Search

Objective: To investigate the prognostic value of chest computed tomography (CT), platelet count (PLT), serum C-reactive protein (CRP) level, and oxygenation index (OI) in patients with severe community-acquired pneumonia (CAP). Methods: We conducted a retrospective analysis of clinical data collected from 226 patients with CAP who received treatment in Huzhou Central Hospital from February 2022 to November 2023. Patients were divided into two groups based on pneumonia severity: Severe group (patients with severe CAP, n=113) and Typical group (patients with typical pneumonia, n=113). Differences in CT score, PLT, CRP, and OI levels between the two groups were analyzed, as well as the prognostic value of the combined CT score, PLT, CRP, and OI levels in severe CAP. Results: The CT Score and CRP level in the Severe group were significantly higher than those in the Typical group, whereas PLT and OI were significantly lower (P<0.05). Of 113 patients with severe pneumonia, 42 died and 71 survived. The CT Score and CRP level in the death group were significantly higher, whereas PLT and OI were lower compared to the survival group (P<0.05). The area under the ROC curve of the combined CT Score, PLT, CRP, OI for the prediction of death in patients with severe CAP was 0.970, sensitivity was 85.7, and specificity was 93.0, which was higher than that of each index alone. Conclusions: The combined chest CT, PLT, CRP, and OI have high prognostic value for severe CAP. [ABSTRACT FROM AUTHOR]

Objectives: To compare and evaluate the usefulness of magnetic resonance imaging (MRI) with computed tomography (CT) in bronchiectasis; to compare MRI and CT scores with pulmonary function tests (PFT) and to evaluate the role of Diffusion-weighted imaging (DWI) in bronchiectasis. Methods: In this prospective study, 25 patients between 7–21 y of age with a clinical/radiological diagnosis of bronchiectasis underwent MDCT and MRI chest. MRI and CT scoring was performed using modified Bhalla-Helbich's score by two independent radiologists for all parameters. A final consensus score was recorded. The overall image quality of different MRI sequences to identify pathologies was also assessed. Appropriate statistical tests were used for inter-observer agreements, and correlation amongst CT and MRI; as well as CT, MRI and PFT. Results: Strong agreement (ICC 0.80–0.95) between CT and MRI was seen for extent and severity of bronchiectasis, number of bullae, sacculation/abscess, emphysema, collapse/ consolidation, mucus plugging, and mosaic perfusion. Overall CT and MRI scores had perfect concordance (ICC 0.978). Statistically significant (p-value <0.01) intra-observer and inter-observer agreement for all CT and MRI score parameters were seen. A strong negative correlation was seen between total CT and MRI severity scores and forced expiratory volume in 1 s (FEV1), forced vital capacity (FVC), forced expiratory flow (FEF) 25–75%. DWI MR, with an apparent diffusion coefficient (ADC) cut-off of 1.62 × 10–3 mm3/s had a sensitivity of 70% and specificity of 75% in detecting true mucus plugs. Conclusions: MRI with DWI can be considered as a radiation-free alternative in the diagnostic algorithm for assessment of lung changes in bronchiectasis, especially in follow-up. [ABSTRACT FROM AUTHOR]

Background: The prevalence of lung cancer among individuals afflicted with interstitial pneumonia (IP) stands at approximately 20%. The early detection of lung cancer via chest computed tomography (CT) surveillance proves challenging in IP patients. Our investigation sought to identify a potential biomarker capable of providing early indications of the presence of lung tumors in such patients. Materials and methods: We examined the attributes of serum tumor markers, imaging characteristics, and histological findings in individuals diagnosed with IP, both with and without concurrent lung cancer. Results: 106 patients diagnosed with IP were included in the study, comprising 36 individuals with concurrent lung cancer and 70 patients solely diagnosed with IP. Serum concentrations of CEA and CA12-5 were notably elevated in IP patients with lung cancer, compared to those with IP alone. Logistic regression analyses revealed that, in comparison to IP patients within the first quartile of CEA levels, the relative risk of developing lung cancer associated with IP escalated by 4.0-fold, 3.1-fold, 11.0-fold, and 13.3-fold in the second, third, fourth, and fifth quartiles, respectively. Upon controlling for gender and age, statistical significance in risk was observed solely for the fourth and fifth quartiles. Receiver operating characteristic (ROC) curve analysis conducted in patients diagnosed with ILD-CA identified a CEA cutoff point of 6.9 ng/mL, demonstrating sensitivities of 61.1% and specificities of 78.5%. The area under the curve was calculated as 0.7(95% CI: 0.63–0.81). Conclusion: The serum levels of CEA were notably elevated in IP patients with concurrent lung cancer in contrast to those who were just suffering from IP. The heightened serum CEA levels correlate with an escalated risk of cancer occurrence among IP patients, suggesting that serum CEA levels could potentially serve as an indicative marker for the presence of cancer in IP patients. [ABSTRACT FROM AUTHOR]

Introduction: A large number of middle-aged and elderly patients have an insufficient understanding of osteoporosis and its harm. This study aimed to establish and validate a convolutional neural network (CNN) model based on unenhanced chest computed tomography (CT) images of the vertebral body and skeletal muscle for opportunistic screening in patients with osteoporosis. Materials and methods: Our team retrospectively collected clinical information from participants who underwent unenhanced chest CT and dual-energy X-ray absorptiometry (DXA) examinations between January 1, 2022, and December 31, 2022, at four hospitals. These participants were divided into a training set (n = 581), an external test set 1 (n = 229), an external test set 2 (n = 198) and an external test set 3 (n = 118). Five CNN models were constructed based on chest CT images to screen patients with osteoporosis and compared with the SMI model to predict the performance of osteoporosis patients. Results: All CNN models have good performance in predicting osteoporosis patients. The average F1 score of Densenet121 in the three external test sets was 0.865. The area under the curve (AUC) of Desenet121 in external test set 1, external test set 2, and external test set 3 were 0.827, 0.859, and 0.865, respectively. Furthermore, the Densenet121 model demonstrated a notably superior performance compared to the SMI model in predicting osteoporosis patients. Conclusions: The CNN model based on unenhanced chest CT vertebral and skeletal muscle images can opportunistically screen patients with osteoporosis. Clinicians can use the CNN model to intervene in patients with osteoporosis and promptly avoid fragility fractures. Critical relevance statement: The CNN model based on unenhanced chest CT vertebral and skeletal muscle images can opportunistically screen patients with osteoporosis. Clinicians can use the CNN model to intervene in patients with osteoporosis and promptly avoid fragility fractures. Key Points: The application of unenhanced chest CT is increasing. Most people do not consciously use DXA to screen themselves for osteoporosis. A deep learning model was constructed based on CT images from four institutions. [ABSTRACT FROM AUTHOR]

