Due to the complex anatomical structure and dynamic involution process of the thymus, segmentation and evaluation of the thymus in medical imaging present significant challenges. The aim of this study is to develop a deep-learning tool "Thy-uNET" for automatic segmentation and measurement of the thymus or thymic region on chest CT imaging, and to validate its performance with multicenter data. Utilizing the segmentation and measurement results from two experts, training of Thy-uNET was conducted on training cohort (n = 500). The segmented regions include thymus or thymic region, and 7 features of the thymic region were measured. The automatic segmentation performance was assessed using Dice and Intersection over Union (IOU) on CT data from three test cohorts (n = 286). Spearman correlation analysis and intraclass correlation coefficient (ICC) were used to evaluate the correlation and reliability of the automatic measurement results. Six radiologists with varying levels of experience were invited to participate in a reader study to assess the measurement performance of Thy-uNET and its ability to assist doctors. Thy-uNET demonstrated consistent segmentation performance across different subgroups, with Dice = 0.83 in the internal test set, and Dice = 0.82 in the external test sets. For automatic measurement of thymic features, Thy-uNET achieved high correlation coefficients and ICC for key measurements (R = 0.829 and ICC = 0.841 for CT attenuation measurement). Its performance was comparable to that of radiology residents and junior radiologists, with significantly shorter measurement time. Providing Thy-uNET measurements to readers reduced their measurement time and improved residents' performance in some thymic feature measurements. Thy-uNET can provide reliable automatic segmentation and automatic measurement information of the thymus or thymic region on routine CT, reducing time costs and improving the consistency of evaluations.

We aimed to evaluate the diagnostic performance of artificial intelligence (AI) in detecting coronary artery stenosis and calcified plaque on CT angiography (CTA), comparing its diagnostic performance with that of radiologists. A thorough search of the literature was performed using PubMed, Web of Science, and Embase, focusing on studies published until October 2024. Studies were included if they evaluated AI models in detecting coronary artery stenosis and calcified plaque on CTA. A bivariate random-effects model was employed to determine combined sensitivity and specificity. Study heterogeneity was assessed using I² statistics. The risk of bias was assessed using the revised quality assessment of diagnostic accuracy studies-2 tool, and the evidence level was graded using the Grading of Recommendations Assessment, Development and Evalutiuon (GRADE) system. Out of 1071 initially identified studies, 17 studies with 5560 patients and images were ultimately included for the final analysis. For coronary artery stenosis ≥50%, AI showed a sensitivity of 0.92 (95% CI: 0.88-0.95), specificity of 0.87 (95% CI: 0.80-0.92), and AUC of 0.96 (95% CI: 0.94-0.97), outperforming radiologists with sensitivity of 0.85 (95% CI: 0.67-0.94), specificity of 0.84 (95% CI: 0.62-0.94), and AUC of 0.91 (95% CI: 0.89-0.93). For stenosis ≥70%, AI achieved a sensitivity of 0.88 (95% CI: 0.70-0.96), specificity of 0.96 (95% CI: 0.90-0.99), and AUC of 0.98 (95% CI: 0.96-0.99). In calcified plaque detection, AI demonstrated a sensitivity of 0.93 (95% CI: 0.84-0.97), specificity of 0.94 (95% CI: 0.88-0.96), and AUC of 0.98 (95% CI: 0.96-0.99)." AI-based CT demonstrated superior diagnostic performance compared to clinicians in identifying ≥50% stenosis in coronary arteries and showed excellent diagnostic performance in recognizing ≥70% coronary artery stenosis and calcified plaque. However, limitations include retrospective study designs and heterogeneity in CTA technologies. Further external validation through prospective, multicenter trials is required to confirm these findings. The original findings of this research are included in the article. For additional inquiries, please contact the corresponding authors.

