Pectoralis Muscle Area as a Predictor of Mortality in Patients Hospitalized with Bronchiectasis Exacerbation.

INTRODUCTION: Data on factors related to mortality in patients with bronchiectasis exacerbation are insufficient. Computed tomography (CT) can measure the pectoralis muscle area (PMA) and is a useful tool to diagnose sarcopenia. This study aimed to evaluate whether PMA can predict mortality in patients with bronchiectasis exacerbation. METHODS: Patients hospitalized due to bronchiectasis exacerbation at a single center were retrospectively divided into survivors and non-survivors based on 1-year mortality. Thereafter, a comparison of the clinical and radiologic characteristics was conducted between the two groups. RESULTS: A total of 66 (14%) patients died at 1 year. In the multivariate analysis, age, BMI <18.4 kg/m2, sex-specific PMA quartile, ≥3 exacerbations in the previous year, serum albumin <3.5 g/dL, cystic bronchiectasis, tuberculosis-destroyed lung, and diabetes mellitus were independent predictors for the 1-year mortality in patients hospitalized with bronchiectasis exacerbation. A lower PMA was associated with a lower overall survival rate in the survival analysis according to sex-specific quartiles of PMA. PMA had the highest area under the curve during assessment of prognostic performance in predicting the 1-year mortality. The lowest sex-specific PMA quartile group exhibited higher disease severity than the highest quartile group. CONCLUSIONS: CT-derived PMA was an independent predictor of 1-year mortality in patients hospitalized with bronchiectasis exacerbation. Patients with lower PMA exhibited higher disease severity. These findings suggest that PMA might be a useful marker for providing additional information regarding prognosis of patients with bronchiectasis exacerbation.

Development and Validation of a Robust and Interpretable Early Triaging Support System for Patients Hospitalized With COVID-19: Predictive Algorithm Modeling and Interpretation Study.

BACKGROUND: Robust and accurate prediction of severity for patients with COVID-19 is crucial for patient triaging decisions. Many proposed models were prone to either high bias risk or low-to-moderate discrimination. Some also suffered from a lack of clinical interpretability and were developed based on early pandemic period data. Hence, there has been a compelling need for advancements in prediction models for better clinical applicability. OBJECTIVE: The primary objective of this study was to develop and validate a machine learning-based Robust and Interpretable Early Triaging Support (RIETS) system that predicts severity progression (involving any of the following events: intensive care unit admission, in-hospital death, mechanical ventilation required, or extracorporeal membrane oxygenation required) within 15 days upon hospitalization based on routinely available clinical and laboratory biomarkers. METHODS: We included data from 5945 hospitalized patients with COVID-19 from 19 hospitals in South Korea collected between January 2020 and August 2022. For model development and external validation, the whole data set was partitioned into 2 independent cohorts by stratified random cluster sampling according to hospital type (general and tertiary care) and geographical location (metropolitan and nonmetropolitan). Machine learning models were trained and internally validated through a cross-validation technique on the development cohort. They were externally validated using a bootstrapped sampling technique on the external validation cohort. The best-performing model was selected primarily based on the area under the receiver operating characteristic curve (AUROC), and its robustness was evaluated using bias risk assessment. For model interpretability, we used Shapley and patient clustering methods. RESULTS: Our final model, RIETS, was developed based on a deep neural network of 11 clinical and laboratory biomarkers that are readily available within the first day of hospitalization. The features predictive of severity included lactate dehydrogenase, age, absolute lymphocyte count, dyspnea, respiratory rate, diabetes mellitus, c-reactive protein, absolute neutrophil count, platelet count, white blood cell count, and saturation of peripheral oxygen. RIETS demonstrated excellent discrimination (AUROC=0.937; 95% CI 0.935-0.938) with high calibration (integrated calibration index=0.041), satisfied all the criteria of low bias risk in a risk assessment tool, and provided detailed interpretations of model parameters and patient clusters. In addition, RIETS showed potential for transportability across variant periods with its sustainable prediction on Omicron cases (AUROC=0.903, 95% CI 0.897-0.910). CONCLUSIONS: RIETS was developed and validated to assist early triaging by promptly predicting the severity of hospitalized patients with COVID-19. Its high performance with low bias risk ensures considerably reliable prediction. The use of a nationwide multicenter cohort in the model development and validation implicates generalizability. The use of routinely collected features may enable wide adaptability. Interpretations of model parameters and patients can promote clinical applicability. Together, we anticipate that RIETS will facilitate the patient triaging workflow and efficient resource allocation when incorporated into a routine clinical practice.

