Use of artificial intelligence in triaging of chest radiographs to reduce radiologists' workload.

OBJECTIVES: To evaluate whether deep learning-based detection algorithms (DLD)-based triaging can reduce outpatient chest radiograph interpretation workload while maintaining noninferior sensitivity. METHODS: This retrospective study included patients who underwent initial chest radiography at the outpatient clinic between June 1 and June 30, 2017. Readers interpreted radiographs with/without a commercially available DLD that detects nine radiologic findings (atelectasis, calcification, cardiomegaly, consolidation, fibrosis, nodules, pneumothorax, pleural effusion, and pneumoperitoneum). The reading order was determined in a randomized, crossover manner. The radiographs were classified into negative and positive examinations. In a 50% worklist reduction scenario, radiographs were sorted in descending order of probability scores: the lower half was regarded as negative exams, while the remaining were read with DLD by radiologists. The primary analysis evaluated noninferiority in sensitivity between radiologists reading all radiographs and simulating a 50% worklist reduction, with the inferiority margin of 5%. The specificities were compared using McNemar's test. RESULTS: The study included 1964 patients (median age [interquartile range], 55 years [40-67 years]). The sensitivity was 82.6% (195 of 236; 95% CI: 77.5%, 87.3%) when readers interpreted all chest radiographs without DLD and 83.5% (197 of 236; 95% CI: 78.8%, 88.1%) in the 50% worklist reduction scenario. The difference in sensitivity was 0.8% (95% CI: - 3.8%, 5.5%), establishing noninferiority of 50% worklist reduction (p = 0.01). The specificity increased from 86.7% (1498 of 1728) to 90.4% (1562 of 1728) (p < 0.001) with DLD-based triage. CONCLUSION: Deep learning-based triaging may substantially reduce workload without lowering sensitivity while improving specificity. CLINICAL RELEVANCE STATEMENT: Substantial workload reduction without lowering sensitivity was feasible using deep learning-based triaging of outpatient chest radiograph; however, the legal responsibility for incorrect diagnoses based on AI-standalone interpretation remains an issue that should be defined before clinical implementation. KEY POINTS: • A 50% workload reduction simulation using deep learning-based detection algorithm maintained noninferior sensitivity while improving specificity. • The CT recommendation rate significantly decreased in the disease-negative patients, whereas it slightly increased in the disease-positive group without statistical significance. • In the exploratory analysis, the noninferiority of sensitivity was maintained until 70% of the workload was reduced; the difference in sensitivity was 0%.

Performance of AI to exclude normal chest radiographs to reduce radiologists' workload.

INTRODUCTION: This study investigates the performance of a commercially available artificial intelligence (AI) system to identify normal chest radiographs and its potential to reduce radiologist workload. METHODS: Retrospective analysis included consecutive chest radiographs from two medical centers between Oct 1, 2016 and Oct 14, 2016. Exclusions comprised follow-up exams within the inclusion period, bedside radiographs, incomplete images, imported radiographs, and pediatric radiographs. Three chest radiologists categorized findings into normal, clinically irrelevant, clinically relevant, urgent, and critical. A commercial AI system processed all radiographs, scoring 10 chest abnormalities on a 0-100 confidence scale. AI system performance was evaluated using the area under the ROC curve (AUC), assessing the detection of normal radiographs. Sensitivity was calculated for the default and a conservative operating point. the detection of negative predictive value (NPV) for urgent and critical findings, as well as the potential workload reduction, was calculated. RESULTS: A total of 2603 radiographs were acquired in 2141 unique patients. Post-exclusion, 1670 radiographs were analyzed. Categories included 479 normal, 332 clinically irrelevant, 339 clinically relevant, 501 urgent, and 19 critical findings. The AI system achieved an AUC of 0.92. Sensitivity for normal radiographs was 92% at default and 53% at the conservative operating point. At the conservative operating point, NPV was 98% for urgent and critical findings, and could result in a 15% workload reduction. CONCLUSION: A commercially available AI system effectively identifies normal chest radiographs and holds the potential to lessen radiologists' workload by omitting half of the normal exams from reporting. CLINICAL RELEVANCE STATEMENT: The AI system is able to detect half of all normal chest radiographs at a clinically acceptable operating point, thereby potentially reducing the workload for the radiologists by 15%. KEY POINTS: The AI system reached an AUC of 0.92 for the detection of normal chest radiographs. Fifty-three percent of normal chest radiographs were identified with a NPV of 98% for urgent findings. AI can reduce the workload of chest radiography reporting by 15%.

