chest radiographs using deep learning

PURPOSE

To develop and test the performance of deep convolutional neural networks (DCNNs) for automated classification of age and sex on chest radiographs (CXR).

METHODS

We obtained 112,120 frontal CXRs from the NIH ChestX-ray14 database performed in 48,780 females (44%) and 63,340 males (56%) ranging from 1 to 95 years old. The dataset was split into training (70%), validation (10%), and test (20%) datasets, and used to fine-tune ResNet-18 DCNNs pretrained on ImageNet for (1) determination of sex (using entire dataset and only pediatric CXRs); (2) determination of age < 18 years old or ≥ 18 years old (using entire dataset); and (3) determination of age < 11 years old or 11-18 years old (using only pediatric CXRs). External testing was performed on 662 CXRs from China. Area under the receiver operating characteristic curve (AUC) was used to evaluate DCNN test performance.

RESULTS

DCNNs trained to determine sex on the entire dataset and pediatric CXRs only had AUCs of 1.0 and 0.91, respectively (p < 0.0001). DCNNs trained to determine age < or ≥ 18 years old and < 11 vs. 11-18 years old had AUCs of 0.99 and 0.96 (p < 0.0001), respectively. External testing showed AUC of 0.98 for sex (p = 0.01) and 0.91 for determining age < or ≥ 18 years old (p < 0.001).

CONCLUSION

DCNNs can accurately predict sex from CXRs and distinguish between adult and pediatric patients in both American and Chinese populations. The ability to glean demographic information from CXRs may aid forensic investigations, as well as help identify novel anatomic landmarks for sex and age.

Deep learning for chest X-ray analysis: A survey

Recent advances in deep learning have led to a promising performance in many medical image analysis tasks. As the most commonly performed radiological exam, chest radiographs are a particularly important modality for which a variety of applications have been researched. The release of multiple, large, publicly available chest X-ray datasets in recent years has encouraged research interest and boosted the number of publications. In this paper, we review all studies using deep learning on chest radiographs published before March 2021, categorizing works by task: image-level prediction (classification and regression), segmentation, localization, image generation and domain adaptation. Detailed descriptions of all publicly available datasets are included and commercial systems in the field are described. A comprehensive discussion of the current state of the art is provided, including caveats on the use of public datasets, the requirements of clinically useful systems and gaps in the current literature.

Chin-on-Chest in Neck of Femur Fracture (COCNOF) sign: A simple radiographic predictor of frailty and mortality in hip fracture patients

INTRODUCTION

Fragility fractures are a significant public health challenge often occurring as a result of frailty. Identifying patients who have increased risk of adverse outcomes can aid treating teams in managing these patients appropriately. We hypothesise that the appearance of the patient's head overlapping the lung fields (named Chin on Chest in Neck of Femur sign (COCNOF)) in the admission chest radiograph was a predictor of increased mortality at 3, 6 and 12 months.

METHODS

All consecutive patients admitted with hip fracture between 1^st January - 31st December 2019 were analysed. We collected patient characteristics, AMTS score, ASA grade, length of stay, place of discharge, Nottingham Hip Fracture Score, Rockwood Frailty score, Charlson Comorbidity Index and presence of COCNOF sign. The main outcome measures were mortality at 90 days, six months and 12 months following admission.

RESULTS

469 patients with a mean age of 81.9 (SD 8.4) were included. 18% of patients were COCNOF positive. Univariate analysis showed positive COCNOF sign to be associated with higher mortality at 90 days (19.1 vs 10.8%; RR 1.95, 95%CI 1.05 - 3.63,p=0.03), six months (31.5% vs 14.2%; RR 2.77, 95%CI 1.62 - 4.72, p<0.001) and twelve months (41.6% vs 17.1%; RR 3.45, 95%CI 1.62-4.72, p<0.001). In the multivariate regression models the strongest predictors of mortality were age, gender and CCI it is therefore likely that the COCNOF sign is acting as a surrogate marker of these variables within the univariate models.

CONCLUSION

Our results suggest that COCNOF sign is a simple radiographic marker which can be used to identify patients with higher levels of frailty and increased risk of mortality following hip fracture.