Clinical Role of Upfront F-18 FDG PET/CT in Determining Biopsy Sites for Lung Cancer Diagnosis.

PURPOSE: This study aimed to investigate the impact of FDG PET/CT timing for biopsy site selection in patients with stage IV lung cancer regarding complications and diagnostic yield. METHODS: This retrospective analysis was performed on 1297 patients (924 men and 373 women with a mean age of 71.4 ± 10.2 years) who underwent percutaneous needle biopsy (PNB) for stage IV lung cancer diagnosis in two hospitals. Data collected included the patient's characteristics, order date of the biopsy and PET/CT exams, biopsy target site (lung or non-lung), guidance modality, complications, sample adequacy, and diagnostic success. Based on the order date of the PNB and PET/CT exams, patients were categorized into upfront and delayed PET/CT groups. RESULTS: PNB for non-lung targets resulted in significantly lower rates of minor (8.1% vs. 16.2%), major (0.2% vs. 3.4%), and overall complications (8.3% vs. 19.6%) compared to PNB for lung targets (p < 0.001 for all types of complications). Compared to the delayed PET/CT group, the upfront PET/CT group exhibited a lower probability of lung target selection of PNB (53.9% vs. 67.1%, p < 0.001), including a reduced incidence of major complications (1.0% vs. 2.9%, p = 0.031). Moreover, there was no significant difference in the occurrence of minor and total complications between the two groups. Upfront PET/CT and delayed PET/CT groups showed no significant difference regarding sample adequacy and diagnostic success. CONCLUSIONS: Upfront PET/CT may have an impact on the selection of the biopsy site for patients with advanced lung cancer, which could result in a lower rate of major complications with no change in the diagnostic yield. Upfront PET/CT demonstrates potential clinical implications for enhancing the safety of lung cancer diagnosis in clinical practice.

Predictive factors for severe radiation-induced lung injury in patients with lung cancer and coexisting interstitial lung disease.

BACKGROUND AND PURPOSE: This study aimed to investigate the predictive factors of severe radiation-induced lung injury (RILI) in patients with lung cancer and coexisting interstitial lung disease (ILD) undergoing conventionally fractionated thoracic radiotherapy. MATERIALS AND METHODS: The study includes consecutive patients treated with thoracic radiotherapy for lung cancer at two tertiary centers between 2010 and 2021. RILI severity was graded using the National Cancer Institute Common Terminology Criteria version 5.0, with severe RILI defined as toxicity grade ≥4, and symptomatic RILI as grade ≥2. The absolute neutrophil count (ANC), absolute lymphocyte count (ALC), and C-reactive protein were collected within 4 weeks before starting radiotherapy. Neutrophil-lymphocyte ratios (NLR) were calculated as ANC/ALC. The median follow-up was 9 (range, 6-114) months. RESULTS: Among 54 patients, 22 (40.7 %) had severe RILI. On multivariate logistic regression analysis, high pretreatment ANC (p = 0.030, OR = 4.313), pretreatment NLR (p = 0.007, OR = 5.784), and ILD severity (p = 0.027, OR = 2.416) were significant predictors of severe RILI. Dosimetric factors were not associated with severe RP. Overall survival was significantly worse for patients with severe RILI than those without, with 1-year cumulative overall survival rates of 7.4 % and 62.8 %, respectively. CONCLUSION: Pretreatment blood NLR, ANC, and ILD severity were associated with severe RILI. Overall survival was dismal for patients with severe RILI.

