This study aimed to develop and validate deep learning models using [¹⁸F]FDG PET/CT to predict PD-L1 status in esophageal cancer (EC) patients. Additionally, we assessed the potential of derived deep learning model scores (DLS) for survival stratification in immunotherapy. In this retrospective study, we included 331 EC patients from two centers, dividing them into training, internal validation, and external validation cohorts. Fifty patients who received immunotherapy were followed up. We developed four 3D ResNet10-based models-PET + CT + clinical factors (CPC), PET + CT (PC), PET (P), and CT (C)-using pre-treatment [¹⁸F]FDG PET/CT scans. For comparison, we also constructed a logistic model incorporating clinical factors (clinical model). The DLS were evaluated as radiological markers for survival stratification, and nomograms for predicting survival were constructed. The models demonstrated accurate prediction of PD-L1 status. The areas under the curve (AUCs) for predicting PD-L1 status were as follows: CPC (0.927), PC (0.904), P (0.886), C (0.934), and the clinical model (0.603) in the training cohort; CPC (0.882), PC (0.848), P (0.770), C (0.745), and the clinical model (0.524) in the internal validation cohort; and CPC (0.843), PC (0.806), P (0.759), C (0.667), and the clinical model (0.671) in the external validation cohort. The CPC and PC models exhibited superior predictive performance. Survival analysis revealed that the DLS from most models effectively stratified overall survival and progression-free survival at appropriate cut-off points (P < 0.05), outperforming stratification based on PD-L1 status (combined positive score ≥ 10). Furthermore, incorporating model scores with clinical factors in nomograms enhanced the predictive probability of survival after immunotherapy. Deep learning models based on [¹⁸F]FDG PET/CT can accurately predict PD-L1 status in esophageal cancer patients. The derived DLS can effectively stratify survival outcomes following immunotherapy, particularly when combined with clinical factors.

Fractional flow reserve (FFR) - a physiological indicator of coronary stenosis significance - has now become a widely used parameter also in the guidance of percutaneous coronary intervention (PCI). Several studies have shown the superiority of FFR compared to visual assessment, contributing to the reduction in clinical endpoints. However, the current approach to FFR assessment requires coronary instrumentation with a dedicated pressure wire and thus increasing invasiveness, cost, and duration of the procedure. Alternative, noninvasive methods of FFR assessment based on computational fluid dynamics are being widely tested; these approaches are generally not fully automated and may sometimes require substantial computational power. Nowadays, one of the most rapidly expanding fields in medicine is the use of artificial intelligence (AI) in therapy optimization, diagnosis, treatment, and risk stratification. AI usage contributes to the development of more sophisticated methods of imaging analysis and allows for the derivation of clinically important parameters in a faster and more accurate way. Over the recent years, AI utility in deriving FFR in a noninvasive manner has been increasingly reported. In this review, we critically summarize current knowledge in the field of AI-derived FFR based on data from computed tomography angiography, invasive angiography, optical coherence tomography, and intravascular ultrasound. Available solutions, possible future directions in optimizing cathlab performance, including the use of mixed reality, as well as current limitations standing behind the wide adoption of these techniques, are overviewed.

Spinal cord injury (SCI) in the pediatric population presents a unique challenge in diagnosis and prognosis due to the complexity of performing clinical assessments on children. Accurate evaluation of structural changes in the spinal cord is essential for effective treatment planning. This study aims to evaluate structural characteristics in pediatric patients with SCI by comparing cross-sectional area (CSA), anterior-posterior (AP) width, and right-left (RL) width across all vertebral levels of the spinal cord between typically developing (TD) and participants with SCI. We employed deep learning techniques to utilize these measures for detecting SCI cases and determining their injury severity. Sixty-one pediatric participants (ages 6-18), including 20 with chronic SCI and 41 TD, were enrolled and scanned by using a 3T MRI scanner. All SCI participants underwent the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) test to assess their neurologic function and determine their American Spinal Injury Association (ASIA) Impairment Scale (AIS) category. T2-weighted MRI scans were utilized to measure CSA, AP width, and RL widths along the entire cervical and thoracic cord. These measures were automatically extracted at every vertebral level of the spinal cord by using the spinal cord toolbox. Deep convolutional neural networks (CNNs) were utilized to classify participants into SCI or TD groups and determine their AIS classification based on structural parameters and demographic factors such as age and height. Significant differences (<i>P</i> < .05) were found in CSA, AP width, and RL width between SCI and TD participants, indicating notable structural alterations due to SCI. The CNN-based models demonstrated high performance, achieving 96.59% accuracy in distinguishing SCI from TD participants. Furthermore, the models determined AIS category classification with 94.92% accuracy. The study demonstrates the effectiveness of integrating cross-sectional structural imaging measures with deep learning methods for classification and severity assessment of pediatric SCI. The deep learning approach outperforms traditional machine learning models in diagnostic accuracy, offering potential improvements in patient care in pediatric SCI management.