Learning from the machine: AI assistance is not an effective learning tool for resident education in chest x-ray interpretation.

OBJECTIVES: To assess whether a computer-aided detection (CADe) system could serve as a learning tool for radiology residents in chest X-ray (CXR) interpretation. METHODS: Eight radiology residents were asked to interpret 500 CXRs for the detection of five abnormalities, namely pneumothorax, pleural effusion, alveolar syndrome, lung nodule, and mediastinal mass. After interpreting 150 CXRs, the residents were divided into 2 groups of equivalent performance and experience. Subsequently, group 1 interpreted 200 CXRs from the "intervention dataset" using a CADe as a second reader, while group 2 served as a control by interpreting the same CXRs without the use of CADe. Finally, the 2 groups interpreted another 150 CXRs without the use of CADe. The sensitivity, specificity, and accuracy before, during, and after the intervention were compared. RESULTS: Before the intervention, the median individual sensitivity, specificity, and accuracy of the eight radiology residents were 43% (range: 35-57%), 90% (range: 82-96%), and 81% (range: 76-84%), respectively. With the use of CADe, residents from group 1 had a significantly higher overall sensitivity (53% [n = 431/816] vs 43% [n = 349/816], p < 0.001), specificity (94% [i = 3206/3428] vs 90% [n = 3127/3477], p < 0.001), and accuracy (86% [n = 3637/4244] vs 81% [n = 3476/4293], p < 0.001), compared to the control group. After the intervention, there were no significant differences between group 1 and group 2 regarding the overall sensitivity (44% [n = 309/696] vs 46% [n = 317/696], p = 0.666), specificity (90% [n = 2294/2541] vs 90% [n = 2285/2542], p = 0.642), or accuracy (80% [n = 2603/3237] vs 80% [n = 2602/3238], p = 0.955). CONCLUSIONS: Although it improves radiology residents' performances for interpreting CXRs, a CADe system alone did not appear to be an effective learning tool and should not replace teaching. CLINICAL RELEVANCE STATEMENT: Although the use of artificial intelligence improves radiology residents' performance in chest X-rays interpretation, artificial intelligence cannot be used alone as a learning tool and should not replace dedicated teaching. KEY POINTS: • With CADe as a second reader, residents had a significantly higher sensitivity (53% vs 43%, p < 0.001), specificity (94% vs 90%, p < 0.001), and accuracy (86% vs 81%, p < 0.001), compared to residents without CADe. • After removing access to the CADe system, residents' sensitivity (44% vs 46%, p = 0.666), specificity (90% vs 90%, p = 0.642), and accuracy (80% vs 80%, p = 0.955) returned to that of the level for the group without CADe.

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.

Artificial intelligence system for identification of false-negative interpretations in chest radiographs.