Background: While chest X-ray (CXR) is an efficient tool for expeditious detection of life-threatening injury, chest computed tomography (CCT) is more sensitive albeit with added time, cost, and radiation. Thus far, there is limited evidence and lack of consensus on the best imaging practices. We sought to determine the association between imaging modality and outcomes in isolated blunt thoracic trauma. Methods: The 2017-2020 TQIP database was queried for adult patients who sustained isolated blunt chest trauma and underwent chest imaging within 24 hours of admission. Patients who underwent CCT were 2:1 propensity-score-matched to those who underwent CXR. The primary outcome was mortality, and the secondary outcomes were hospital and ICU length of stay (LOS), ICU admission, need for and days requiring mechanical ventilation, complications, and discharge location. Results: Propensity score matching yielded 17 716 patients with CCT and 8861 with CXR. While bivariate analysis showed lower 24-hr (CCT.2% vs CXR.4%, P =.0015) and in-hospital mortality (CCT 1.2% vs CXR 1.5%, P =.0454) in the CCT group, there was no difference in survival probability between groups (P =.1045). A higher percentage of CCT patients were admitted to the ICU (CCT 26.9% vs CXR 21.9%, P <.0001) and discharged to rehab (CCT.8% vs CXR.5%, P =.0178). Discussion: CT offers no survival benefit over CXR in isolated blunt thoracic trauma. While CCT should be considered if clinically unclear, CXR likely suffices as an initial screening tool. These findings facilitate optimal resource allocation in constrained environments. [ABSTRACT FROM AUTHOR]

Abstract Background Lung cancer is a leading cause of cancer-related mortality. Non-small cell lung cancer (NSCLC) comprises 85% of cases with rising incidence among never-smokers (NS). This study seeks to compare clinical, imaging, pathology, and outcomes between NS and ever-smokers (S) NSCLC patients to identify significant differences if any. Methods Retrospective cohort study of 155 NSCLC patients (88 S and 67 NS). The main predictor was smoking. Clinical, imaging, and pathology findings were evaluated at initial biopsy for staging. The primary outcome was all-cause mortality, and the secondary outcome was 12-month progression-free survival. Results Imaging: NS and S had similar nodule size (0.81), calcification (> 0.99), and invasion of adjacent structures (> 0.99) (p values). NS slightly trended to more commonly involve the RLL vs S the RUL (p = 0.11). NS had higher numbers of extrathoracic metastases at initial biopsy for staging (p = 0.055). Pathology: NS more commonly had adenocarcinoma compared to S, who had equal numbers of adenocarcinoma and squamous cell carcinoma (p = 0.001). Rates of lymphovascular and pleural invasion were similar (p = 0.84 and 0.28). Initial staging: NS were more often initially diagnosed with stage IV disease (p = 0.046), positive nodal disease (p = 0.002), and metastatic disease (p = 0.004). Outcomes: S had a non-significant trend toward worse 12-month progression-free survival (rate ratio = 1.31, p = 0.31; HR = 1.33, p = 0.28). NS and S had similar 1-year all-cause mortality (HR = 1.06, p = 0.90). S had nearly double the risk of all-cause mortality in 5 years (HR = 1.73, p = 0.056) and 10 years (HR = 1.77, p = 0.02). Median survival was 6.6 years for NS and 3.9 years for S, with NS surviving 2.7 years longer on average (p = 0.045). Conclusions CT nodule features were similar in NS and S. NS more often had metastatic adenopathy, distant metastases, and stage IV disease at initial biopsy. Despite similar 12-month progression-free survival and 1-year all-cause mortality, S had nearly double the risk of mortality in the first 5 and 10 years post-diagnosis. Trial registration Retrospectively registered.

Abstract Distinguishing between primary adenocarcinoma (AC) and squamous cell carcinoma (SCC) within non-small cell lung cancer (NSCLC) tumours holds significant management implications. We assessed the performance of radiomics-based models in distinguishing primary there is from SCC presenting as lung nodules on Computed Tomography (CT) scans. We studied individuals with histopathologically proven adenocarcinoma or SCC type NSCLC tumours, detected as lung nodules on Chest CT. The workflow comprised manual nodule segmentation, regions of interest creation, preprocessing data, feature extraction, and nodule classification using machine learning algorithms. The dataset comprised 46 adenocarcinoma and 28 SCC cases. For feature extraction, 101 radiomic features were extracted from the tumour regions using the ‘pyradiomics’ module in Python. After extensive experimentation with various feature importance techniques, the top 10 most significant radiomic features for differentiating between adenocarcinoma and squamous cell carcinoma (SCC) were identified. The Synthetic Minority Over-Sampling Technique was used to achieve a balanced distribution. Lung nodules were classified using 13 machine-learning algorithms, including Linear Discriminant Analysis, Random Forest, AdaBoost, and eXtreme Gradient Boosting. The Multilayer Perceptron (MLP) Classifier with Rectified Linear Unit (ReLu) activation was the most accurate (83% accuracy) with 83% precision and 86% sensitivity in distinguishing SCC from adenocarcinoma. It achieved a balanced F1 score of 83%, indicating well-rounded performance in both precision and sensitivity. The average Area Under the Curve score was 88%, representing good discrimination between the two classes of lung nodules. Radiomics is a powerful non-invasive tool that could potentially add to visual information obtained on CT. The MLP Classifier with ReLu activation showed good accuracy in distinguishing primary lung adenocarcinoma from SCC nodules. However, widespread multicentre trials are required to realize the full potential of radiomics in lung nodules.

Preserved ratio impaired spirometry (PRISm) is a common pulmonary functional impairment, considered a pre-chronic obstructive pulmonary disease stage and has received increased attention from the academic community in recent years. Despite a comprehensive review of the etiology, epidemiology, and risk factors of the PRISm population, there is a lack of systematic review for chest CT imaging. To gain a more comprehensive understanding of this population, this article summarizes the chest CT imaging features of the PRISm population using both visual and quantitative assessment methods, including the characteristic changes of airways, lung parenchyma, and vessels. The article indicates that the reference value of visual assessment of chest CT results for the PRISm population is limited, while quantitative assessment of chest CT, combined with pulmonary function testing, is advantageous in gaining a deeper understanding of the lung structural features of the PRISm population. Future studies are expected to employ a more systematic approach through prospective, large-sample, multi-center cohort studies to elucidate the characteristics of the PRISm population in chest imaging.