Immunoglobulin G4-Related Lung Disease with Waxing and Waning Pulmonary Infiltrates: A Case Report.

Immunoglobulin G4 (IgG4)-related lung disease can have various clinical courses. To our knowledge, reports of IgG4-related lung disease with waxing and waning pulmonary infiltrates only are very rare. A few lung nodules and ground glass opacities were incidentally found in a pre-operative evaluation in a 36-year-old female. The lung lesions showed waxing and waning in the follow-up chest CT. She underwent a surgical biopsy, and IgG4-related lung disease was confirmed.

Efficacy of Bronchial Artery Embolization for Clinically Suspected Bronchial Dieulafoy's Disease.

BACKGROUND: The efficacy of bronchial artery embolization (BAE) for bronchial Dieulafoy's disease (BDD) has not been well established. OBJECTIVE: This study aimed to evaluate the safety and efficacy of BAE in patients with clinically suspected BDD presenting with major hemoptysis, and to describe angiographic findings. METHODS: 17 patients (all men; mean age, 53.5 years) diagnosed with clinically suspected BDD by bronchoscopy (n = 7) or CT angiography (CTA) (n = 10) and who underwent BAE after directional and segmental localization of the target bronchus were enrolled. BAE was performed at the culprit bronchial artery traveling toward the target bronchus, regardless of the pathologic angiographic findings. Angiographic findings and clinical outcomes of BAE, including technical and clinical success, complication, recurrent hemoptysis, and follow-up imaging, were retrospectively reviewed. RESULTS: Representative angiographic findings included parenchymal hypervascularity prominent in the lobe where the BDD was located (82.4%), bronchial artery hypertrophy (70.6%), and contrast extravasation into the bleeding bronchus (17.6%). BAE was technically successful in all patients. All hemoptysis ceased within 24 h. No procedure-related complications occurred. During a mean follow-up of 491.9 days, 1 (6%) patient experienced recurrent hemoptysis. Follow-up bronchoscopy or CT performed in 10 (58.8%) patients showed the disappearance of pre-existing lesions (n = 9) or glue cast within the target bronchial artery (n = 1). CONCLUSION: Bronchial angiography showed pathologic findings in most patients with clinically suspected BDD. BAE assisted by bronchoscopy or CTA localization is a safe and effective treatment for patients with clinically suspected BDD with excellent short- to mid-term results.

Hemoptysis as the presenting manifestation of bronchiectasis-associated hospitalization in Korea.

Background: Patients with bronchiectasis commonly experience disease exacerbations, which cause significant morbidity and mortality. However, data regarding the clinical features of bronchiectasis patients hospitalized with hemoptysis are scarce. Methods: We retrospectively collected the data of patients with bronchiectasis-associated hospitalization at a tertiary referral center in Korea, and classified them into the hemoptysis and infective exacerbation (IE) groups. The presence of hemoptysis was defined as a volume of expectorated blood larger than 10 mL per 24 hours. The clinical, radiological, and laboratory parameters were compared between the two groups. Results: Patients were classified into the hemoptysis [267 (54.5%)] and IE [223 (45.5%)] groups. Among the 44 patients of the hemoptysis group, 37 (84.1%) presented with hemoptysis than with IE at the recurrent episode. The hemoptysis group had a significantly lower 30-day mortality than that of the IE group. Previous pulmonary tuberculosis (TB), mycetoma, and bronchial artery hypertrophy were independently associated with the hemoptysis group. In contrast, male sex, poor performance status, colonization of Pseudomonas aeruginosa, ≥3 involved lobes, cystic bronchiectasis, and emphysema were inversely associated with the hemoptysis group. The absence of hemoptysis was one of the independent predictors of 30-day mortality in patients with bronchiectasis-associated hospitalization. Conclusions: In Korea, bronchiectasis patients hospitalized with hemoptysis exhibit a distinct phenotype, and are more likely to have previous pulmonary TB, mycetoma, and bronchial artery hypertrophy. Hemoptysis is associated with a lower risk of short-term mortality compared to IE in bronchiectasis-associated hospitalization.