OBJECTIVES: To investigate the efficacy of an artificial intelligence (AI) system for the identification of false negatives in chest radiographs that were interpreted as normal by radiologists. METHODS: We consecutively collected chest radiographs that were read as normal during 1 month (March 2020) in a single institution. A commercialized AI system was retrospectively applied to these radiographs. Radiographs with abnormal AI results were then re-interpreted by the radiologist who initially read the radiograph ("AI as the advisor" scenario). The reference standards for the true presence of relevant abnormalities in radiographs were defined by majority voting of three thoracic radiologists. The efficacy of the AI system was evaluated by detection yield (proportion of true-positive identification among the entire examination) and false-referral rate (FRR, proportion of false-positive identification among all examinations). Decision curve analyses were performed to evaluate the net benefits of applying the AI system. RESULTS: A total of 4208 radiographs from 3778 patients (M:F = 1542:2236; median age, 56 years) were included. The AI system identified initially overlooked relevant abnormalities with a detection yield and an FRR of 2.4% and 14.0%, respectively. In the "AI as the advisor" scenario, radiologists detected initially overlooked relevant abnormalities with a detection yield and FRR of 1.2% and 0.97%, respectively. In a decision curve analysis, AI as an advisor scenario exhibited a positive net benefit when the cost-to-benefit ratio was below 1:0.8. CONCLUSION: An AI system could identify relevant abnormalities overlooked by radiologists and could enable radiologists to correct their false-negative interpretations by providing feedback to radiologists. KEY POINTS: • In consecutive chest radiographs with normal interpretations, an artificial intelligence system could identify relevant abnormalities that were initially overlooked by radiologists. • The artificial intelligence system could enable radiologists to correct their initial false-negative interpretations by providing feedback to radiologists when overlooked abnormalities were present.

Reproducibility of a protocol for standardized reading of chest X-rays of children household contact of patients with tuberculosis.

BACKGROUND: The interpretation of the chest radiograph may vary because it depends on the reader and due to the non-specificity of findings in tuberculosis (TB). We aim to assess the reproducibility of a standardized chest radiograph reading protocol in contacts of patients with pulmonary TB under the 5 years of age. METHODS: Descriptive, cross-sectional study with children under the age of five, household contacts of patients with confirmed pulmonary TB from Medellín, Bello and Itagüí (Colombia) between Jan-01-2015 and May-31-2016. Standardized reading protocol: two radiologists, blinded independent reading, use of template (Dr. Andronikou design) in case of disagreement a third reading was performed. Kappa coefficient for intra and inter observer agreement, and prevalence ratio were estimated of sociodemographic characteristics, TB exposure and interpretation of chest X-ray. RESULTS: From 278 children, standardized reading found 255 (91.7%) normal X-rays, 10 (3.6%) consistent with TB, and 13 (4.7%) other alterations. Global agreement was 91.3% (Kappa = 0.51). Inter-observer agreement between readers 1-2 was 90.0% (Kappa = 0.59) and 1-3 93.2% (Kappa = 0.59). Intra-observer agreement for reader 1 was 95.5% (Kappa = 0.86), 2 84.0% (Kappa = 0.51), and 3 94.7% (Kappa = 0.68). Greater inter-observer disagreement was between readers 1-2 for soft tissue density suggestive of adenopathy (4.6%), airspace opacification (1.17%) and pleural effusion (0.58%); between readers 1-3 for soft tissue density suggestive of adenopathy (4.2%), opacification of airspace (2.5%) and cavities (0.8%). CONCLUSIONS: Chest radiographs are an affordable tool that contributes to the diagnosis of TB, so having a standardized reading protocol showed good agreement and improves the reproducibility of radiograph interpretation.

Dose reduction and image quality improvement of chest radiography by using bone-suppression technique and low tube voltage: a phantom study.

OBJECTIVE: To clarify the relationship between entrance surface dose (ESD) and physical image quality of original and bone-suppressed chest radiographs acquired using high and low tube voltages. METHODS: An anthropomorphic chest phantom and a 12-mm diameter spherical simulated nodule with a CT value of approximately + 100 HU were used. The lung field in the chest radiograph was divided into seven areas, and the nodule was set in a total of 66 positions. A total of 264 chest radiographs were acquired using four ESD conditions: approximately 0.3 mGy at 140 and 70 kVp and approximately 0.2 and 0.1 mGy at 70 kVp. The radiographs were processed to produce bone-suppressed images. Differences in contrast and contrast-to-noise ratio (CNR) values of the nodule between each condition and between the original and bone-suppressed images were analyzed by a two-sided Wilcoxon signed-rank test. RESULTS: In the areas not overlapping with the ribs, both contrast and CNR values were significantly increased with the bone-suppression technique (p < 0.01). In the bone-suppressed images, these values of the three conditions at 70 kVp were equal to or significantly higher than those of the condition at 140 kVp. There was no apparent decrease in these values between the ESD of approximately 0.3 and 0.1 mGy at 70 kVp. CONCLUSION: By using the shortest exposure time and the lowest tube voltage possible not to increase in blurring artifact and image noise, it is possible to improve the image quality of bone-suppressed images and reduce the patient dose. KEY POINTS: • The effectiveness of bone-suppression techniques differs in areas of lung field. • Image quality of bone-suppressed chest radiographs is improved by lower tube voltage. • Applying lower tube voltage to bone-suppressed chest radiographs leads to dose reduction.