Abstract Purpose Opportunistic osteoporosis screening, conducted during routine medical examinations such as chest computed tomography (CT), presents a potential solution for early detection. This study aims to investigate the feasibility of utilizing radiomics technology based on chest CT images to screen for opportunistic osteoporosis. Methods This Study is a Multicenter Retrospective Investigation. Relevant clinical data, including demographics and DXA results, would be collected for each participant. The radiomics analysis in this study focuses on the extraction of features from the 11th or 12th thoracic vertebral bodies from chest CT images. SVM machine learning models would be trained using these radiomic features, with DXA results as the ground truth for osteoporosis classification. Results In the training group, Clinical models had an accuracy of 0.684 and an AUC of 0.744, Radiomics models had an accuracy of 0.828 and an AUC of 0.896, Nomogram models had an accuracy of 0.839 and an AUC of 0.901. In the internal validation group, Clinical models had an accuracy of 0.769 and an AUC of 0.829, Radiomics models had an accuracy of 0.832 and an AUC of 0.892, Nomogram models had an accuracy of 0.839 and an AUC of 0.918. In the external validation group, Clinical models had an accuracy of 0.715 and an AUC of 0.741, Radiomics models had an accuracy of 0.777 and an AUC of 0.796, Nomogram models had an accuracy of 0.785 and an AUC of 0.807. In all three datasets, the Nomogram model exhibited a statistically significant difference in screening effectiveness compared to the clinical models. Conclusion Our research demonstrates that by leveraging radiomics features extracted from a single thoracic spine using chest CT, and incorporating these features with patient basic information, opportunistic screening for osteoporosis can be achieved.

Abstract We aimed to implement a fully automatic computed tomography (CT) image-detection programming algorithm as a pectus excavatum (PE) diagnostic tool, facilitating comprehensive chest wall deformity evaluation. We developed our algorithm using MATLAB, leveraging the Hounsfield unit threshold and region growing methods. The MATLAB graphical user interface enables the direct use of our program. We validated the model using CT images of anthropomorphic phantoms and one normal individual. The measurement values obtained by our algorithm demonstrated very small differences compared to the known anthropomorphic phantom model data and manual measurement. For algorithm testing, 17,214 chest CT scans obtained from 57 PE patients were processed by the algorithm and independently reviewed by a radiologist and a thoracic surgeon. The measurements of transverse, anteroposterior, and sternum-to-vertebral distance of the thoracic cavity, along with the calculated data of four indices, exhibited high positive correlations (0.94–0.99). The asymmetry index and maximum anteroposterior hemithorax distance exhibited moderate correlation (0.40–0.83). Our automatic PE diagnostic tool demonstrated high accuracy; four chest wall deformity indices were obtained simultaneously without any initial manual marking, correlating well with manual measurements.

Distinguishing between primary adenocarcinoma (AC) and squamous cell carcinoma (SCC) within non-small cell lung cancer (NSCLC) tumours holds significant management implications. We assessed the performance of radiomics-based models in distinguishing primary there is from SCC presenting as lung nodules on Computed Tomography (CT) scans. We studied individuals with histopathologically proven adenocarcinoma or SCC type NSCLC tumours, detected as lung nodules on Chest CT. The workflow comprised manual nodule segmentation, regions of interest creation, preprocessing data, feature extraction, and nodule classification using machine learning algorithms. The dataset comprised 46 adenocarcinoma and 28 SCC cases. For feature extraction, 101 radiomic features were extracted from the tumour regions using the 'pyradiomics' module in Python. After extensive experimentation with various feature importance techniques, the top 10 most significant radiomic features for differentiating between adenocarcinoma and squamous cell carcinoma (SCC) were identified. The Synthetic Minority Over-Sampling Technique was used to achieve a balanced distribution. Lung nodules were classified using 13 machine-learning algorithms, including Linear Discriminant Analysis, Random Forest, AdaBoost, and eXtreme Gradient Boosting. The Multilayer Perceptron (MLP) Classifier with Rectified Linear Unit (ReLu) activation was the most accurate (83% accuracy) with 83% precision and 86% sensitivity in distinguishing SCC from adenocarcinoma. It achieved a balanced F1 score of 83%, indicating well-rounded performance in both precision and sensitivity. The average Area Under the Curve score was 88%, representing good discrimination between the two classes of lung nodules. Radiomics is a powerful non-invasive tool that could potentially add to visual information obtained on CT. The MLP Classifier with ReLu activation showed good accuracy in distinguishing primary lung adenocarcinoma from SCC nodules. However, widespread multicentre trials are required to realize the full potential of radiomics in lung nodules. [ABSTRACT FROM AUTHOR]

Background: To validate the automated quantification of cardiac chamber volumes and myocardial mass on non-contrast chest CT using cardiac MR (CMR) as a reference. Methods: We retrospectively included 53 consecutive patients who received non-contrast chest CT and CMR within three weeks. A deep learning model created cardiac segmentations on axial soft-tissue reconstructions from CT, covering all four cardiac chambers and the left ventricular myocardium. Segmentations on CMR cine short-axis and long-axis images served as a reference. Standard estimates of diagnostic accuracy were calculated for ventricular volumes at end-diastole and end-systole (LVEDV, LVESV, RVEDV, RVESV), left ventricular mass (LVM), and atrial volumes (LA, RA) at ventricular end-diastole. A qualitative assessment noted segmentation issues. Results: The deep learning model generated CT measurements for 52 of the 53 patients (98%). Based on CMR measurements, the average LVEDV was 166 ± 64 mL, RVEDV was 144 ± 51 mL, and LVM was 115 ± 39 g. The CT measurements correlated well with CMR measurements for LVEDV, LVESV, and LVM (ICC = 0.85, ICC = 0.84, and ICC = 0.91; all p < 0.001) and RVEDV and RVESV (ICC = 0.79 and ICC= 0.78; both p < 0.001), and moderately well with LA and RA (ICC = 0.74 and ICC = 0.61; both p < 0.001). Absolute agreements likewise favored LVEDV, LVM, and RVEDV. ECG-gating did not relevantly influence the results. The CT results correctly identified 7/15 LV and 1/1 RV as dilated (one and six false positives, respectively). Major qualitative issues were found in three cases (6%). Conclusions: Automated cardiac chamber volume and myocardial mass quantification on non-contrast chest CT produced viable measurements in this retrospective sample. Relevance Statement: An automated cardiac assessment on non-contrast chest CT provides quantitative morphological data on the heart, enabling a preliminary organ evaluation that aids in incidentally identifying at-risk patients who may benefit from a more targeted diagnostic workup. [ABSTRACT FROM AUTHOR]

Some of the viruses may cause lung parenchyma and airway involvement. Usually, viral pneumonia causes ground-glass opacities, bilateral peripheral distribution, consolidation, vascular thickening, and reticular opacity. These features are common in COVID-19 rather than Non-Covid-19 viral pneumonia. However, in advanced cases, COVID-19 viral pneumonia may cause organising pneumonia and fibrosis of the lung. Atypical findings of Non-Covid-19 pneumonia have included central peripheral distribution, pleural effusion, lymphadenopathy, nodules, tree-in-bud opacities, and pneumothorax. Therefore, differentiating Non-Covid-19 pneumonia from COVID-19 pneumonia at chest computed tomography (CT) is necessary. In that case, CT scans of the thorax are one of the essential tools for early identification and future prognosis of viral pneumonia. We have proposed a Computer-Aided Diagnostic (CAD) system that can detect features of chest CT using a Deep Neural Network (DNN) with Long Short-Term Memory (LSTM). Transfer learning using pre-trained DNN models (ResNet50, VGG19, InceptionV3, Xception, DenseNet121, and VGG16) is applied to retain both high-level and low-level features effectively. The deep features are passed to the LSTM layer. The LSTM is utilised as a classifier and detects long short-term dependencies. The proposed method employs a hybrid DNN-LSTM network for automatic detection to take advantage of the uniqueness of the two models. The proposed models are trained with common and different features present in the chest CT of COVID-19 and Non-Covid-19 viral pneumonia. The 5-fold cross-validation (CV) method validated and tested the proposed model. The proposed DNN model's performance is quite improved with LSTM and CV. As a result, the proposed LM-DNN (VGG16+LSTM+CV) model has achieved the classification test accuracy of 91.58% and specificity of 93.86%, which offers superior performance with state-of-the-art. Also, the DenseNet121+LSTM+CV model has reached the classification test accuracy of 90.1% and sensitivity of 92%. [ABSTRACT FROM AUTHOR]