Deep learning models often struggle to maintain generalizability in medical imaging, particularly under domain-fracture scenarios where distribution shifts arise from varying imaging techniques, acquisition protocols, patient populations, demographics, and equipment. In practice, each hospital may need to train distinct models - differing in learning task, width, and depth - to match local data. For example, one hospital may use Euclidean architectures such as MLPs and CNNs for tabular or grid-like image data, while another may require non-Euclidean architectures such as graph neural networks (GNNs) for irregular data like brain connectomes. How to train such heterogeneous models coherently across datasets, while enhancing each model's generalizability, remains an open problem. We propose unified learning, a new paradigm that encodes each model into a graph representation, enabling unification in a shared graph learning space. A GNN then guides optimization of these unified models. By decoupling parameters of individual models and controlling them through a unified GNN (uGNN), our method supports parameter sharing and knowledge transfer across varying architectures (MLPs, CNNs, GNNs) and distributions, improving generalizability. Evaluations on MorphoMNIST and two MedMNIST benchmarks - PneumoniaMNIST and BreastMNIST - show that unified learning boosts performance when models are trained on unique distributions and tested on mixed ones, demonstrating strong robustness to unseen data with large distribution shifts. Code and benchmarks: https://github.com/basiralab/uGNN

This study develops a machine learning model using dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) radiomics and clinical data to preoperatively differentiate hepatocellular carcinoma (HCC), intrahepatic cholangiocarcinoma (ICC), and hepatic inflammatory pseudotumor (HIPT), addressing limitations of conventional diagnostics. This retrospective study included 280 patients (HCC = 160, ICC = 80, HIPT = 40) who underwent DCE-MRI from 2008 to 2024 at three hospitals. Radiomics features and clinical data were extracted and analyzed using LASSO regression and machine learning algorithms (Logistic Regression, Random Forest, and Extreme Gradient Boosting), with class weighting (HCC:ICC:HIPT = 1:2:4) to address class imbalance. Models were compared using macro-average Area Under the Curve (AUC), accuracy, recall, and precision. The fusion model, integrating radiomics and clinical features, achieved an AUC of 0.933 (95% CI: 0.91-0.95) and 84.5% accuracy, outperforming radiomics-only (AUC = 0.856, 72.6%) and clinical-only (AUC = 0.795, 66.7%) models (p < 0.05). Rim enhancement is a key model feature for distinguishing HCC from ICC and HIPT, while hepatic lobe atrophy distinguishes ICC and HIPT from HCC. This study developed a novel preoperative imaging-based model to differentiate HCC, ICC, and HIPT. The fusion model performed exceptionally well, demonstrating superior accuracy in ICC identification, significantly outperforming traditional diagnostic methods (e.g., radiology and biomarkers) and single-modality machine learning models (p < 0.05). This noninvasive approach enhances diagnostic precision and supports personalized treatment planning in liver disease management. This study develops a novel preoperative imaging-based machine learning model to differentiate hepatocellular carcinoma (HCC), intrahepatic cholangiocarcinoma (ICC), and hepatic inflammatory pseudotumor (HIPT), improving diagnostic accuracy and advancing personalized treatment strategies in clinical radiology. A machine learning model integrates DCE-MRI radiomics and clinical data for liver lesion differentiation. The fusion model outperforms single-modality models with 0.933 AUC and 84.5% accuracy. This model provides a noninvasive, reliable tool for personalized liver disease diagnosis and treatment planning.