Establishment of a Nationwide Korean Imaging Cohort of Coronavirus Disease 2019.

BACKGROUND: The Korean Society of Thoracic Radiology (KSTR) recently constructed a nation-wide coronavirus disease 2019 (COVID-19) database and imaging repository, referred to the Korean imaging cohort of COVID-19 (KICC-19) based on the collaborative efforts of its members. The purpose of this study was to provide a summary of the clinico-epidemiological data and imaging data of the KICC-19. METHODS: The KSTR members at 17 COVID-19 referral centers retrospectively collected imaging data and clinical information of consecutive patients with reverse transcription polymerase chain reaction-proven COVID-19 in respiratory specimens from February 2020 through May 2020 who underwent diagnostic chest computed tomography (CT) or radiograph in each participating hospital. RESULTS: The cohort consisted of 239 men and 283 women (mean age, 52.3 years; age range, 11-97 years). Of the 522 subjects, 201 (38.5%) had an underlying disease. The most common symptoms were fever (n = 292) and cough (n = 245). The 151 patients (28.9%) had lymphocytopenia, 86 had (16.5%) thrombocytopenia, and 227 patients (43.5%) had an elevated CRP at admission. The 121 (23.4%) needed nasal oxygen therapy or mechanical ventilation (n = 38; 7.3%), and 49 patients (9.4%) were admitted to an intensive care unit. Although most patients had cured, 21 patients (4.0%) died. The 465 (89.1%) subjects underwent a low to standard-dose chest CT scan at least once during hospitalization, resulting in a total of 658 CT scans. The 497 subjects (95.2%) underwent chest radiography at least once during hospitalization, which resulted in a total of 1,475 chest radiographs. CONCLUSION: The KICC-19 was successfully established and comprised of 658 CT scans and 1,475 chest radiographs of 522 hospitalized Korean COVID-19 patients. The KICC-19 will provide a more comprehensive understanding of the clinical, epidemiological, and radiologic characteristics of patients with COVID-19.

Visual assessment of calcification in solitary pulmonary nodules on chest radiography: correlation with volumetric quantification of calcification.

PURPOSE: To assess the ability of digital chest radiography (CXR) to reveal calcification in solitary pulmonary nodules (SPNs), and to examine the correlation between a visual assessment and volumetric quantification of the calcification. MATERIALS AND METHODS: This study was a retrospective review of 220 SPNs identified by both CXR and chest CT. Eleven observers did blind review of the CXR images and scored nodule calcification on a confidence scale of 1 to 5. The area under the receiver operating characteristics (ROC) curve (AUC) was obtained to analyze the diagnostic performance. The intraclass correlation coefficient (ICC) for interrater reliability was calculated. The AUC and ICC were calculated according to the following nodule diameter groups: group 1 (< 10 mm), group 2 (≥ 10 mm and < 20 mm), and group 3 (≥ 20 mm). RESULTS: Of the 220 SPNs, 145 SPNs (65.6%) were identified as non-calcified and 75 (34.4%) as calcified. The average percentage of calcification volume in SPN > 160 HU (Vol160HU) among the 75 calcified nodules was 47.5%. The mean Vol160HU of the 68 SPNs classified as having definite calcification was 51.1%. The overall AUC was 0.71. The AUCs for groups 1, 2, and 3 was 0.835, 0.639, and 0.620, respectively. The ICCs for groups 1, 2, 3 was 0.65, 0.48, and 0.33, respectively. CONCLUSION: The overall diagnostic performance of digital CXR to predict calcification in SPNs was moderately accurate and the diagnostic performance for predicting calcification in SPNs was significantly higher, and interobserver reproducibility was good when SPN < 10 mm compared with ≥ 10 mm in diameter. KEY POINTS: • The misdiagnosis of a non-calcified nodule as a calcified one by CXR could lead to poor management choices for the SPN. • The diagnostic performance of CXR in predicting calcification was best for nodules < 10 mm in diameter. SPNs with calcification of approximately 50% of their volume tend to be considered calcified. • The diagnostic performance of CXR in identifying calcification was low for nodules ≥ 10 mm in diameter; therefore, we should carefully evaluate calcification carefully for nodules ≥ 10 mm.