Background and Objectives: Severe and critical COVID-19 pneumonia can lead to long-term complications, especially affecting pulmonary function and immune health. However, the extent and progression of these complications over time are not well understood. This study aimed to assess lung function, radiological changes, and some immune parameters in survivors of severe and critical COVID-19 up to 12 months after hospital discharge. Materials and Methods: This prospective observational cohort study followed 85 adult patients who were hospitalized with severe or critical COVID-19 pneumonia at a tertiary care hospital in Vilnius, Lithuania, for 12 months post-discharge. Pulmonary function tests (PFTs), including forced vital capacity (FVC), forced expiratory volume in 1 s (FEV1), and diffusion capacity for carbon monoxide (DLCO), were conducted at 3, 6, and 12 months. High-resolution chest computed tomography (CT) scans assessed residual inflammatory and profibrotic/fibrotic abnormalities. Lymphocyte subpopulations were evaluated via flow cytometry during follow-up visits to monitor immune status. Results: The median age of the cohort was 59 years (IQR: 51–64). Fifty-three (62.4%) patients had critical COVID-19 disease. Pulmonary function improved significantly over time, with increases in FVC, FEV1, VC, TLC, and DLCO. Residual volume (RV) did not change significantly over time, suggesting that some aspects of lung function, such as air trapping, remained stable and may require attention in follow-up care. The percentage of patients with restrictive spirometry patterns decreased from 24.71% at 3 months to 14.8% at 12 months (p < 0.05). Residual inflammatory changes on CT were present in 77.63% at 6 months, decreasing to 69.62% at 12 months (p < 0.001). Profibrotic changes remained prevalent, affecting 82.89% of patients at 6 months and 73.08% at 12 months. Lymphocyte counts declined significantly from 3 to 12 months (2077 cells/µL vs. 1845 cells/µL, p = 0.034), with notable reductions in CD3+ (p = 0.040), CD8+ (p = 0.007), and activated CD3HLA-DR+ cells (p < 0.001). This study found that higher CD4+ T cell counts were associated with worse lung function, particularly reduced total lung capacity (TLC), while higher CD8+ T cell levels were linked to improved pulmonary outcomes, such as increased forced vital capacity (FVC) and vital capacity (VC). Multivariable regression analyses revealed that increased levels of CD4+/CD28+/CD192+ T cells were associated with worsening lung function, while higher CD8+/CD28+/CD192+ T cell counts were linked to better pulmonary outcomes, indicating that immune dysregulation plays a critical role in long-term respiratory recovery. Conclusions: Survivors of severe and critical COVID-19 pneumonia continue to experience significant long-term impairments in lung function and immune system health. Regular monitoring of pulmonary function, radiological changes, and immune parameters is essential for guiding personalized post-COVID-19 care and improving long-term outcomes. Further research is needed to explore the mechanisms behind these complications and to develop targeted interventions for long COVID-19. [ABSTRACT FROM AUTHOR]

Purpose: To describe the first experience of patient dose optimization in developing AQD, SSDE and image quality-related DRLs for common CT examinations in the adult age group using the concept of AQD. Materials and methods: The recent published IQSC from 0 to 4 were used by radiologists for the assessment of image quality. The entire data were collected for five types (brain CT, chest CT, chest HRCT, abdomen KUB CT and abdomen + pelvic CT) CT investigations based on anatomic region (head, chest and abdomen + pelvic). The entire datasets of 264 patients were categorized into three groups based on their weights: group-1 (41–60 kg), group-2 (61–80 kg) and group-3 (81–100 kg). Only score-3 images were considered to assess median and 75th percentile values of CTDIvol and DLP to obtain AQDs and DRLs, respectively. Results: Following the practical training of four radiologists on image quality scoring criteria for CT images, 264 (92%) out of 288 patient images were clinically acceptable as per IQSC for the study. The AQD (median) values in terms of CTDIvol for the mentioned weight groups were 25.8, 2.7, and 30.6 mGy, while the median DLP values for these groups were 496, 510 and 557 mGycm, respectively, for brain CT. The 75th percentile values in terms of CTDIvol were 30.2, 35.3 and 36.2 mGy, while in terms of DLP, they were 583, 619 and 781 mGycm for brain CT, respectively. Similar results are presented for the above-mentioned procedures, as well as in terms of SSDE. Conclusion: The first ever experience in obtaining AQD, SSDE and DRLs values for specific CT procedures couples image quality with dose indices, showing comparable values with other relevant studies. Hence, it will provide a baseline for comparison within the facility for future studies and facilitate dose optimization for other facilities aiming for improvement. [ABSTRACT FROM AUTHOR]

Simple Summary: Lung nodules detected by computed tomography (CT) often require invasive procedures for definitive diagnosis. With the increasing use of CT, incidental lung nodules have increased significantly. An adjunctive blood-based biomarker test that predicts lung cancer risk could reduce unnecessary interventions and focus diagnostic efforts on high-risk patients. This study introduces a blood-based biomarker test that predicts lung cancer risk in CT-detected nodules with a sensitivity of 78.4% (95% CI: 75.7–81.1) and a specificity of 93.1% (95% CI: 90.0–96.3). Background and Objectives: Lung nodules detected by chest computed tomography (CT) often require invasive biopsies for definitive diagnosis, leading to unnecessary procedures for benign lesions. A blood-based biomarker test that predicts lung cancer risk in CT-detected nodules could help stratify patients and direct invasive diagnostics toward high-risk individuals. Methods: In this multicenter, single-blinded clinical trial, we evaluated a test measuring plasma levels of p53, anti-p53 autoantibodies, CYFRA 21-1, and anti-CYFRA 21-1 autoantibodies in patients with CT-detected lung nodules. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated, and subgroup analyses by gender, age, and smoking status were performed. A total of 1132 patients who had CT-detected lung nodules, including 885 lung cancer cases and 247 benign lesions, were enrolled from two academic hospitals in South Korea. Results: The test demonstrated a sensitivity of 78.4% (95% CI: 75.7–81.1) and specificity of 93.1% (95% CI: 90.0–96.3) in predicting lung cancer in CT-detected nodules. The PPV was 97.6%, and the NPV was 54.6%. Performance was consistent across gender (sensitivity 79.3% in men and 76.8% in women) and age groups, with a specificity of 93.4% in men and 92.7% in women. Stage I lung cancer was detected with a sensitivity of 80.6%. Conclusions: The Lung Cancer test based on 9G technology presented here offers a non-invasive method for stratifying lung cancer risk in patients with CT-detected nodules. Its integration into clinical practice could reduce unnecessary interventions and foster earlier detection. [ABSTRACT FROM AUTHOR]