To construct a multimodal ultrasound (US) radiomics model for predicting axillary lymph node metastasis (ALNM) in breast cancer and evaluated its application value in predicting ALNM and patient prognosis. From March 2014 to December 2022, data from 682 breast cancer patients from four hospitals were collected, including preoperative grayscale US, color Doppler flow imaging (CDFI), contrast-enhanced ultrasound (CEUS) imaging data, and clinical information. Data from the First Medical Center of PLA General Hospital were used as the training and internal validation sets, while data from Peking University First Hospital, the Cancer Hospital of the Chinese Academy of Medical Sciences, and the Fourth Medical Center of PLA General Hospital were used as the external validation set. LASSO regression was employed to select radiomic features (RFs), while eight machine learning algorithms were utilized to construct radiomic models based on US, CDFI, and CEUS. The prediction efficiency of ALNM was assessed to identify the optimal model. In the meantime, Radscore was computed and integrated with immunoinflammatory markers to forecast Disease-Free Survival (DFS) in breast cancer patients. Follow-up methods included telephone outreach and in-person hospital visits. The analysis employed Cox regression to pinpoint prognostic factors, while clinical-imaging models were developed accordingly. The performance of the model was evaluated using the C-index, Receiver Operating Characteristic (ROC) curves, calibration curves, and Decision Curve Analysis (DCA). In the training cohort (n = 400), 40% of patients had ALNM, with a mean age of 55 ± 10 years. The US + CDFI + CEUS-based radiomics model achieved Area Under the Curves (AUCs) of 0.88, 0.81, and 0.77 for predicting N0 versus N+ (≥ 1) in the training, internal, and external validation sets, respectively, outperforming the US-only model (P < 0.05). For distinguishing N+ (1-2) from N+ (≥ 3), the model achieved AUCs of 0.89, 0.74, and 0.75. Combining radiomics scores with clinical immunoinflammatory markers (platelet count and neutrophil-to-lymphocyte ratio) yielded a clinical-radiomics model predicting disease-free survival (DFS), with C-indices of 0.80, 0.73, and 0.79 across the three cohorts. In the external validation cohort, the clinical-radiomics model achieved higher AUCs for predicting 2-, 3-, and 5-year DFS compared to the clinical model alone (2-year: 0.79 vs. 0.66; 3-year: 0.83 vs. 0.70; 5-year: 0.78 vs. 0.64; all P < 0.05). Calibration and decision curve analyses demonstrated good model agreement and clinical utility. The multimodal ultrasound radiomics model based on US, CDFI, and CEUS could effectively predict ALNM in breast cancer. Furthermore, the combined application of radiomics and immune inflammation markers might predict the DFS of breast cancer patients to some extent.

Subepithelial lesions (SELs) present significant diagnostic challenges in gastrointestinal endoscopy, particularly in differentiating malignant types, such as gastrointestinal stromal tumors (GISTs) and neuroendocrine tumors, from benign types like leiomyomas. Misdiagnosis can lead to unnecessary interventions or delayed treatment. To address this challenge, we developed ECMAI-WME, a parallel fusion deep learning model integrating white light endoscopy (WLE) and microprobe endoscopic ultrasonography (EUS), to improve SEL classification and lesion characterization. A total of 523 SELs from four hospitals were used to develop serial and parallel fusion AI models. The Parallel Model, demonstrating superior performance, was designated as ECMAI-WME. The model was tested on an external validation cohort (n = 88) and a multicenter test cohort (n = 274). Diagnostic performance, lesion characterization, and clinical decision-making support were comprehensively evaluated and compared with endoscopists' performance. The ECMAI-WME model significantly outperformed endoscopists in diagnostic accuracy (96.35% vs. 63.87-86.13%, p < 0.001) and treatment decision-making accuracy (96.35% vs. 78.47-86.13%, p < 0.001). It achieved 98.72% accuracy in internal validation, 94.32% in external validation, and 96.35% in multicenter testing. For distinguishing gastric GISTs from leiomyomas, the model reached 91.49% sensitivity, 100% specificity, and 96.38% accuracy. Lesion characteristics were identified with a mean accuracy of 94.81% (range: 90.51-99.27%). The model maintained robust performance despite class imbalance, confirmed by five complementary analyses. Subgroup analyses showed consistent accuracy across lesion size, location, or type (p > 0.05), demonstrating strong generalizability. The ECMAI-WME model demonstrates excellent diagnostic performance and robustness in the multiclass SEL classification and characterization, supporting its potential for real-time deployment to enhance diagnostic consistency and guide clinical decision-making.