[The correlation between gastric bubble size and laryngopharyngeal reflux pattern in patients with laryngopharyngeal reflux disease].

Objective: To investigate the correlation between gastric bubble size and laryngopharyngeal reflux pattern in patients with laryngopharyngeal reflux disease(LPRD). Methods: A total of 52 LPRD patients who underwent Dx-pH monitoring and anteroposterior chest radiography at the same time from February 2016 to November 2018 were retrospectively studied. Patients were devided into three position-related groups according to the Ryan score of upright and supine: isolated upright reflux(IUR), isolated supine reflux(ISR) and bipositional reflux(BR) groups. In addition, 13 healthy volunteers with negative pH monitoring were selected as the control group. Gastric bubble size and pH monitoring data among the four groups were compared. SPSS 24.0 was used for statistical analysis. Results: In all the 52 patients, 35 cases (67.3%) were classified as IUR, 9 cases (17.3%) as ISR, and 8 cases (15.4%) as BR. The height of gastric bubbles in the four groups were: IUR (26±14) mm, ISR (9±8) mm, BR (20±13) mm, control (17±15) mm, and statistical difference was found among the four groups(P=0.004). Post Hoc Multiple Comparisons found that IUR group had statistical difference between ISR group and control group (P=0.001, P=0.034 respectively). There was no statistical difference of gastric bubble width and area among the four groups(P=0.340, P=0.186 respectively). The ROC curve of the gastric bubble height with isolated upright and supine reflux patterns was obtained, and the optimal cutoff value of the gastric bubble height was 11 mm. Accordingly, we divided the patiens into two groups with high and low gastric bubble. LPRD reflux pattern distribution was significantly different between the two groups(P<0.001). The comparison of reflux parameters in pH monitoring also showed that the supine reflux parameters in the lower group were significantly higher than those in the higher group, and the upright reflux parameters in the higher group were significantly higher than those in the lower group(P<0.001). Conclusions: The height of gastric bubble is significantly correlated with the reflux patterns in LPRD patients. The gastric bubble of patients with IUR is significantly higher than that of patients with ISR. Taking 11 mm as the cutoff value, patients with higher gastric bubble are more prone to upright laryngopharyngeal reflux, while those with lower gastric bubble are more prone to supine laryngopharyngeal reflux.

[Application of chest CT scan in gestational trophoblastic neoplasia with lung metastasis].