Correction: Deep Learning With Chest Radiographs for Making Prognoses in Patients With COVID-19: Retrospective Cohort Study.

[This corrects the article DOI: 10.2196/42717.].

Erratum: Correction of Affiliations in the Article "Establishment of a Nationwide Korean Imaging Cohort of Coronavirus Disease 2019".

This corrects the article on p. e413 in vol. 35, PMID: 33258333.

Nontuberculous Mycobacterial Infection Mimicking Lung Cancer in A Patient with Usual Interstitial Pneumonia Pattern Interstitial Lung Disease: A Case Report.

BACKGROUND: It has been reported that structure damage in the parenchymal lung disease such as idiopathic pulmonary fibrosis (IPF) is associated with high susceptibility to nontuberculous mycobacterial (NTM) infection. Radiologic features of NTM lung disease in destructive lung parenchyma can be atypical, which can cause confusion with other diseases including malignancy. Prompt and accurate identification of newly developed lesions in the follow-up computed tomography (CT) of IPF patients is challenging but crucial. CASE REPORT: We reported a case of an NTM infection in a patient with IPF, manifested as a mass-like consolidation with cavitation on chest CT, mimicking lung cancer. CONCLUSION: Being aware of the unusual radiologic features of NTM lung disease in IPF patients can be useful in the differential diagnosis of newly detected lesions.

Generative Adversarial Network-Based Image Conversion Among Different Computed Tomography Protocols and Vendors: Effects on Accuracy and Variability in Quantifying Regional Disease Patterns of Interstitial Lung Disease.

OBJECTIVE: To assess whether computed tomography (CT) conversion across different scan parameters and manufacturers using a routable generative adversarial network (RouteGAN) can improve the accuracy and variability in quantifying interstitial lung disease (ILD) using a deep learning-based automated software. MATERIALS AND METHODS: This study included patients with ILD who underwent thin-section CT. Unmatched CT images obtained using scanners from four manufacturers (vendors A-D), standard- or low-radiation doses, and sharp or medium kernels were classified into groups 1-7 according to acquisition conditions. CT images in groups 2-7 were converted into the target CT style (Group 1: vendor A, standard dose, and sharp kernel) using a RouteGAN. ILD was quantified on original and converted CT images using a deep learning-based software (Aview, Coreline Soft). The accuracy of quantification was analyzed using the dice similarity coefficient (DSC) and pixel-wise overlap accuracy metrics against manual quantification by a radiologist. Five radiologists evaluated quantification accuracy using a 10-point visual scoring system. RESULTS: Three hundred and fifty CT slices from 150 patients (mean age: 67.6 ± 10.7 years; 56 females) were included. The overlap accuracies for quantifying total abnormalities in groups 2-7 improved after CT conversion (original vs. converted: 0.63 vs. 0.68 for DSC, 0.66 vs. 0.70 for pixel-wise recall, and 0.68 vs. 0.73 for pixel-wise precision; P < 0.002 for all). The DSCs of fibrosis score, honeycombing, and reticulation significantly increased after CT conversion (0.32 vs. 0.64, 0.19 vs. 0.47, and 0.23 vs. 0.54, P < 0.002 for all), whereas those of ground-glass opacity, consolidation, and emphysema did not change significantly or decreased slightly. The radiologists' scores were significantly higher (P < 0.001) and less variable on converted CT. CONCLUSION: CT conversion using a RouteGAN can improve the accuracy and variability of CT images obtained using different scan parameters and manufacturers in deep learning-based quantification of ILD.

Clinical predictors and outcomes of non-expandable lung following percutaneous catheter drainage in lung cancer patients with malignant pleural effusion.

