While deep convolutional neural networks (DCNNs) have achieved remarkable performance in chest X-ray interpretation, their success typically depends on access to large-scale, expertly annotated datasets. However, collecting such data in real-world clinical settings can be difficult because of limited labeling resources, privacy concerns, and patient variability. In this study, we applied a multimodal Transformer pretrained on free-text reports and their paired CXRs to evaluate the effectiveness of this method in settings with limited labeled data. Our dataset consisted of more than 1 million CXRs, each accompanied by reports from board-certified radiologists and 31 structured labels. The results indicated that a linear model trained on embeddings from the pretrained model achieved AUCs of 0.907 and 0.903 on internal and external test sets, respectively, using only 128 cases and 384 controls; the results were comparable those of DenseNet trained on the entire dataset, whose AUCs were 0.908 and 0.903, respectively. Additionally, we demonstrated similar results by extending the application of this approach to a subset annotated with structured echocardiographic reports. Furthermore, this multimodal model exhibited excellent small sample learning capabilities when tested on external validation sets such as CheXpert and ChestX-ray14. This research significantly reduces the sample size necessary for future artificial intelligence advancements in CXR interpretation.

Vision-grounded medical report generation aims to produce clinically accurate descriptions of medical images, anchored in explicit visual evidence to improve interpretability and facilitate integration into clinical workflows. However, existing methods often rely on separately trained detection modules that require extensive expert annotations, introducing high labeling costs and limiting generalizability due to pathology distribution bias across datasets. To address these challenges, we propose Self-Supervised Anatomical Consistency Learning (SS-ACL) -- a novel and annotation-free framework that aligns generated reports with corresponding anatomical regions using simple textual prompts. SS-ACL constructs a hierarchical anatomical graph inspired by the invariant top-down inclusion structure of human anatomy, organizing entities by spatial location. It recursively reconstructs fine-grained anatomical regions to enforce intra-sample spatial alignment, inherently guiding attention maps toward visually relevant areas prompted by text. To further enhance inter-sample semantic alignment for abnormality recognition, SS-ACL introduces a region-level contrastive learning based on anatomical consistency. These aligned embeddings serve as priors for report generation, enabling attention maps to provide interpretable visual evidence. Extensive experiments demonstrate that SS-ACL, without relying on expert annotations, (i) generates accurate and visually grounded reports -- outperforming state-of-the-art methods by 10\% in lexical accuracy and 25\% in clinical efficacy, and (ii) achieves competitive performance on various downstream visual tasks, surpassing current leading visual foundation models by 8\% in zero-shot visual grounding.

The precise contouring of gross tumor volume lymph nodes (GTVnd) is an essential step in clinical target volume delineation. This study aims to propose and evaluate a deep learning model for segmenting GTVnd specifically in lung cancer, representing one of the pioneering investigations into automated segmentation of GTVnd specifically for lung cancer. Ninety computed tomography (CT) scans of patients with stage Ш-Ⅳ small cell lung cancer (SCLC) were collected, of which 75 patients were assembled into a training dataset and 15 were used in a testing dataset. A new segmentation model was constructed to enable the automatic and accurate delineation of the GTVnd in lung cancer. This model integrates a contextual cue enhancement module and an edge-guided feature enhancement decoder. The contextual cues enhancement module was used to enforce the consistency of the contextual cues encoded in the deepest feature, and the edge-guided feature enhancement decoder was used to obtain edge-aware and edge-preserving segmentation predictions. The model was quantitatively evaluated using the three-dimensional Dice Similarity Coefficient (3D DSC) and the 95th Hausdorff Distance (95HD). Additionally, comparative analysis was conducted between predicted treatment plans derived from auto-contouring GTVnd and established clinical plans. The ECENet achieved a mean 3D DSC of 0.72 ± 0.09 and a 95HD of 6.39 ± 4.59 mm, showing significant improvement compared to UNet, with a DSC of 0.46 ± 0.19 and a 95HD of 12.24 ± 13.36 mm, and nnUNet, with a DSC of 0.52 ± 0.18 and a 95HD of 9.92 ± 6.49 mm. Its performance was intermediate, falling between mid-level physicians, with a DSC of 0.81 ± 0.06, and junior physicians, with a DSC of 0.68 ± 0.10. And the clinical and predicted treatment plans were further compared. The dosimetric analysis demonstrated excellent agreement between predicted and clinical plans, with average relative deviation of < 0.17% for PTV D2/D50/D98, < 3.5% for lung V30/V20/V10/V5/Dmean, and < 6.1% for heart V40/V30/Dmean. Furthermore, the TCP (66.99% ± 0.55 vs. 66.88% ± 0.45) and NTCP (3.13% ± 1.33 vs. 3.25% ± 1.42) analyses revealed strong concordance between predicted and clinical outcomes, confirming the clinical applicability of the proposed method. The proposed model could achieve the automatic delineation of the GTVnd in the thoracic region of lung cancer and showed certain advantages, making it a potential choice for the automatic delineation of the GTVnd in lung cancer, particularly for young radiation oncologists.