To explore the predictive factors of axillary lymph node metastasis in breast cancer, and to provide a basis for clinical decision-making under the DIP payment mode of medical insurance. Methods A total of 715 patients with breast cancer were divided into the metastasis group (n=309) and the non-metastasis group (n=406) according to the postoperative paraffin pathological results. Data of age > 60 years old, menopausal status, body mass index (BMI) > 24 kg/m², hyperglycemia (GLU > 6.1 mmol/L), high triglycerides (TG > 1.7 mmol/L), maximum diameter of the tumor, the distance between the tumor and nipple and the quadrant where the tumor located were compared between the two groups. The expression levels of estrogen receptor (ER), progesterone receptor (PR), nuclear proliferation antigen (Ki-67) and human epidermal growth factor receptor-2 (Her-2) in breast cancer tissue samples were detected by histological grading and immunohistochemistry. The consistency, sensitivity and specificity of chest CT and breast ultrasound were examined, taken the pathological diagnosis as the gold standard. Results Compared with the non-metastatic group, the proportion of maximum diameter of tumor > 2 cm, histological grade III, high Ki-67 and high ER expression, tumor located in the outer upper quadrant, the distance > 3 cm between tumor and nipple were increased in the metastatic group, and the proportion of high level of TG was decreased in the metastatic group (P < 0.05). The consistency between chest CT and pathological diagnosis was better than that of breast ultrasound (Kappa was 0.493 and 0.353 respectively, P < 0.05). Logistic regression analysis showed that histological grade III, high expression of ER, maximum diameter of tumor > 2 cm, and chest CT diagnosis were risk factors for axillary lymph node metastasis (P < 0.05). Conclusion The combined application of the predictive factors of axillary lymph node metastasis of breast cancer could provide certain reference for clinical decision-making under the background of DIP payment mode of medical insurance. [ABSTRACT FROM AUTHOR]

Background: We have previously demonstrated that high-risk obstructive sleep apnea (HR-OSA), based on a modified Berlin Questionnaire (mBQ), is linked to worse clinical outcomes. Chest computed tomography (CT) imaging with the implementation of an artificial intelligence (AI) analysis program has been a valuable tool for the speedy assessment of huge numbers of patients during the COVID-19 epidemic. In the current study, we addressed how the severity of AI-guided, CT-based total opacity ratio (TOR) scores are associated with high-risk OSA and short-term outcomes in the same cohort. Methods: The ratio of the volume of high opacity areas to that of the total lung volume constituted the TOR. We arbitrarily applied thresholds of <5 (no or mild TOR), ≥5 and <15 (moderate TOR), and ≥15 (severe TOR). Results: In total, 221 patients were included. HR-OSA was observed among 11.0% of the no or mild TOR group, 22.2% of the moderate TOR group, and 38.7% of the severe TOR group (p < 0.001). In a logistic regression analysis, HR-OSA was associated with a severe TOR with an adjusted odds ratio of 3.06 (95% confidence interval [CI] 1.27–7.44; p = 0.01). A moderate TOR predicted clinical worsening with an adjusted hazard ratio (HR) of 1.93 (95% CI 1.00–3.72; p = 0.05) and a severe TOR predicted worsening with an HR of 3.06 (95% CI 1.56–5.99; p = 0.001). Conclusions: Our results offer additional radiological proof of the relationship between HR-OSA and worse outcomes in patients with COVID-19 pneumonia. A TOR may also potentially indicate the individuals that are at higher risk of HR-OSA, enabling early intervention and management strategies. The clinical significance of TOR thresholds needs further evaluation in larger samples. [ABSTRACT FROM AUTHOR]

Copyright of Die Radiologie is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Purpose: Opportunistic osteoporosis screening, conducted during routine medical examinations such as chest computed tomography (CT), presents a potential solution for early detection. This study aims to investigate the feasibility of utilizing radiomics technology based on chest CT images to screen for opportunistic osteoporosis. Methods: This Study is a Multicenter Retrospective Investigation. Relevant clinical data, including demographics and DXA results, would be collected for each participant. The radiomics analysis in this study focuses on the extraction of features from the 11th or 12th thoracic vertebral bodies from chest CT images. SVM machine learning models would be trained using these radiomic features, with DXA results as the ground truth for osteoporosis classification. Results: In the training group, Clinical models had an accuracy of 0.684 and an AUC of 0.744, Radiomics models had an accuracy of 0.828 and an AUC of 0.896, Nomogram models had an accuracy of 0.839 and an AUC of 0.901. In the internal validation group, Clinical models had an accuracy of 0.769 and an AUC of 0.829, Radiomics models had an accuracy of 0.832 and an AUC of 0.892, Nomogram models had an accuracy of 0.839 and an AUC of 0.918. In the external validation group, Clinical models had an accuracy of 0.715 and an AUC of 0.741, Radiomics models had an accuracy of 0.777 and an AUC of 0.796, Nomogram models had an accuracy of 0.785 and an AUC of 0.807. In all three datasets, the Nomogram model exhibited a statistically significant difference in screening effectiveness compared to the clinical models. Conclusion: Our research demonstrates that by leveraging radiomics features extracted from a single thoracic spine using chest CT, and incorporating these features with patient basic information, opportunistic screening for osteoporosis can be achieved. [ABSTRACT FROM AUTHOR]

Background: There has been growing interest in using artificial intelligence/deep learning (DL) to help diagnose prevalent diseases earlier. In this study we sought to survey the landscape of externally validated DL-based computer-aided diagnostic (CADx) models, and assess their diagnostic performance for predicting the risk of malignancy in computed tomography (CT)-detected pulmonary nodules. Methods: An electronic search was performed in four databases (from inception to 10 August 2023). Studies were eligible if they were peer-reviewed experimental or observational articles comparing the diagnostic performance of externally validated DL-based CADx models with models widely used in clinical practice to predict the risk of malignancy. A bivariate random-effect approach for the meta-analysis on the included studies was used. Results: Seventeen studies were included, comprising 8553 participants and 9884 nodules. Pooled analyses showed DL-based CADx models were 11.6% more sensitive than physician judgement alone, and 14.5% more than clinical risk models alone. They had a similar pooled specificity to physician judgement alone [0.77 (95% CI 0.68–0.84) v 0.81 (95% CI 0.71–0.88)], and were 7.4% more specific than clinical risk models alone. They had superior pooled areas under the receiver operating curve (AUC), with relative pooled AUCs of 1.03 (95% CI 1.00–1.07) and 1.10 (95% CI 1.07–1.13) versus physician judgement and clinical risk models alone, respectively. Conclusion: DL-based models are already used in clinical practice in certain settings for nodule management. Our results show their diagnostic performance potentially justifies wider, more routine deployment alongside experienced physician readers to help inform multidisciplinary team decision-making. [ABSTRACT FROM AUTHOR]