Objective: To explore the role of CT scan for the diagnosis of lung metastasis in stage Ⅲ gestational trophoblastic neoplasia (GTN) . Methods: To figure out the role of CT scan for lung metastasis in GTN initial diagnosis, treatment and follow-up, 93 GTN patients with lung metastasis from January, 2015 to December, 2016 were retrospectively analyzed in Obstetrics and Gynecology Hospital of Fudan University. Results: (1) Among 93 GTN patients with lung metastasis, 70 patients with the International Federation of Gynecology and Obstetrics (FIGO) score ≤6 were defined as low risk GTN and 23 patients score score ≥7 were defined as high risk GTN. Forty nine patients had negative chest X-ray findings and 39 cases with pulmonary lesions were identified both by chest X-ray compared to CT scan. Five cases were excluded due to no consensus could make for the results of chest X-ray. The true positive rate of chest X-ray for lung metastasis were 41% (29/70) in low risk GTN and 43% (10/23) in high risk GTN patients without statistical difference (χ(2)=0.090, P=0.925) . For those patients with positive chest CT scan and negative chest X-ray finding, pulmonary lesions in 32 (65%, 32/49) cases were blocked by heart, chest wall or diaphragm in chest X-ray. Seventeen (35%,17/49) patients with lung lesions less than 5 mm had negative chest X-ray results due to the lower sensitivity compared to CT scan. (2) In 88 patients with stage Ⅲ, 78 patients had successful initial treatment, but 4 of them were recurrence in twelve months follow-up. Ten patients were chemotherapy resistance for the initial treatment. The initial chemotherapy remission rate in low risk GTN patients was higher than that in high risk ones (χ(2)=4.911, P=0.027) . In 49 cases with negative chest X-ray, there was no correlation with the rate of remission,chemotherapy resistance and recurrence in stage Ⅲ patients (P>0.05) . (3) For those patients who had poorly response to initial chemotherapy, the diameters of lesions in lung were unchanged or increased during the treatment, form (5.1±4.1) mm to (7.4±2.8) mm. The pulmonary lesions were continuously shrunk from (7.8±5.3) mm to (4.7±4.4) mm for those patients with complete and partial remission including the recurrent GTN patients (Z=-2.713, P=0.007) . Conclusions: Patients with GTN in stage Ⅲ have down staging if only use chest X-ray for imaging at the initial diagnosis. Chest CT scan is recommended for primary imaging evaluation of FIGO staging in qualified medical organization. For those patients with persistent abnormal serum hCG level and negative chest X-ray, chest CT scan is strongly recommended to identify the persist or resistant lung lesions and follow up.

[Results of conservative treatment in patients with occult pneumothorax]. / Resultados del tratamiento conservador en pacientes con neumotórax oculto.

INTRODUCTION: An occult pneumothorax is found in 2-15% trauma patients. Observation (without tube thoracostomy) in these patients presents still some controversies in the clinical practice. The objective of the study is to evaluate the efficacy and the adverse effects when observation is performed. METHODS: A retrospective observational study was undertaken in our center (university hospital level II). Data was obtained from a database with prospective registration. A total of 1087 trauma patients admitted in the intensive care unit from 2006 to 2013 were included. RESULTS: In this period, 126 patients with occult pneumothorax were identified, 73 patients (58%) underwent immediate tube thoracostomy and 53 patients (42%) were observed. Nine patients (12%) failed observation and required tube thoracostomy for pneumothorax progression or hemothorax. No patient developed a tension pneumothorax or experienced another adverse event related to the absence of tube thoracostomy. Of the observed patients 16 were under positive pressure ventilation, in this group 3 patients (19%) failed observation. There were no differences in mortality, hospital length of stay or intensive care length of stay between the observed and non-observed group. CONCLUSION: Observation is a safe treatment in occult pneumothorax, even in pressure positive ventilated patients.

[Update on the radiological study of pulmonary tuberculosis]. / Puesta al día en el estudio radiológico de la tuberculosis pulmonary.

Tuberculosis has made a comeback in recent years. This upsurge has been attributed to factors such as increased immigration and the human immunodeficiency virus epidemic. Primary pulmonary tuberculosis manifests radiologically with parenchymal involvement, lymph node involvement, pleural effusion, and/or miliary disease. In post-primary tuberculosis, the earliest radiological sign is small nodules and branching centrilobular lesions that increase in size and coalesce to form ill-defined patchy consolidations; cavitations are very characteristic of active disease. The aim of this article is to describe the radiologic findings for pulmonary tuberculosis and its complications.

[Lung cancer screening]. / Cribado del cáncer de pulmón.