Performance of SURF and SIFT Keypoints for the Automated Differentiation of Abnormality in Chest Radiographs

In this work, automated abnormality detection using keypoint information from Speeded-Up Robust feature (SURF) and Scale Invariant Feature Transform (SIFT) descriptors in chest Radiographic (CR) images is investigated and compared. Computerized image analysis using artificial intelligence is crucial to detect subtle and non-specific alterations of Tuberculosis (TB). For this, the healthy and TB CRs are subjected to lung field segmentation. SURF and SIFT keypoints are extracted from the segmented lung images. Statistical features from keypoints, its scale and orientation are computed. Discrimination of TB from healthy is performed using SVM. Results show that the SURF and SIFT methods are able to extract local keypoint information in CRs. Linear SVM is found to perform better with precision of 88.9% and AUC of 91% in TB detection for combined features. Hence, the application of keypoint techniques is found to have clinical relevance in the automated screening of non-specific TB abnormalities using CRs.

Risk factors for persistent abnormality on chest radiographs at 12-weeks post hospitalisation with PCR confirmed COVID-19

Background

The long-term consequences of COVID-19 remain unclear. There is concern a proportion of patients will progress to develop pulmonary fibrosis. We aimed to assess the temporal change in CXR infiltrates in a cohort of patients following hospitalisation for COVID-19.

Methods

We conducted a single-centre prospective cohort study of patients admitted to University Hospital Southampton with confirmed SARS-CoV2 infection between 20th March and 3rd June 2020. Patients were approached for standard-of-care follow-up 12-weeks after hospitalisation. Inpatient and follow-up CXRs were scored by the assessing clinician for extent of pulmonary infiltrates; 0–4 per lung (Nil = 0, < 25% = 1, 25–50% = 2, 51–75% = 3, > 75% = 4).

Results

101 patients with paired CXRs were included. Demographics: 53% male with a median (IQR) age 53.0 (45–63) years and length of stay 9 (5–17.5) days. The median CXR follow-up interval was 82 (77–86) days with median baseline and follow-up CXR scores of 4.0 (3–5) and 0.0 (0–1) respectively. 32% of patients had persistent CXR abnormality at 12-weeks. In multivariate analysis length of stay (LOS), smoking-status and obesity were identified as independent risk factors for persistent CXR abnormality. Serum LDH was significantly higher at baseline and at follow-up in patients with CXR abnormalities compared to those with resolution. A 5-point composite risk score (1-point each; LOS ≥ 15 days, Level 2/3 admission, LDH > 750 U/L, obesity and smoking-status) strongly predicted risk of persistent radiograph abnormality (0.81).

Conclusion

Persistent CXR abnormality 12-weeks post COVID-19 was common in this cohort. LOS, obesity, increased serum LDH, and smoking-status were risk factors for radiograph abnormality. These findings require further prospective validation.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12931-021-01750-8.

Chest X-ray in the emergency department during COVID-19 pandemic descending phase in Italy: correlation with patients' outcome

PURPOSE

The aims of our study are: (1) to estimate admission chest X-ray (CXR) accuracy during the descending phase of pandemic; (2) to identify specific CXR findings strictly associated with COVID-19 infection; and (3) to correlate lung involvement of admission CXR with patients' outcome.

MATERIALS AND METHODS

We prospectively evaluated the admission CXR of 327 patients accessed to our institute during the Italian pandemic descending phase (April 2020). For each CXR were searched ground glass opacification (GGO), consolidation (CO), reticular-nodular opacities (RNO), nodules, excavations, pneumothorax, pleural effusion, vascular congestion and cardiac enlargement. For lung alterations was defined the predominance (upper or basal, focal or diffuse, central or peripheric, etc.). Then radiologists assessed whether CXRs were suggestive or not for COVID-19 infection. For COVID-19 patients, a prognostic score was applied and correlated with the patients' outcome.

RESULTS

CXR showed 83% of specificity and 60% of sensitivity. GGO, CO, RNO and a peripheric, diffuse and basal prevalence showed good correlation with COVID-19 diagnosis. A logistic regression analysis pointed out GGO and a basal or diffuse distribution as independent predictors of COVID-19 diagnosis. The prognostic score showed good correlation with the patients' outcome.