Decisions regarding veno-venous extracorporeal membrane oxygenation (vv-ECMO) in patients with acute respiratory distress syndrome (ARDS) are often based solely on clinical and physiological parameters, which may insufficiently reflect severity and heterogeneity of lung injury. This study aimed to develop a predictive model integrating machine learning-derived quantitative features from admission chest computed tomography (CT) with selected clinical variables to support early individualized decision-making regarding vv-ECMO therapy. In this retrospective single-center cohort study, 375 consecutive patients with COVID-19-associated ARDS admitted to the ICU between March 2020 and April 2022 were included. Lung segmentation from initial CTs was performed using a convolutional neural network (CNN) to generate high-resolution, anatomically accurate masks of the lungs. Subsequently, 592 radiomic features, quantifying lung aeration, density and morphology, were extracted. Four clinical parameters - age, mean airway pressure, lactate, and C-reactive protein, were selected on the basis of clinical relevance. Three logistic regression models were developed: (1) Imaging Model, (2) Clinical Model, and (3) Combined Model integrating different features. Predictive performance was assessed via the area under the receiver operating characteristic curve (AUROC), accuracy, sensitivity, and specificity. A total of 375 patients were included: 172 in the training and 203 in the validation cohort. In the training cohort, the AUROCs were 0.743 (Imaging), 0.828 (Clinical), and 0.842 (Combined). In the validation cohort, the Combined Model achieved the highest AUROC (0.705), outperforming the Clinical (0.674) and Imaging (0.639) Models. Overall accuracy in the validation cohort was 64.0% (Combined), 66.5% (Clinical), and 59.1% (Imaging). The Combined Model showed 68.1% sensitivity and 58.9% specificity. Kaplan-Meier analysis confirmed a significantly greater cumulative incidence of ECMO therapy in patients predicted as high risk (p < 0.001), underscoring its potential to support individualized, timely ECMO decisions in ARDS by providing clinicians with objective data-driven risk estimates. Quantitative CT features based on machine learning-derived lung segmentation allow early individualized prediction of the need for vv-ECMO in ARDS. While clinical data remain essential, radiomic markers enhance prognostic accuracy. The Combined Model demonstrates considerable potential to support timely and evidence-based ECMO initiation, facilitating individualized critical care in both specialized and general ICU environments.Trial registration: The study is registered with the German Clinical Trials Register under the number DRKS00027856. Registered 18.01.2022, retrospectively registered due to retrospective design of the study.

The purpose of this study was to evaluate the relationship between structural abnormalities on CT and lung function prior to and after initiation of elexacaftor-tezacaftor-ivacaftor (ETI) in adults with cystic fibrosis (CF) using a deep learning model. A deep learning quantification model was developed using 100 chest computed tomography (CT) examinations of patients with CF and 150 chest CT examinations of patients with various other bronchial diseases to quantify seven types of abnormalities. This model was then applied to an independent dataset of CT examinations of 218 adults with CF who were treated with ETI. The relationship between structural abnormalities and percent predicted forced expiratory volume in one second (ppFEV₁) was examined using general linear regression models. The deep learning model performed as well as radiologists for the quantification of the seven types of abnormalities. Chest CT examinations obtained before to and one year after the initiation of ETI were analyzed. The independent structural predictors of ppFEV₁ prior to ETI were bronchial wall thickening (P = 0.011), mucus plugging (P < 0.001), consolidation/atelectasis (P < 0.001), and mosaic perfusion (P < 0.001). An increase in ppFEV₁ after initiation of ETI independently correlated with a decrease in bronchial wall thicknening (-49 %; P = 0.004), mucus plugging (-92 %; P < 0.001), centrilobular nodules (-78 %; P = 0.009) and mosaic perfusion (-14 %; P < 0.001). Younger age (P < 0.001), greater mucus plugging extent (P = 0.016), and centrilobular nodules (P < 0.001) prior to ETI initiation were independent predictors of ppFEV₁ improvement. A deep learning model can quantify CT lung abnormalities in adults with CF. Lung function impairment in adults with CF is associated with muco-inflammatory lesions on CT, which are largely reversible with ETI, and with mosaic perfusion, which appear less reversible and is presumably related to irreversible damage. Predictors of lung function improvement are a younger age and a greater extent of muco-inflammatory lesions obstructing the airways.