Lung cancer is a very important disease, curable in early stages. There have been trials trying to show the utility of chest x-ray or computed tomography in Lung Cancer Screening for decades. In 2011, National Lung Screening Trial results were published, showing a 20% reduction in lung cancer mortality in patients with low dose computed tomography screened for three years. These results are very promising and several scientific societies have included lung cancer screening in their guidelines. Nevertheless we have to be aware of lung cancer screening risks, such as: overdiagnosis, radiation and false positive results. Moreover, there are many issues to be solved, including choosing the appropriate group to be screened, the duration of the screening program, intervals between screening and its cost-effectiveness. Ongoing trials will probably answer some of these questions. This article reviews the current evidence on lung cancer screening.

Is a Camera-Type Portable X-Ray Device Clinically Feasible in Chest Imaging?: Image Quality Comparison with Chest Radiographs Taken with Traditional Mobile Digital X-Ray Devices.

Purpose: To evaluate whether the image quality of chest radiographs obtained using a camera-type portable X-ray device is appropriate for clinical practice by comparing them with traditional mobile digital X-ray devices. Materials and Methods: Eighty-six patients who visited our emergency department and underwent endotracheal intubation, central venous catheterization, or nasogastric tube insertion were included in the study. Two radiologists scored images captured with traditional mobile devices before insertion and those captured with camera-type devices after insertion. Identification of the inserted instruments was evaluated on a 5-point scale, and the overall image quality was evaluated on a total of 20 points scale. Results: The identification score of the instruments was 4.67 ± 0.71. The overall image quality score was 19.70 ± 0.72 and 15.02 ± 3.31 (p < 0.001) for the mobile and camera-type devices, respectively. The scores of the camera-type device were significantly lower than those of the mobile device in terms of the detailed items of respiratory motion artifacts, trachea and bronchus, pulmonary vessels, posterior cardiac blood vessels, thoracic intervertebral disc space, subdiaphragmatic vessels, and diaphragm (p = 0.013 for the item of diaphragm, p < 0.001 for the other detailed items). Conclusion: Although caution is required for general diagnostic purposes as image quality degrades, a camera-type device can be used to evaluate the inserted instruments in chest radiographs.

Enhancing diagnostic deep learning via self-supervised pretraining on large-scale, unlabeled non-medical images.

BACKGROUND: Pretraining labeled datasets, like ImageNet, have become a technical standard in advanced medical image analysis. However, the emergence of self-supervised learning (SSL), which leverages unlabeled data to learn robust features, presents an opportunity to bypass the intensive labeling process. In this study, we explored if SSL for pretraining on non-medical images can be applied to chest radiographs and how it compares to supervised pretraining on non-medical images and on medical images. METHODS: We utilized a vision transformer and initialized its weights based on the following: (i) SSL pretraining on non-medical images (DINOv2), (ii) supervised learning (SL) pretraining on non-medical images (ImageNet dataset), and (iii) SL pretraining on chest radiographs from the MIMIC-CXR database, the largest labeled public dataset of chest radiographs to date. We tested our approach on over 800,000 chest radiographs from 6 large global datasets, diagnosing more than 20 different imaging findings. Performance was quantified using the area under the receiver operating characteristic curve and evaluated for statistical significance using bootstrapping. RESULTS: SSL pretraining on non-medical images not only outperformed ImageNet-based pretraining (p < 0.001 for all datasets) but, in certain cases, also exceeded SL on the MIMIC-CXR dataset. Our findings suggest that selecting the right pretraining strategy, especially with SSL, can be pivotal for improving diagnostic accuracy of artificial intelligence in medical imaging. CONCLUSIONS: By demonstrating the promise of SSL in chest radiograph analysis, we underline a transformative shift towards more efficient and accurate AI models in medical imaging. RELEVANCE STATEMENT: Self-supervised learning highlights a paradigm shift towards the enhancement of AI-driven accuracy and efficiency in medical imaging. Given its promise, the broader application of self-supervised learning in medical imaging calls for deeper exploration, particularly in contexts where comprehensive annotated datasets are limited.