This study aims to develop and validate an ultrasoundbased habitat imaging and peritumoral radiomics model for predicting high-risk capsule characteristics for recurrence of pleomorphic adenoma (PA) of the parotid gland while also exploring the optimal range of peritumoral region. Retrospective analysis was conducted on 325 patients (171 in training set, 74 in validation set and 80 in testing set) diagnosed with PA at two medical centers. Univariate and multivariate logistic regression analyses were performed to identify clinical risk factors. The tumor was segmented into four habitat subregions using K-means clustering, with peri-tumor regions expanded at thicknesses of 1/3/5mm. Radiomics features were extracted from intra-tumor, habitat subregions, and peritumoral regions respectively to construct predictive models, integrating three machine learning classifiers: SVM, RandomForest, and XGBoost. Additionally, a combined model was developed by incorporating peritumoral features and clinical factors based on habitat imaging. Model performance was evaluated using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA). SHAP analysis was employed to improve the interpretability. The RandomForest model in habitat imaging consistently outperformed other models in predictive performance, with AUC values of 0.881, 0.823, and 0.823 for the training set, validation set, and testing set respectively. Incorporating peri-1mm features and clinical factors into the combined model slightly improved its performance, resulting in AUC values of 0.898, 0.833, and 0.829 for each set. The calibration curves and DCA exhibited excellent fit for the combined model while providing great clinical net benefit. The combined model exhibits robust predictive performance in identifying high-risk capsule characteristics for recurrence of PA in the parotid gland. This model may assist in determining optimal surgical margin and assessing patients' prognosis.

Chest computed tomography (CT) radiomics can be utilized for categorical predictions; however, models predicting pulmonary function indices directly are lacking. This study aimed to develop machine-learning-based regression models to predict pulmonary function using chest CT radiomics. This retrospective study enrolled patients who underwent chest CT and pulmonary function tests between January 2018 and April 2024. Machine-learning regression models were constructed and validated to predict pulmonary function indices, including forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV₁). The models incorporated radiomics of the whole lung and clinical features. Model performance was evaluated using mean absolute error, mean squared error, root mean squared error, concordance correlation coefficient (CCC), and R-squared (R²) value and compared to spirometry results. Individual explanations of the models' decisions were analyzed using an explainable approach based on SHapley Additive exPlanations. In total, 1585 cases were included in the analysis, with 102 of them being external cases. Across the training, validation, test, and external test sets, the combined model consistently achieved the best performance in the regression task for predicting FVC (e.g. external test set: CCC, 0.745 [95% confidence interval 0.642-0.818]; R², 0.601 [0.453-0.707]) and FEV₁ (e.g. external test set: CCC, 0.744 [0.633-0.824]; R², 0.527 [0.298-0.675]). Age, sex, and emphysema were important factors for both FVC and FEV₁, while distinct radiomics features contributed to each. Whole-lung-based radiomics features can be used to construct regression models to improve pulmonary function prediction.

Diagnosing Parkinson's disease (PD) typically relies on clinical evaluations, often detecting it in advanced stages. Recently, artificial intelligence has increasingly been applied to imaging for neurodegenerative disorders. This study aims to develop a diagnostic prediction model using T1-weighted magnetic resonance imaging (T1-MRI) data from the caudate and putamen in individuals with early-stage PD. This retrospective case-control study included 69 early-stage PD patients and 22 controls, recruited through the Parkinson's Progression Markers Initiative. T1-MRI scans were acquired using a 3-tesla system. 432 radiomic features were extracted from images of the segmented caudate and putâmen in an automated way. Feature selection was performed using Pearson's correlation and recursive feature elimination to identify the most relevant variables. Three machine learning algorithms-random forest (RF), support vector machine and logistic regression-were evaluated for diagnostic prediction effectiveness using a cross-validation method. The Shapley Additive Explanations technique identified the most significant features distinguishing between the groups. The metrics used to evaluate the performance were discrimination, expressed in area under the ROC curve (AUC), sensitivity and specificity; and calibration, expressed as accuracy. The RF algorithm showed superior performance with an average accuracy of 92.85%, precision of 100.00%, sensitivity of 86.66%, specificity of 96.65% and AUC of 0.93. The three most influential features were contrast, elongation, and gray-level non-uniformity, all from the putamen. Machine learning-based models can differentiate early-stage PD from controls using T1-weighted MRI radiomic features.