Acute pulmonary embolism (PE) is a life-threatening condition often diagnosed using CT pulmonary angiography (CTPA). However, CTPA is contraindicated in patients with contrast allergies or at risk for contrast-induced nephropathy. This study explores an AI-driven approach to generate synthetic contrast-enhanced images from non-contrast CT scans for accurate diagnosis of acute PE without contrast agents. This retrospective study used dual-energy and standard CT datasets from two institutions. The internal dataset included 84 patients: 41 PE-negative cases for generative model training and 43 patients (30 PE-positive) for diagnostic evaluation. An external dataset of 62 patients (26 PE-positive) was used for further validation. We developed a generative adversarial network (GAN) based on U-Net, trained on paired non-contrast and contrast-enhanced images. The model was optimized using contrast-enhanced L1-loss with hyperparameter λ to improve anatomical accuracy. A ConvNeXt-based classifier trained on the RSNA dataset (N = 7,122) generated per-slice PE probabilities, which were aggregated for patient-level prediction via a Random Forest model. Diagnostic performance was assessed using five-fold cross-validation on both internal and external datasets. The GAN achieved optimal image similarity at λ = 0.5, with the lowest mean absolute error (0.0089) and highest MS-SSIM (0.9674). PE classification yielded AUCs of 0.861 and 0.836 in the internal dataset, and 0.787 and 0.680 in the external dataset, using real and synthetic images, respectively. No statistically significant differences were observed. Our findings demonstrate that synthetic contrast CT can serve as a viable alternative for PE diagnosis in patients contraindicated for CTPA, supporting safe and accessible imaging strategies.

To quantify the impact of workflow parameters on time savings in report turnaround time due to an AI triage device that prioritized pulmonary embolism (PE) in chest CT pulmonary angiography (CTPA) examinations. This retrospective study analyzed 11,252 adult CTPA examinations conducted for suspected PE at a single tertiary academic medical center. Data was divided into two periods: pre-artificial intelligence (AI) and post-AI. For PE-positive examinations, turnaround time (TAT)-defined as the duration from patient scan completion to the first preliminary report completion-was compared between the two periods. Time savings were reported separately for work-hour and off-hour cohorts. To characterize radiologist workflow, 527,234 records were retrieved from the PACS and workflow parameters such as examination interarrival time and radiologist read time extracted. These parameters were input into a computational model to predict time savings after deployment of an AI triage device and to study the impact of workflow parameters. The pre-AI dataset included 4,694 chest CTPA examinations with 13.3% being PE-positive. The post-AI dataset comprised 6,558 examinations with 16.2% being PE-positive. The mean TAT for pre-AI and post-AI during work hours are 68.9 (95% confidence interval 55.0-82.8) and 46.7 (38.1-55.2) min, respectively, and those during off-hours are 44.8 (33.7-55.9) and 42.0 (33.6-50.3) min. Clinically observed time savings during work hours (22.2 [95% confidence interval: 5.85-38.6] min) were significant (P = .004), while off-hour (2.82 [-11.1 to 16.7] min) were not (P = .345). Observed time savings aligned with model predictions (29.6 [95% range: 23.2-38.1] min for work hours; 2.10 [1.76, 2.58] min for off-hours). Consideration and quantification of the clinical workflow contributes to the accurate assessment of the expected time savings in report TAT after deployment of an AI triage device.

Despite early detection via low-dose computed tomography and complete surgical resection for early-stage lung adenocarcinoma, postoperative recurrence remains high, particularly in patients with tumor spread through air spaces. A reliable preoperative prediction model is urgently needed to adjust the treatment modality. In this multicenter retrospective study, 609 patients with pathological stage I lung adenocarcinoma from 3 independent centers were enrolled. Regions of interest for the primary tumor and peritumoral areas (extended by three, six, and twelve voxel units) were manually delineated from preoperative CT imaging. Quantitative imaging features were extracted and filtered by correlation analysis and Random forest Ranking to yield 40 candidate features. Fifteen machine learning methods were evaluated, and a ten-fold cross-validated elastic net regression model was selected to construct the radiomics-based prediction model. A clinical model based on five key clinical variables and a combined model integrating imaging and clinical features were also developed. The radiomics model achieved accuracies of 0.801, 0.866, and 0.831 in the training set and two external test sets, with AUC of 0.791, 0.829, and 0.807. In one external test set, the clinical model had an AUC of 0.689, significantly lower than the radiomics model (0.807, p < 0.05). The combined model achieved the highest performance, with AUC of 0.834 in the training set and 0.894 in an external test set (p < 0.01 and p < 0.001, respectively). Interpretability analysis revealed that wavelet-transformed features dominated the model, with the highest contribution from a feature reflecting small high-intensity clusters within the tumor and the second highest from a feature representing low-intensity clusters in the six-voxel peritumoral region. Kaplan-Meier analysis demonstrated that patients with either pathologically confirmed or model-predicted spread had significantly shorter progression-free survival (p < 0.001). Our novel machine learning model, integrating imaging features from both tumor and peritumoral regions, preoperatively predicts tumor spread through air spaces in stage I lung adenocarcinoma. It outperforms traditional clinical models, highlighting the potential of quantitative imaging analysis in personalizing treatment. Future prospective studies and further optimization are warranted.