A study on changes in lung function, neutralizing antibodies, and symptoms of adult patients hospitalized with COVID-19.

BACKGROUND/AIMS: To identify changes in symptoms and pulmonary sequelae in patients with coronavirus disease 2019 (COVID-19). METHODS: Patients with COVID-19 hospitalized at seven university hospitals in Korea between February 2020 and February 2021 were enrolled, provided they had ≥ 1 outpatient follow-up visit. Between January 11 and March 9, 2021 (study period), residual symptom investigations, chest computed tomography (CT) scans, pulmonary function tests (PFT), and neutralizing antibody tests (NAb) were performed at the outpatient visit (cross-sectional design). Additionally, data from patients who already had follow-up outpatient visits before the study period were collected retrospectively. RESULTS: Investigation of residual symptoms, chest CT scans, PFT, and NAb were performed in 84, 35, 31, and 27 patients, respectively. After 6 months, chest discomfort and dyspnea persisted in 26.7% (4/15) and 33.3% (5/15) patients, respectively, and 40.0% (6/15) and 26.7% (4/15) patients experienced financial loss and emotional distress, respectively. When the ratio of later CT score to previous ones was calculated for each patient between three different time intervals (1-14, 15-60, and 61-365 days), the median values were 0.65 (the second interval to the first), 0.39 (the third to the second), and 0.20 (the third to the first), indicating that CT score decreases with time. In the high-severity group, the ratio was lower than in the low-severity group. CONCLUSION: In COVID-19 survivors, chest CT score recovers over time, but recovery is slower in severely ill patients. Subjects complained of various ongoing symptoms and socioeconomic problems for several months after recovery.

Dynamic chest radiography: a state-of-the-art review.

Dynamic chest radiography (DCR) is a real-time sequential high-resolution digital X-ray imaging system of the thorax in motion over the respiratory cycle, utilising pulsed image exposure and a larger field of view than fluoroscopy coupled with a low radiation dose, where post-acquisition image processing by computer algorithm automatically characterises the motion of thoracic structures. We conducted a systematic review of the literature and found 29 relevant publications describing its use in humans including the assessment of diaphragm and chest wall motion, measurement of pulmonary ventilation and perfusion, and the assessment of airway narrowing. Work is ongoing in several other areas including assessment of diaphragmatic paralysis. We assess the findings, methodology and limitations of DCR, and we discuss the current and future roles of this promising medical imaging technology.Critical relevance statement Dynamic chest radiography provides a wealth of clinical information, but further research is required to identify its clinical niche.

Imaging of bronchopulmonary dysplasia.

Bronchopulmonary dysplasia (BPD) is a multifactorial disease with many associated co-morbidities, responsible for most cases of chronic lung disease in childhood. The use of imaging exams is pivotal for the clinical care of BPD and the identification of candidates for experimental therapies and a closer follow-up. Imaging is also useful to improve communication with the family and objectively evaluate the clinical evolution of the patient's disease. BPD imaging has been classically performed using only chest X-rays, but several modern techniques are currently available, such as lung ultrasound, thoracic tomography, magnetic resonance imaging and electrical impedance tomography. These techniques are more accurate and provide clinically meaningful information. We reviewed the most recent evidence published in the last five years regarding these techniques and analyzed their advantages and disadvantages.

CT Findings of Thymic Epidermoid Cyst in the Anterior Mediastinum: A Case Report and Literature Review.

An epidermoid cyst is a benign tumor found anywhere in the body. However, the occurrence of epidermoid cysts in the thymus is extremely rare, with only six cases reported worldwide. The correct diagnosis of thymic epidermoid cysts is often difficult due to the unusual location and nonspecific imaging findings. Herein, we present a case of a thymic epidermoid cyst in a 37-year-old female with clinical information and chest CT findings. Further, we have reviewed previous literature reports describing imaging findings of thymic epidermoid cysts.