Non-expandable lung (NEL) often occurs during pleural fluid drainage in patients with malignant pleural effusion (MPE). However, data regarding the predictors and prognostic impact of NEL on primary lung cancer patients with MPE receiving pleural fluid drainage, compared to malignant pleural mesothelioma (MPM), are limited. This study was aimed to investigate the clinical characteristics of lung cancer patients with MPE developing NEL following ultrasonography (USG)-guided percutaneous catheter drainage (PCD) and compare the clinical outcomes between those with and without NEL. Clinical, laboratory, pleural fluid, and radiologic data and survival outcomes of lung cancer patients with MPE undergoing USG-guided PCD were retrospectively reviewed and compared between those with and without NEL. Among 121 primary lung cancer patients with MPE undergoing PCD, NEL occurred in 25 (21%). Higher pleural fluid lactate dehydrogenase (LDH) levels and presence of endobronchial lesions were associated with development of NEL. The median time to catheter removal was significantly extended in those with NEL compared to those without (P = .014). NEL was significantly associated with poor survival outcome in lung cancer patients with MPE undergoing PCD, along with poor Eastern Cooperative Oncology Group (ECOG) performance status (PS), the presence of distant metastasis, higher serum C-reactive protein (CRP) levels, and not receiving chemotherapy. NEL developed in one-fifth of lung cancer patients undergoing PCD for MPE and was associated with high pleural fluid LDH levels and the presence of endobronchial lesions. NEL may negatively affect overall survival in lung cancer patients with MPE receiving PCD.

Ultrasound-guided lung biopsy for small (≤2 cm) subpleural lung lesions: comparison of diagnostic yield and safety with larger lesions.

Background: Ultrasound (US)-guided percutaneous core needle biopsy (PCNB) has been used to diagnose subpleural lung lesions with high diagnostic performance and acceptable complication rates. However, with regard to the role of US-guided needle biopsy for the diagnosis of small (≤2 cm) subpleural lesions, limited information is available. Methods: From April 2011 to October 2021, a total of 572 US-guided PCNBs in 572 patients were retrospectively reviewed. The lesion size, pleural contact length (PCL), lesion location, and operator's experience were analyzed. Computed tomography features including peri-lesional emphysema, air-bronchogram, and cavitary change were also included in image analysis. The patients were divided into three groups according to lesion size (lesions ≤2 cm vs. 2 cm< lesions ≤5 cm vs. lesions >5 cm). The sample adequacy, diagnostic success rate, diagnostic accuracy, and complication rate was calculated. For statistical analysis, one-way ANOVA, Kruskal-Wallis test, or the chi-square test were used. Results: The overall sample adequacy, diagnostic success rate, and diagnostic accuracy were 96.2%, 82.9%, and 90.4%, respectively. In the subgroup analysis, sample adequacy (93.1% vs. 96.1% vs. 96.9%, P=0.307), diagnostic success rate (75.0% vs. 81.6% vs. 85.7%, P=0.079), and diagnostic accuracy (84.7% vs. 90.8% vs. 90.5%, P=0.301) were not significantly different. Operator's experience (OR, 0.64; 95% CI: 0.49-0.80; P<0.001), lesion size (OR, 0.68; 95% CI: 0.54-0.83; P<0.001), PCL (OR, 0.68; 95% CI: 0.52-0.84; P=0.001), and presence of air-bronchogram (OR, 14.36; 95% CI: 4.18-48.53; P<0.001) were independently associated with complication rate. Conclusions: US-guided PCNB performed by an experienced radiologist could be an effective and safe diagnostic approach for subpleural lesions, even in small lesions.

Atypical Radiological Findings of Capillary Hemangioma in Right Atrium: A Case Report.

BACKGROUND: Cardiac hemangiomas account for only 2.8% of primary cardiac tumors and are caused by the abnormal proliferation of endothelial cells and excess blood vessels. Typical radiological findings of cardiac hemangioma demonstrate intense contrast enhancement. CASE PRESENTATION: Here we report the case of a 69-year-old man who presented with a right atrial mass found incidentally with multimodal imaging findings, including echocardiography, computed tomography (CT), and magnetic resonance imaging (MRI). This case presented with poor enhancement within right atrial mass on dynamic contrast-enhanced CT scan and gadolinium-enhanced first-pass perfusion image of cardiac MRI. After surgical resection, the pathologic diagnosis of cardiac capillary hemangioma was made. CONCLUSION: Cardiac hemangioma could be included in the differential diagnosis for cardiac neoplasms demonstrating poor enhancements on CT and MRI scans.