This study constructed an interpretable machine learning model based on multi-parameter MRI sub-region habitat radiomics and clinicopathological features, aiming to preoperatively evaluate the microsatellite instability (MSI) status of rectal cancer (RC) patients. This retrospective study recruited 291 rectal cancer patients with pathologically confirmed MSI status and randomly divided them into a training cohort and a testing cohort at a ratio of 8:2. First, the K-means method was used for cluster analysis of tumor voxels, and sub-region radiomics features and classical radiomics features were respectively extracted from multi-parameter MRI sequences. Then, the synthetic minority over-sampling technique method was used to balance the sample size, and finally, the features were screened. Prediction models were established using logistic regression based on clinicopathological variables, classical radiomics features, and MSI-related sub-region radiomics features, and the contribution of each feature to the model decision was quantified by the Shapley-Additive-Explanations (SHAP) algorithm. The area under the curve (AUC) of the sub-region radiomics model in the training and testing groups was 0.848 and 0.8, respectively, both better than that of the classical radiomics and clinical models. The combined model performed the best, with AUCs of 0.908 and 0.863 in the training and testing groups, respectively. We developed and validated a robust combined model that integrates clinical variables, classical radiomics features, and sub-region radiomics features to accurately determine the MSI status of RC patients. We visualized the prediction process using SHAP, enabling more effective personalized treatment plans and ultimately improving RC patient survival rates.

Current ultrasound-based screening for endometrial cancer (EC) primarily relies on endometrial thickness (ET) and morphological evaluation, which suffer from low specificity and high interobserver variability. This study aimed to develop and validate an artificial intelligence (AI)-driven diagnostic model to improve diagnostic accuracy and reduce variability. A total of 1,861 consecutive postmenopausal women were enrolled from two centers between April 2021 and April 2024. Super-resolution (SR) technique was applied to enhance image quality before feature extraction. Radiomics features were extracted using Pyradiomics, and deep learning features were derived from convolutional neural network (CNN). Three models were developed: (1) R model: radiomics-based machine learning (ML) algorithms; (2) CNN model: image-based CNN algorithms; (3) DLR model: a hybrid model combining radiomics and deep learning features with ML algorithms. Using endometrium-level regions of interest (ROI), the DLR model achieved the best diagnostic performance, with an area under the receiver operating characteristic curve (AUROC) of 0.893 (95% CI: 0.847-0.932), sensitivity of 0.847 (95% CI: 0.692-0.944), and specificity of 0.810 (95% CI: 0.717-0.910) in the internal testing dataset. Consistent performance was observed in the external testing dataset (AUROC 0.871, sensitivity 0.792, specificity 0.829). The DLR model consistently outperformed both the R and CNN models. Moreover, endometrium-level ROIs yielded better results than uterine-corpus-level ROIs. This study demonstrates the feasibility and clinical value of AI-enhanced ultrasound analysis for EC detection. By integrating radiomics and deep learning features with SR-based image preprocessing, our model improves diagnostic specificity, reduces false positives, and mitigates operator-dependent variability. This non-invasive approach offers a more accurate and reliable tool for EC screening in postmenopausal women. Not applicable.