CONCLUSION

In our study, admission CXR showed a fair specificity and a good correlation with patients' outcome. GGO and others CXR findings showed a good correlation with COVID-19 diagnosis; besides GGO a diffuse or bibasal distribution resulted in independent variables highly suggestive for COVID-19 infection thus enabling radiologists to signal to clinicians radiologically suspect patients during the pandemic descending phase.

Cost benefit analysis of portable chest radiography through glass: Initial experience at a tertiary care centre during COVID-19 pandemic

Introduction

Portable chest radiography through glass (TG-CXR) is a novel technique, particularly useful during the COVID-19 (Coronavirus disease 2019) pandemic. The purpose of this study was to understand the cost and benefit of adopting TG-CXR in quantifiable terms.

Methods

Portable or bedside radiographs are typically performed by a team of two technologists. The TG-CXR method has the benefit of allowing one technologist to stay outside of the patient room while operating the portable radiography machine, reducing PPE use, decreasing the frequency of radiography machine sanitization and decreasing technologists’ exposures to potentially infectious patients. The cost of implementing this technique during the current COVID-19 pandemic was obtained from our department's operational database. The direct cost of routinely used PPE and sanitization materials and the cost of the time taken by the technologists to clean the machine was used to form a quantitative picture of the benefit associated with TG-CXR technique.

Results

Technologists were trained on the TG-CXR method during a 15 min shift change briefing. This translated to a one-time cost of $424.88 USD. There was an average reduction of portable radiography machine downtime of 4 min and 48 s per study. The benefit of adopting the TG-CXR technique was $9.87 USD per patient imaged. This will result in a projected net cost savings of $51,451.84 USD per annum.

Conclusion

Adoption of the TG-CXR technique during the COVID-19 pandemic involved minimal one-time cost, but is projected to result in a net-benefit of over $51,000 USD per annum in our emergency department.

Unsupervised Deep Anomaly Detection in Chest Radiographs

The purposes of this study are to propose an unsupervised anomaly detection method based on a deep neural network (DNN) model, which requires only normal images for training, and to evaluate its performance with a large chest radiograph dataset. We used the auto-encoding generative adversarial network (α-GAN) framework, which is a combination of a GAN and a variational autoencoder, as a DNN model. A total of 29,684 frontal chest radiographs from the Radiological Society of North America Pneumonia Detection Challenge dataset were used for this study (16,880 male and 12,804 female patients; average age, 47.0 years). All these images were labeled as "Normal," "No Opacity/Not Normal," or "Opacity" by board-certified radiologists. About 70% (6,853/9,790) of the Normal images were randomly sampled as the training dataset, and the rest were randomly split into the validation and test datasets in a ratio of 1:2 (7,610 and 15,221). Our anomaly detection system could correctly visualize various lesions including a lung mass, cardiomegaly, pleural effusion, bilateral hilar lymphadenopathy, and even dextrocardia. Our system detected the abnormal images with an area under the receiver operating characteristic curve (AUROC) of 0.752. The AUROCs for the abnormal labels Opacity and No Opacity/Not Normal were 0.838 and 0.704, respectively. Our DNN-based unsupervised anomaly detection method could successfully detect various diseases or anomalies in chest radiographs by training with only the normal images.

Imaging algorithm for COVID-19: A practical approach

The global pandemic of COVID-19 pneumonia caused by the novel coronavirus (SARS-CoV-2) has strained healthcare resources across the world with emerging challenges of mass testing, resource allocation and management. While reverse transcriptase-polymerase chain reaction (RT-PCR) test is the most commonly utilized test and considered the current gold standard for diagnosis, the role of chest imaging has been highlighted by several studies demonstrating high sensitivity of computed tomography (CT). Many have suggested using CT chest as a first-line screening tool for the diagnosis of COVID-19. However, with advancement of laboratory testing and challenges in obtaining a CT scan without significant risk to healthcare providers, the role of imaging in diagnosis has been questioned. Several imaging societies have released consensus statements and guidelines on utilizing imaging resources and optimal reporting. In this review, we highlight the current evidence on various modalities in thoracic imaging for the diagnosis of COVID-19 and describe an algorithm on how to use these resources in an optimal fashion in accordance with the guidelines and statements released by major imaging societies.