Bronchogenic cysts, a rare congenital pulmonary disorder, typically affect young adults and are often managed conservatively. However, large cysts with recurrent infections require surgical intervention. This case study highlights the successful management of a large bronchogenic cyst. A 53-year-old female presented with a decade-long history of recurrent respiratory infections manifesting as cough, yellow purulent sputum, and shortness of breath. Chest computed tomography revealed a large bronchogenic cyst in the right middle lobe, causing cardiac compression. Despite conservative management, the recurrent symptoms persisted. After multidisciplinary consultation, a thoracoscopic right middle lobectomy was planned. Severe pleural adhesions and bleeding complicated the procedure; therefore, thoracotomy was performed. Postoperatively, the patient developed transient fever and elevated white blood cell count, both of which resolved with appropriate antibiotic therapy. The patient was discharged in stable condition, with no recurrence of symptoms at follow-up. Large, symptomatic bronchogenic cysts that cause recurrent infections require surgical resection. [ABSTRACT FROM AUTHOR]

Purpose: Childhood and young adult cancer survivors exposed to chest radiotherapy are at increased risk of lung cancer. In other high-risk populations, lung cancer screening has been recommended. Data is lacking on prevalence of benign and malignant pulmonary parenchymal abnormalities in this population. Methods: We conducted a retrospective review of pulmonary parenchymal abnormalities in chest CTs performed more than 5 years post-cancer diagnosis in survivors of childhood, adolescent, and young adult cancer. We included survivors exposed to radiotherapy involving the lung field and followed at a high-risk survivorship clinic between November 2005 and May 2016. Treatment exposures and clinical outcomes were abstracted from medical records. Risk factors for chest CT–detected pulmonary nodule were assessed. Results: Five hundred and ninety survivors were included in this analysis: median age at diagnosis, 17.1 years (range, 0.4–39.8); and median time since diagnosis, 22.3 years (range, 1–58.6). At least one chest CT more than 5 years post-diagnosis was performed in 338 survivors (57%). Among these, 193 (57.1%) survivors had at least one pulmonary nodule detected on a total of 1057 chest CTs, resulting in 305 CTs with 448 unique nodules. Follow-up was available for 435 of these nodules; 19 (4.3%) were malignant. Risk factors for first pulmonary nodule were older age at time of CT, CT performed more recently, and splenectomy. Conclusions: Benign pulmonary nodules are very common among long-term survivors of childhood and young adult cancer. Implications for Cancer Survivors: High prevalence of benign pulmonary nodules in cancer survivors exposed to radiotherapy could inform future guidelines on lung cancer screening in this population. [ABSTRACT FROM AUTHOR]

Multidisciplinary pulmonary embolism response teams (PERTs) have shown that timely triage expedites treatment. The use of artificial intelligence (AI) may help improve pulmonary embolism (PE) management with early CT pulmonary angiogram (CTPA) screening and accelerate PERT coordination. This study aimed to test the clinical validity of an FDA-approved PE AI algorithm. CTPA scan data of 200 patients referred due to automated AI detection of suspected PE were retrospectively reviewed. In our institution, all patients suspected of PE received a CTPA. The AI app was then used to analyze CTPA for the presence of PE and calculate the right-ventricle/left-ventricle (RV/LV) ratio. We compared the AI's output with the radiologists' report. Inclusion criteria included segmental PE with and without RV dysfunction and high-risk PE. The primary endpoint was false positive rate. Secondary end points included clinical outcomes according to the therapy selected, including catheter-directed interventions, systemic thrombolytics, and anticoagulation. Fifty-seven of 200 exams (28.5%) were correctly identified as positive for PE by the algorithm. A total of 143 exams (71.5%) were incorrectly reported as positive. In 8% of cases, PERT was consulted. Four patients (7%) received systemic thrombolytics without any complications. There were six patients (10.5%) who developed high-risk PE and underwent thrombectomy, one of whom died. Among 46 patients with acute PE without right heart strain, 44 (95%) survived. The false positive rate of our AI algorithm was 71.5%, higher than what was reported in the AI's prior clinical validity study (91% sensitivity, 100% specificity). A high rate of discordant AI auto-detection of suspected PE raises concerns about its diagnostic accuracy. This can lead to increased workloads for PERT consultants, alarm/notification fatigue, and automation bias. The AI direct notification process to the PERT team did not improve PERT triage efficacy. [ABSTRACT FROM AUTHOR]

Histiocytic sarcoma is an extremely rare disease that's hard to diagnose and treat, often leading to a poor prognosis. Here, we present a case report detailing a rare occurrence of HS in a 37-year-old pregnant woman who first presented with left shoulder pain, palpitations, and a productive cough at 20 weeks of gestation. Her diagnostic evaluations were performed, including different imaging modalities such as chest X-rays, CT scans, and MRI. Imaging revealed a large mediastinal mass with extensive involvement of the adrenal glands, lungs, and lymph nodes. The definitive diagnosis of HS is based on pathological and morphological features, and the immunohistochemistry report plays a key role. In our case, the diagnosis of HS was confirmed through pathological evaluation and immunohistochemistry, with a positive CD68 result obtained from a supraclavicular lymph node biopsy. A hospital committee comprising medical specialists like hematologists-oncologists, pathologists, pulmonologists, and obstetricians was brought together to assess the case collectively. The patient received chemotherapy, which alleviated her symptoms and maintained her condition. Based on the committee's recommendations, despite a healthy fetus and normal obstetric sonograms, the decision was made to terminate the pregnancy with the consent of the patient and her family. Despite initial improvement postchemotherapy, the patient's condition worsened, necessitating intubation. Tragically, two months after the initial admission, the patient passed away due to severe complications. In this case report, we provide a literature review and review of the patient's imaging reports. Since the patient is pregnant and HS is uncommon, it's important to highlight that this case is unique and worth sharing.

Atopobium minutum (A. minutum) has rarely been documented in human infections. However, this report describes a case involving a 52-year-old woman who developed empyema and lung collapse due to A. minutum. She initially presented to the emergency department with nausea, vomiting, diarrhea, and abdominal pain. Her condition quickly declined within the first day of arrival, leading to respiratory failure and requiring intubation and ICU-level care. Despite receiving intensive antibiotic treatment, the patient needed prolonged intubation and a tracheostomy. Initial cultures indicated Streptococcus intermedius and Lactobacillus minutus, but final culture results identified A. minutum as the cause. This case highlights the difficulty in diagnosing A. minutum infections, often necessitating advanced DNA sequencing, and raises concerns about potential multidrug resistance. It highlights the importance of prompt identification of the pathogen by laboratories to allow for effective treatment of these rare infections.