The aim of this study was to develop and validate a fully automatic anatomical landmark localization and trajectory planning method for external ventricular drain (EVD) placement using CT or MRI. The authors used 125 preoperative CT and 137 contrast-enhanced T1-weighted MRI scans to generate 3D surface meshes of patients' skin and ventricular systems. Seven anatomical landmarks were manually annotated to train a neural network for automatic landmark localization. The model's accuracy was assessed by calculating the mean Euclidian distance of predicted landmarks to the ground truth. Kocher's point and EVD trajectories were automatically calculated with the foramen of Monro as the target. Performance was evaluated using Kakarla grades, as assessed by 3 clinicians. Interobserver agreement was measured with Pearson correlation, and scores were aggregated using majority voting. Ordinal linear regressions were used to assess whether modality or placement side had an effect on Kakarla grades. The impact of landmark localization error on the final EVD plan was also evaluated. The automated landmark localization model achieved a mean error of 4.0 mm (SD 2.6 mm). Trajectory planning generated a trajectory for all patients, with a Kakarla grade of 1 in 92.9% of cases. Statistical analyses indicated a strong interobserver agreement and no significant differences between modalities (CT vs MRI) or EVD placement sides. The location of Kocher's point and the target point were significantly correlated to nasion landmark localization error, with median drifts of 9.38 mm (95% CI 1.94-19.16 mm) and 3.91 mm (95% CI 0.18-26.76 mm) for Kocher's point and the target point, respectively. The presented method was efficient and robust for landmark localization and accurate EVD trajectory planning. The short processing time thereby also provides a base for use in emergency settings.

To generate virtual T1 contrast-enhanced (T1CE) sequences from plain spinal MRI sequences using the denoising diffusion probabilistic model (DDPM) and to compare its performance against one baseline model pix2pix and three advanced models. A total of 1195 consecutive spinal tumor patients who underwent contrast-enhanced MRI at two hospitals were divided into a training set (n = 809, 49 ± 17 years, 437 men), an internal test set (n = 203, 50 ± 16 years, 105 men), and an external test set (n = 183, 52 ± 16 years, 94 men). Input sequences were T1- and T2-weighted images, and T2 fat-saturation images. The output was T1CE images. In the test set, one radiologist read the virtual images and marked all visible enhancing lesions. Results were evaluated using sensitivity (SE) and false discovery rate (FDR). We compared differences in lesion size and enhancement degree between reference and virtual images, and calculated signal-to-noise (SNR) and contrast-to-noise ratios (CNR) for image quality assessment. In the external test set, the mean squared error was 0.0038±0.0065, and structural similarity index 0.78±0.10. Upon evaluation by the reader, the overall SE of the generated T1CE images was 94% with FDR 2%. There was no difference in lesion size or signal intensity ratio between the reference and generated images. The CNR was higher in the generated images than the reference images (9.241 vs. 4.021; P＜0.001). The proposed DDPM demonstrates potential as an alternative to gadolinium contrast in spinal MRI examinations of oncologic patients.

<b>Objective</b>  The aim of this study was to explore an innovative approach for developing deep learning (DL) algorithm for renal cell carcinoma (RCC) detection and subtyping on computed tomography (CT): clear cell RCC (ccRCC) versus non-ccRCC using two-dimensional (2D) neural network architecture and feature consistency modules. <b>Materials and Methods</b>  This retrospective study included baseline CT scans from 196 histopathologically proven RCC patients: 143 ccRCCs and 53 non-ccRCCs. Manual tumor annotations were performed on axial slices of corticomedullary phase images, serving as ground truth. After image preprocessing, the dataset was divided into training, validation, and testing subsets. The study tested multiple 2D DL architectures, with the FocalNet-DINO demonstrating highest effectiveness in detecting and classifying RCC. The study further incorporated spatial and class consistency modules to enhance prediction accuracy. Models' performance was evaluated using free-response receiver operating characteristic curves, recall rates, specificity, accuracy, F1 scores, and area under the curve (AUC) scores. <b>Results</b>  The FocalNet-DINO architecture achieved the highest recall rate of 0.823 at 0.025 false positives per image (FPI) for RCC detection. The integration of spatial and class consistency modules into the architecture led to 0.2% increase in recall rate at 0.025 FPI, along with improvements of 0.1% in both accuracy and AUC scores for RCC classification. These enhancements allowed detection of cancer in an additional 21 slices and reduced false positives in 126 slices. <b>Conclusion</b>  This study demonstrates high performance for RCC detection and classification using DL algorithm leveraging 2D neural networks and spatial and class consistency modules, to offer a novel, computationally simpler, and accurate DL approach to RCC characterization.