The limited availability of annotated data presents a major challenge for applying deep learning methods to medical image analysis. Few-shot learning methods aim to recognize new classes from only a small number of labeled examples. These methods are typically studied under the standard few-shot learning setting, where all classes in a task are new. However, medical applications such as pathology classification from chest X-rays often require learning new classes while simultaneously leveraging knowledge of previously known ones, a scenario more closely aligned with generalized few-shot classification. Despite its practical relevance, few-shot learning has been scarcely studied in this context. In this work, we present MetaChest, a large-scale dataset of 479,215 chest X-rays collected from four public databases. MetaChest includes a meta-set partition specifically designed for standard few-shot classification, as well as an algorithm for generating multi-label episodes. We conduct extensive experiments evaluating both a standard transfer learning approach and an extension of ProtoNet across a wide range of few-shot multi-label classification tasks. Our results demonstrate that increasing the number of classes per episode and the number of training examples per class improves classification performance. Notably, the transfer learning approach consistently outperforms the ProtoNet extension, despite not being tailored for few-shot learning. We also show that higher-resolution images improve accuracy at the cost of additional computation, while efficient model architectures achieve comparable performance to larger models with significantly reduced resource requirements.

Sepsis, a severe complication of infection, often leads to acute kidney injury (AKI), which significantly increases the risk of death. Despite its clinical importance, early prediction of AKI remains challenging. Current tools rely on blood and urine tests, which are costly, variable, and not always available in time for intervention. Pneumonia is the most common cause of sepsis, accounting for over one-third of cases. In such patients, pulmonary inflammation and perilesional tissue alterations may serve as surrogate markers of systemic disease progression. However, these imaging features are rarely used in clinical decision-making. To overcome this limitation, our study aims to extract informative imaging features from pneumonia-associated sepsis cases using deep learning, with the goal of predicting the development of AKI. This dual-center retrospective study included pneumonia-associated sepsis patients (Jan 2020-Jul 2024). Chest CT images, clinical records, and laboratory data at admission were collected. We propose MCANet (Multimodal Cross-Attention Network), a two-stage deep learning framework designed to predict the occurrence of pneumonia-associated sepsis-related acute kidney injury (pSA-AKI). In the first stage, region-specific features were extracted from the lungs, epicardial adipose tissue, and T4-level subcutaneous adipose tissue using ResNet-18, which was chosen for its lightweight architecture and efficiency in processing multi-regional 2D CT slices with low computational cost. In the second stage, the extracted features were fused via a Multiscale Feature Attention Network (MSFAN) employing cross-attention mechanisms to enhance interactions among anatomical regions, followed by classification using ResNet-101, selected for its deeper architecture and strong ability to model global semantic representations and complex patterns.Model performance was evaluated using AUC, accuracy, precision, recall, and F1-score. Grad-CAM and PyRadiomics were employed for visual interpretation and radiomic analysis, respectively. A total of 399 patients with pneumonia-associated sepsis were included in this study. The modality ablation experiments demonstrated that the model integrating features from the lungs, T4-level subcutaneous adipose tissue, and epicardial adipose tissue achieved the best performance, with an accuracy of 0.981 and an AUC of 0.99 on the external test set from an independent center. For the prediction of AKI onset time, the LightGBM model incorporating imaging and clinical features achieved the highest accuracy of 0.8409 on the external test set. Furthermore, the multimodal model combining deep features, radiomics features, and clinical data further improved predictive performance, reaching an accuracy of 0.9773 and an AUC of 0.961 on the external test set. This study developed MCAnet, a multimodal deep learning framework that integrates imaging features from the lungs, epicardial adipose tissue, and T4-level subcutaneous adipose tissue. The framework significantly improved the accuracy of AKI occurrence and temporal prediction in pneumonia-associated sepsis patients, highlighting the synergistic role of adipose tissue and lung characteristics. Furthermore, explainability analysis revealed potential decision-making mechanisms underlying the temporal progression of pSA-AKI, offering new insights for clinical management.