Introduction: Coronavirus disease 2019 (COVID-19), caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), presents a broad spectrum of clinical manifestations, from asymptomatic cases to severe pneumonia and acute respiratory distress syndrome (ARDS). Chest computed tomography (CT) has become a critical diagnostic tool, especially in regions with high disease prevalence. This study aims to assess the common CT findings, patterns of lung involvement, and severity of disease in RT-PCR positive COVID-19 patients to better understand and manage this pervasive disease. Materials and Methods: This observational cross-sectional study was conducted at Dr. Kamakshi Memorial Hospital, Chennai, from 2020 to 2021. The study included RT-PCR positive COVID-19 patients aged 18 to 95 years who presented to the fever clinic or COVID casualty and were referred to the radiology department for chest CT evaluation. Exclusion criteria included pregnancy, age under 18 years, and refusal to consent. Non-contrast chest CT scans were performed using a TOSHIBA Aquilion Lightening 16-slice CT machine. Scans were acquired in a single inspiratory breath-hold from the lung apex to the costophrenic angle. CT findings were analyzed and reported by two experienced radiologists, with discrepancies resolved through consensus. Results: Out of 349 patients, 213 (61%) were male and 136 (39%) were female, with a mean age of 47.7 years. The distribution of CT findings showed significant variability among the four groups. Group A had the highest percentage of normal CT scans (22%) and ground-glass opacities (52%). Group B exhibited a reduction in normal CT scans (9%) and an increase in ground-glass opacities (57%). Group C showed further decrease in normal CT scans (10.6%) with increased crazy paving (17.3%) and reticulation (14.6%). Group D had similar normal CT scans (10.8%) but significantly higher incidences of reticulation (24.3%) and ground-glass opacities (64%). Conclusion: This study highlights the critical role of chest CT in monitoring the progression of COVID-19 pneumonia. The findings demonstrate a clear temporal evolution of lung involvement, from ground-glass opacities in the early stages to more complex patterns such as crazy paving and reticulation in later stages.

Abstract Background Liver injuries such as metabolic-associated fatty liver disease, liver fibrosis, and steatosis are common in COVID-19 patients. Unenhanced CT can be used to diagnose the morphological traits of steatosis and cirrhosis. This study aims to provide a clear overview on the association between liver injuries and decreased hepatic CT attenuation values on chest CT images in patients with COVID-19. Main text Measuring HU values can be used as an additional method to diagnose liver injuries, even though HU values alone cannot definitively diagnose specific liver diseases. Chest CT is a common imaging procedure for diagnosing pneumonia, and during this CT examination, the upper abdomen, including the liver and spleen, is incidentally captured on the CT scan. Therefore, the assessment of liver injuries in chest CT of patients with COVID-19 can be performed by measuring the HU value of the liver and spleen. In this review, we summarize all the currently available CT findings in liver injuries associated with decreased hepatic CT attenuation values. Conclusion We found out that liver injuries such as hepatic steatosis and metabolic disease were more frequent in the COVID-19 patient, especially in severe and ICU patients. Compared to control group and COVID-19 patients with mild symptoms, the hepatic CT attenuation values and L/S ratios were lower in research group and severe COVID-19 patients.

Tinghua Hu,1,&ast; Shanshan Dai,2,&ast; Lan Yang,1 Bo Zhu1 1Department of Respiratory and Critical Care Medicine, The First Affiliated Hospital of Xi’an Jiaotong University, Xi’an, Shaanxi, 710000, People’s Republic of China; 2Department of Respiratory and Critical Care Medicine, Xi’an No. 9 Hospital, Xi’an, Shaanxi, People’s Republic of China&ast;These authors contributed equally to this workCorrespondence: Bo Zhu, Department of Respiratory and Critical Care Medicine, The First Affiliated Hospital of Xi’an Jiaotong University, 277 Yanta West Road, Xi’an, Shaanxi, 710061, People’s Republic of China, Email zhubo685689@163.comBackground: COPD, combined with Osteoporosis, has a high incidence and potential for great harm. Choosing an optimal diagnostic method to achieve bone mineral density (BMD) screening is crucial for COPD patients. Studies on COPD patients with BMD reduction are lacking.Purpose: To identify the risk factors of BMD reduction and osteoporosis in COPD patients.Patients and Methods: We included a total of 81 patients with AECOPD, who were admitted to the hospital from July 1, 2019, to January 31, 2020. Patients were grouped into BMD normal group, BMD reduced group and OP group. The areas under ROC curve were used to explore the value of CT values in the diagnosis of bone abnormality, and clinical indicators were collected.Results: The CT value of the vertebral cancellous bone is highly correlated with the T value of BMD (R > 5.5, P < 0.0001). Using multivariate Logistic regression analysis, we showed that COPD duration, BMI, 25-hydroxyvitamin D3, and long-term inhaled glucocorticoid were independent factors affecting different BMD levels in COPD patients. No significant difference in bone formation indexes between groups. β-crossL was negatively correlated with serum IL-6 (r=− 0.254, P=0.022), and ALP was positively correlated with serum TNF-α (r=0.284, P=0.023).Conclusion: Thoracolumbar vertebral cancellous bone CT has potential value in the diagnosis of bone abnormality. COPD duration, BMI, 25-hydroxyvitamin D3, and long-term inhaled glucocorticoid may contribute to the BMD reduction in COPD patients, and serum IL-6 and TNF-α regulate bone metabolism in COPD patients.Keywords: chronic obstructive pulmonary disease, osteoporosis, bone mineral density, chest CT

The COVID-19 pandemic has spread worldwide for over two years. It was considered a significant threat to global health due to its transmissibility and high pathogenicity. The standard test for COVID-19, namely, reverse transcription polymerase chain reaction (RT–PCR), is somehow inaccurate and might have a high false-negative rate (FNR). As a result, an infected person with a negative test result may unknowingly continue to spread the virus, especially if they are infected with an undiscovered COVID-19 strain. Thus, a more accurate diagnostic technique is required. In this study, we propose 3Cs, which is a capsule neural network (CapsNet) used to classify computed tomography (CT) images as novel coronavirus pneumonia (NCP), common pneumonia (CP), or normal lungs. Using 6123 CT images of healthy patients’ lungs and those of patients with CP and NCP, the 3Cs method achieved an accuracy of around 98% and an FNR of about 2%, demonstrating CapNet’s ability to extract features from CT images that distinguish between healthy and infected lungs. This research confirmed that using CapsNet to detect COVID-19 from CT images results in a lower FNR compared to RT–PCR. Thus, it can be used in conjunction with RT–PCR to diagnose COVID-19 regardless of the variant.