Despite the potential clinical utility for acute ischemic stroke patients, predicting short-term operational outcomes like length of stay (LOS) and long-term functional outcomes such as the 90-day Modified Rankin Scale (mRS) remain a challenge, with limited current clinical guidance on expected patient trajectories. Machine learning approaches have increasingly aimed to bridge this gap, often utilizing admission-based clinical features; yet, the integration of imaging biomarkers remains underexplored, especially regarding whole 2.5D image fusion using advanced deep learning techniques. This study introduces a novel method leveraging autoencoders to integrate 2.5D diffusion weighted imaging (DWI) with clinical features for refined outcome prediction. Results on a comprehensive dataset of AIS patients demonstrate that our autoencoder-based method has comparable performance to traditional convolutional neural networks image fusion methods and clinical data alone (LOS > 8 days: AUC 0.817, AUPRC 0.573, F1-Score 0.552; 90-day mRS > 2: AUC 0.754, AUPRC 0.685, F1-Score 0.626). This novel integration of imaging and clinical data for post-intervention stroke prognosis has numerous computational and operational advantages over traditional image fusion methods. While further validation of the presented models is necessary before adoption, this approach aims to enhance personalized patient management and operational decision-making in healthcare settings. Not applicable.

Stroke is a leading cause of death and long-term disability worldwide, and early detection remains a significant clinical challenge. This study proposes an Effective Brain Stroke Diagnosis Strategy (EBDS). The hybrid deep learning framework integrates Vision Transformer (ViT) and VGG16 to enable accurate and interpretable stroke detection from CT images. The model was trained and evaluated using a publicly available dataset from Kaggle, achieving impressive results: a test accuracy of 99.6%, a precision of 1.00 for normal cases and 0.98 for stroke cases, a recall of 0.99 for normal cases and 1.00 for stroke cases, and an overall F1-score of 0.99. These results demonstrate the robustness and reliability of the EBDS model, which outperforms several recent state-of-the-art methods. To enhance clinical trust, the model incorporates explainability techniques, such as Grad-CAM and LIME, which provide visual insights into its decision-making process. The EBDS framework is designed for real-time application in emergency settings, offering both high diagnostic performance and interpretability. This work addresses a critical research gap in early brain stroke diagnosis and contributes a scalable, explainable, and clinically relevant solution for medical imaging diagnostics.

Identifying syndesmotic instability in ankle fractures using conventional radiographs is still a major challenge. In this study we trained a convolutional neural network (CNN) to classify the fracture utilizing the AO-classification (AO-44 A/B/C) and to simultaneously detect syndesmosis instability in the conventional radiograph by leveraging the intraoperative stress testing as the gold standard. In this retrospective exploratory study we identified 700 patients with rotational ankle fractures at a university hospital from 2019 to 2024, from whom 1588 digital radiographs were extracted to train, validate, and test a CNN. Radiographs were classified based on the therapy-decisive gold standard of the intraoperative hook-test and the preoperatively determined AO-classification from the surgical report. To perform internal validation and quality control, the algorithm results were visualized using Guided Score Class activation maps (GSCAM).The AO44-classification sensitivity over all subclasses was 91%. Furthermore, the syndesmosis instability could be identified with a sensitivity of 0.84 (95% confidence interval (CI) 0.78, 0.92) and specificity 0.8 (95% CI 0.67, 0.9). Consistent visualization results were obtained from the GSCAMs. The integration of an explainable deep-learning algorithm, trained on an intraoperative gold standard showed a 0.84 sensitivity for syndesmotic stability testing. Thus, providing clinically interpretable outputs, suggesting potential for enhanced preoperative decision-making in complex ankle trauma.