Different characteristics of pleural abnormalities on computed tomography between tuberculous and malignant pleural effusions.

BACKGROUND: Computed tomography (CT) is the mainstay imaging modality for suspected pleural malignancy. Tuberculous pleural effusion (TPE) can present with various pleural abnormalities. However, few studies have evaluated the different characteristics of pleural abnormalities on chest CT between TPE and malignant pleural effusion (MPE). METHODS: Pleural abnormalities on contrast-enhanced CT in 277 and 289 patients with confirmed TPE and MPE diagnoses, respectively, were retrospectively assessed and compared between the two groups. Discriminating factors and diagnostic performance for MPE were evaluated using multivariate analysis and receiver operating characteristic curves. RESULTS: Focal pleural thickening was present in 44 (16%) cases of TPEs and 202 (70%) of MPEs. Further characterization of focal pleural thickening showed that MPEs had a significantly greater number, larger maximal thickness, and more nodular contour form, compared to TPEs. On the other hand, diffuse and circumferential pleural thickening were significantly more common in TPEs. In multivariate analysis, independent predictors for MPE included focally thickened pleurae ≥7, maximum thickness ≥6 mm, nodular contour pattern, and the absence of diffuse pleural thickening. Out of all the individual or combined predictors for MPE, the presence of any one of the three sub-parameters of focal pleural thickening provided the best diagnostic yield with 66% sensitivity and 92% specificity. CONCLUSION: Although focal pleural thickening in TPE mimics that in MPE, the features of MPE are significantly different from those of TPE in terms of size, number, and contour. These different characteristics may help differentiate MPE from TPE in patients with suspected MPE.

Peritumoral imaging features of thymic epithelial tumors for the prediction of transcapsular invasion: beyond intratumoral analysis

PURPOSE: The purpose of this study was to differentiate cases without transcapsular invasion (Masaoka-Koga stage I) from cases with transcapsular invasion (Masaoka-Koga stage II or higher) in patients with thymic epithelial tumors (TETs) using tumoral and peritumoral computed tomography (CT) features. METHODS: This retrospective study included 116 patients with pathological diagnoses of TETs. Two radiologists evaluated clinical variables and CT features, including size, shape, capsule integrity, presence of calcification, internal necrosis, heterogeneous enhancement, pleural effusion, pericardial effusion, and vascularity grade. Vascularity grade was defined as the extent of peritumoral vascular structures in the anterior mediastinum. The factors associated with transcapsular invasion were analyzed using multivariable logistic regression. In addition, the interobserver agreement for CT features was assessed using Cohen's or weighted kappa coefficients. The difference between the transcapsular invasion group and that without transcapsular invasion was evaluated statistically using the Student's t-test, Mann-Whitney U test, chi-square test, and Fisher's exact test. RESULTS: Based on pathology reports, 37 TET cases without and 79 with transcapsular invasion were identified. Lobular or irregular shape [odds ratio (OR): 4.19; 95% confidence interval (CI): 1.53-12.09; P = 0.006], partial complete capsule integrity (OR: 5.03; 95% CI: 1.85-15.13; P = 0.002), and vascularity grade 2 (OR: 10.09; 95% CI: 2.59-45.48; P = 0.001) were significantly associated with transcapsular invasion. The interobserver agreement for shape classification, capsule integrity, and vascularity grade was 0.840, 0.526, and 0.752, respectively (P < 0.001 for all). CONCLUSION: Shape, capsule integrity, and vascularity grade were independently associated with transcapsular invasion of TETs. Furthermore, three CT TET features demonstrated good reproducibility and help differentiate between TET cases with and without transcapsular invasion.

Clinical relevance of bronchiectasis in patients with community-acquired pneumonia.