PURPOSE: Since the prompt recognition of acute pulmonary embolism (PE) and the immediate initiation of treatment can significantly reduce the risk of death, we developed a deep learning (DL)-based application aimed to automatically detect PEs on chest computed tomography angiograms (CTAs) and alert radiologists for an urgent interpretation. Convolutional neural networks (CNNs) were used to design the application. The associated algorithm used a hybrid 3D/2D UNet topology. The training phase was performed on datasets adequately distributed in terms of vendors, patient age, slice thickness, and kVp. The objective of this study was to validate the performance of the algorithm in detecting suspected PEs on CTAs. METHODS: The validation dataset included 387 anonymized real-world chest CTAs from multiple clinical sites (228 U.S. cities). The data were acquired on 41 different scanner models from five different scanner makers. The ground truth (presence or absence of PE on CTA images) was established by three independent U.S. board-certified radiologists. RESULTS: The algorithm correctly identified 170 of 186 exams positive for PE (sensitivity 91.4% [95% CI: 86.4-95.0%]) and 184 of 201 exams negative for PE (specificity 91.5% [95% CI: 86.8-95.0%]), leading to an accuracy of 91.5%. False negative cases were either chronic PEs or PEs at the limit of subsegmental arteries and close to partial volume effect artifacts. Most of the false positive findings were due to contrast agent-related fluid artifacts, pulmonary veins, and lymph nodes. CONCLUSIONS: The DL-based algorithm has a high degree of diagnostic accuracy with balanced sensitivity and specificity for the detection of PE on CTAs.

Introduction: This study aimed to identify useful clinical indicators for predicting the relapse of interstitial lung disease (ILD) complicated with anti-aminoacyltRNA synthetase (ARS) antibodies (anti-ARS-ILD), being treated with prednisolone and calcineurin inhibitors. Methods: Fifty patients with anti-ARS-ILD were enrolled between October 2014 and August 2022. All patients were treated with prednisolone and calcineurin inhibitors as remission induction therapy and followed up for over a year with these combination therapies. We examined patients who experienced ILD relapse after immunosuppressive treatment. We explored the risk factors for predicting ILD relapse in these patients by comparing demographic, clinical, laboratory, and radiological findings and treatments between the relapsed and non-relapsed groups on admission. Results: Of the 50 patients, 19 (38%) relapsed during a median follow-up of 4.8 years. Univariate and multivariate Cox regression analyses identified the presence of acute/subacute (A/S)-ILD, higher serum aldolase (ALD) and surfactant protein-D (SP-D) levels, and lower %forced vital capacity (FVC) as risk factors for relapse in patients with anti-ARS-ILD. Using the receiver operating curve analysis, ALD =6.3 U/L, SP-D =207 ng/mL, and %FVC =76.8% were determined as the cut-off levels for indicating a poor prognosis. The 5-year relapse rate was significantly higher in patients with A/S-ILD, serum ALD=6.3 U/L, serum SP-D =207 ng/mL, or %FVC of =76.8% than in those without these parameters. (P=0.009, 0.0005, 0.0007, 0.0004, respectively) Serum ALD levels were significantly correlated with the disease activity indicators of anti-ARS-ILD. Conclusion: The presence of A/S-ILD, higher serum ALD and SP-D levels, and lower %FVC are useful indicators for predicting anti-ARS-ILD relapse. [ABSTRACT FROM AUTHOR]

Background/Objectives: The accurate quantification of ground-glass opacities (GGOs) and consolidation volumes has prognostic value in COVID-19 patients. Nevertheless, the accurate manual quantification of the corresponding volumes remains a time-consuming task. Deep learning (DL) has demonstrated good performance in the segmentation of normal lung parenchyma and COVID-19 pneumonia. We introduce a Human-in-the-Loop (HITL) strategy for the segmentation of normal lung parenchyma and COVID-19 pneumonia that is both time efficient and quality effective. Furthermore, we propose a Gaussian Mixture Model (GMM) to classify GGO and consolidation based on a probabilistic characterization and case-sensitive thresholds. Methods: A total of 65 Computed Tomography (CT) scans from 64 patients, acquired between March 2020 and June 2021, were randomly selected. We pretrained a 3D-UNet with an international dataset and implemented a HITL strategy to refine the local dataset with delineations by teams of medical interns, radiology residents, and radiologists. Following each HITL cycle, 3D-UNet was re-trained until the Dice Similarity Coefficients (DSCs) reached the quality criteria set by radiologists (DSC = 0.95/0.8 for the normal lung parenchyma/COVID-19 pneumonia). For the probabilistic characterization, a Gaussian Mixture Model (GMM) was fitted to the Hounsfield Units (HUs) of voxels from the CT scans of patients with COVID-19 pneumonia on the assumption that two distinct populations were superimposed: one for GGO and one for consolidation. Results: Manual delineation of the normal lung parenchyma and COVID-19 pneumonia was performed by seven teams on 65 CT scans from 64 patients (56 ± 16 years old (μ ± σ), 46 males, 62 with reported symptoms). Automated lung/COVID-19 pneumonia segmentation with a DSC > 0.96/0.81 was achieved after three HITL cycles. The HITL strategy improved the DSC by 0.2 and 0.5 for the normal lung parenchyma and COVID-19 pneumonia segmentation, respectively. The distribution of the patient-specific thresholds derived from the GMM yielded a mean of −528.4 ± 99.5 HU (μ ± σ), which is below most of the reported fixed HU thresholds. Conclusions: The HITL strategy allowed for fast and effective annotations, thereby enhancing the quality of segmentation for a local CT dataset. Probabilistic characterization of COVID-19 pneumonia by the GMM enabled patient-specific segmentation of GGO and consolidation. The combination of both approaches is essential to gain confidence in DL approaches in our local environment. The patient-specific probabilistic approach, when combined with the automatic quantification of COVID-19 imaging findings, enhances the understanding of GGO and consolidation during the course of the disease, with the potential to improve the accuracy of clinical predictions. [ABSTRACT FROM AUTHOR]

Following the great success of various deep learning methods in image and object classification, the biomedical image processing society is also overwhelmed with their applications to various automatic diagnosis cases. Unfortunately, most of the deep learning-based classification attempts in the literature solely focus on the aim of extreme accuracy scores, without considering interpretability, or patient-wise separation of training and test data. For example, most lung nodule classification papers using deep learning randomly shuffle data and split it into training, validation, and test sets, causing certain images from the Computed Tomography (CT) scan of a person to be in the training set, while other images of the same person to be in the validation or testing image sets. This can result in reporting misleading accuracy rates and the learning of irrelevant features, ultimately reducing the real-life usability of these models. When the deep neural networks trained on the traditional, unfair data shuffling method are challenged with new patient images, it is observed that the trained models perform poorly. In contrast, deep neural networks trained with strict patient-level separation maintain their accuracy rates even when new patient images are tested. Heat map visualizations of the activations of the deep neural networks trained with strict patient-level separation indicate a higher degree of focus on the relevant nodules. We argue that the research question posed in the title has a positive answer only if the deep neural networks are trained with images of patients that are strictly isolated from the validation and testing patient sets. [ABSTRACT FROM AUTHOR]