Multi-Radiologist User Study for Artificial Intelligence-Guided Grading of COVID-19 Lung Disease Severity on Chest Radiographs

RATIONALE AND OBJECTIVES

Radiographic findings of COVID-19 pneumonia can be used for patient risk stratification; however, radiologist reporting of disease severity is inconsistent on chest radiographs (CXRs). We aimed to see if an artificial intelligence (AI) system could help improve radiologist interrater agreement.

MATERIALS AND METHODS

We performed a retrospective multi-radiologist user study to evaluate the impact of an AI system, the PXS score model, on the grading of categorical COVID-19 lung disease severity on 154 chest radiographs into four ordinal grades (normal/minimal, mild, moderate, and severe). Four radiologists (two thoracic and two emergency radiologists) independently interpreted 154 CXRs from 154 unique patients with COVID-19 hospitalized at a large academic center, before and after using the AI system (median washout time interval was 16 days). Three different thoracic radiologists assessed the same 154 CXRs using an updated version of the AI system trained on more imaging data. Radiologist interrater agreement was evaluated using Cohen and Fleiss kappa where appropriate. The lung disease severity categories were associated with clinical outcomes using a previously published outcomes dataset using Fisher's exact test and Chi-square test for trend.

RESULTS

Use of the AI system improved radiologist interrater agreement (Fleiss κ = 0.40 to 0.66, before and after use of the system). The Fleiss κ for three radiologists using the updated AI system was 0.74. Severity categories were significantly associated with subsequent intubation or death within 3 days.

CONCLUSION

An AI system used at the time of CXR study interpretation can improve the interrater agreement of radiologists.

Prognostic value of the chest X-ray in patients hospitalised for heart failure

Background

Patients admitted to hospital with heart failure will have had a chest X-ray (CXR), but little is known about their prognostic significance. We aimed to report the prevalence and prognostic value of the initial chest radiograph findings in patients admitted to hospital with heart failure (acute heart failure, AHF).

Methods

The erect CXRs of all patients admitted with AHF between October 2012 and November 2016 were reviewed for pulmonary venous congestion, Kerley B lines, pleural effusions and alveolar oedema. Film projection (whether anterior–posterior [AP] or posterior–anterior [PA]) and cardiothoracic ratio (CTR) were also recorded. Trial registration: ISRCTN96643197

Results

Of 1145 patients enrolled, 975 [median (interquartile range) age 77 (68–83) years, 61% with moderate, or worse, left ventricular systolic dysfunction, and median NT-proBNP 5047 (2337–10,945) ng/l] had an adequate initial radiograph, of which 691 (71%) were AP. The median CTR was 0.57 (IQR 0.53–0.61) in PA films and 0.60 (0.55–0.64) in AP films. Pulmonary venous congestion was present in 756 (78%) of films, Kerley B lines in 688 (71%), pleural effusions in 649 (67%) and alveolar oedema in 622 (64%).

A CXR score was constructed using the above features. Increasing score was associated with increasing age, urea, NT-proBNP, and decreasing systolic blood pressure, haemoglobin and albumin; and with all-cause mortality on multivariable analysis (hazard ratio 1.10, 95% confidence intervals 1.07–1.13, p < 0.001).

Conclusions

Radiographic evidence of congestion on a CXR is very common in patients with AHF and is associated with other clinical measures of worse prognosis.

Graphic abstract

Decoding the neonatal chest radiograph: An insight into neonatal respiratory distress

Respiratory distress is one of the leading causes of neonatal morbidity and mortality. Factors such as gestational age at birth, pulmonary maturity, and congenital factors are peculiar to this demographic. Clinical evaluation accompanied by chest radiography is the standard protocol for evaluating the underlying causative factors. Knowledge of the radiographic appearances of various pathologies and associations with certain congenital factors is quintessential for radiologists and primary neonatal care providers to steer the management in the right direction.