BACKGROUND: Data regarding the clinical characteristics and treatment outcomes of patients with community-acquired pneumonia (CAP) and bronchiectasis (BE) are rare. This study aims to elucidate the clinical relevance of BE in patients with CAP. METHODS: Patients hospitalized with CAP in a single center were retrospectively analyzed and divided into significant BE (BE with ≥ 3 lobes or cystic BE on computed tomography) and control groups. Clinical and microbiological characteristics were compared between the two groups. RESULTS: In the final analysis, 2112 patients were included, and 104 (4.9%) had significant BE. The significant BE group exhibited a higher prevalence of sputum production, dyspnea, and complicated parapneumonic effusion or empyema than the control group. Pseudomonas aeruginosa was more frequently isolated in the significant BE group than in the control group, whereas Mycoplasma pneumoniae was less commonly identified. Length of hospital stay (LOS) was significantly longer in the significant BE group than the control group (12 [8-17] days vs. 9 [6-13] days, p < 0.001). In contrast, 30-day and in-hospital mortality rates did not significantly differ between the two groups. Furthermore, significant BE was an independent predictor of prolonged hospitalization in two models based on CURB-65 and pneumonia severity index. CONCLUSIONS: Significant BE occurred in approximately 5% of patients with CAP and was more likely to be associated with sputum, dyspnea, complicated parapneumonic effusion or empyema, and isolation of P. aeruginosa. Significant BE was an independent predictor of LOS in patients with CAP.

Deep Learning With Chest Radiographs for Making Prognoses in Patients With COVID-19: Retrospective Cohort Study.

BACKGROUND: An artificial intelligence (AI) model using chest radiography (CXR) may provide good performance in making prognoses for COVID-19. OBJECTIVE: We aimed to develop and validate a prediction model using CXR based on an AI model and clinical variables to predict clinical outcomes in patients with COVID-19. METHODS: This retrospective longitudinal study included patients hospitalized for COVID-19 at multiple COVID-19 medical centers between February 2020 and October 2020. Patients at Boramae Medical Center were randomly classified into training, validation, and internal testing sets (at a ratio of 8:1:1, respectively). An AI model using initial CXR images as input, a logistic regression model using clinical information, and a combined model using the output of the AI model (as CXR score) and clinical information were developed and trained to predict hospital length of stay (LOS) ≤2 weeks, need for oxygen supplementation, and acute respiratory distress syndrome (ARDS). The models were externally validated in the Korean Imaging Cohort of COVID-19 data set for discrimination and calibration. RESULTS: The AI model using CXR and the logistic regression model using clinical variables were suboptimal to predict hospital LOS ≤2 weeks or the need for oxygen supplementation but performed acceptably in the prediction of ARDS (AI model area under the curve [AUC] 0.782, 95% CI 0.720-0.845; logistic regression model AUC 0.878, 95% CI 0.838-0.919). The combined model performed better in predicting the need for oxygen supplementation (AUC 0.704, 95% CI 0.646-0.762) and ARDS (AUC 0.890, 95% CI 0.853-0.928) compared to the CXR score alone. Both the AI and combined models showed good calibration for predicting ARDS (P=.079 and P=.859). CONCLUSIONS: The combined prediction model, comprising the CXR score and clinical information, was externally validated as having acceptable performance in predicting severe illness and excellent performance in predicting ARDS in patients with COVID-19.

Unilateral Pulmonary Vein Atresia Initially Presenting as Interstitial Lung Disease in an Elderly Female: Serial Chest Radiograph Changes and Its Literature Review.

Unilateral pulmonary vein atresia (PVA) is a rare congenital cardiovascular anomaly occurring after the common pulmonary vein fails to incorporate into the left atrium. It is most commonly diagnosed in childhood, and diagnosis after reaching adulthood is extremely rare. Dyspnea on exertion and hemoptysis are common clinical features in adult PVA patients, whereas lung parenchymal abnormalities are indirect signs of PVA, which can manifest as interstitial lung disease. Herein, we present the case of a 62-year-old female diagnosed with unilateral PVA presenting as unilateral interstitial lung disease and report the changes in her chest radiographs over 12 years.