We aimed to develop and validate multimodal models integrating computed tomography (CT) images, text and tabular clinical data to predict poor functional outcomes and in-hospital mortality in patients with intracerebral hemorrhage (ICH). These models were designed to assist non-specialists in emergency settings with limited access to stroke specialists. A retrospective analysis of 527 patients with ICH admitted to a Japanese tertiary hospital between April 2019 and February 2022 was conducted. Deep learning techniques were used to extract features from three-dimensional CT images and unstructured data, which were then combined with tabular data to develop an L1-regularized logistic regression model to predict poor functional outcomes (modified Rankin scale score 3-6) and in-hospital mortality. The model's performance was evaluated by assessing discrimination metrics, calibration plots, and decision curve analysis (DCA) using temporal validation data. The multimodal model utilizing both imaging and text data, such as medical interviews, exhibited the highest performance in predicting poor functional outcomes. In contrast, the model that combined imaging with tabular data, including physiological and laboratory results, demonstrated the best predictive performance for in-hospital mortality. These models exhibited high discriminative performance, with areas under the receiver operating curve (AUROCs) of 0.86 (95% CI: 0.79-0.92) and 0.91 (95% CI: 0.84-0.96) for poor functional outcomes and in-hospital mortality, respectively. Calibration was satisfactory for predicting poor functional outcomes, but requires refinement for mortality prediction. The models performed similar to or better than conventional risk scores, and DCA curves supported their clinical utility. Multimodal prediction models have the potential to aid non-specialists in making informed decisions regarding ICH cases in emergency departments as part of clinical decision support systems. Enhancing real-world data infrastructure and improving model calibration are essential for successful implementation in clinical practice.

Image segmentation plays a central role in computer vision applications such as medical imaging, industrial inspection, and environmental monitoring. However, evaluating segmentation performance can be particularly challenging when ground truth is not clearly defined, as is often the case in tasks involving subjective interpretation. These challenges are amplified by inter- and intra-observer variability, which complicates the use of human annotations as a reliable reference. To address this, we propose a novel validation framework-referred to as the three-blind validation strategy-that enables rigorous assessment of segmentation models in contexts where subjectivity and label variability are significant. The core idea is to have a third independent expert, blind to the labeler identities, assess a shuffled set of segmentations produced by multiple human annotators and/or automated models. This allows for the unbiased evaluation of model performance and helps uncover patterns of disagreement that may indicate systematic issues with either human or machine annotations. The primary objective of this study is to introduce and demonstrate this validation strategy as a generalizable framework for robust model evaluation in subjective segmentation tasks. We illustrate its practical implementation in a mammography use case involving dense tissue segmentation while emphasizing its potential applicability to a broad range of segmentation scenarios.

Desmoid tumors are rare, locally invasive soft-tissue tumors with unpredictable clinical behavior. Imaging plays a crucial role in their diagnosis, measurement of disease burden, and assessment of treatment response. However, desmoid tumors' unique imaging features present challenges to conventional imaging metrics. The heterogeneous nature of these tumors, with a variable composition (fibrous, myxoid, or cellular), complicates accurate delineation of tumor boundaries and volumetric assessment. Furthermore, desmoid tumors can demonstrate prolonged stability or spontaneous regression, and biologic quiescence is often manifested by collagenization rather than bulk size reduction, making traditional size-based response criteria, such as Response Evaluation Criteria in Solid Tumors (RECIST), suboptimal. To overcome these limitations, advanced imaging techniques offer promising opportunities. Functional and parametric imaging methods, such as diffusion-weighted MRI, dynamic contrast-enhanced MRI, and T2 relaxometry, can provide insights into tumor cellularity and maturation. Radiomics and artificial intelligence approaches may enhance quantitative analysis by extracting and correlating complex imaging features with biological behavior. Moreover, imaging biomarkers could facilitate earlier detection of treatment efficacy or resistance, enabling tailored therapy. By integrating advanced imaging into clinical practice, it may be possible to refine the evaluation of disease burden and treatment response, ultimately improving the management and outcomes of patients with desmoid tumors.

Tuberous breast deformity (TBD) is a congenital condition characterized by constriction of the breast base, parenchymal hypoplasia, and areolar herniation. The absence of a universally accepted classification system complicates diagnosis and surgical planning, leading to variability in clinical outcomes. Artificial intelligence (AI) has emerged as a powerful adjunct in medical imaging, enabling objective, reproducible, and data-driven diagnostic assessments. This study introduces an AI-driven diagnostic tool for tuberous breast deformity (TBD) classification using a Siamese Network trained on paired frontal and lateral images. Additionally, the model generates a continuous Tuberosity Score (ranging from 0 to 1) based on embedding vector distances, offering an objective measure to enhance surgical planning and improved clinical outcomes. A dataset of 200 expertly classified frontal and lateral breast images (100 tuberous, 100 non-tuberous) was used to train a Siamese Network with contrastive loss. The model extracted high-dimensional feature embeddings to differentiate tuberous from non-tuberous breasts. Five-fold cross-validation ensured robust performance evaluation. Performance metrics included accuracy, precision, recall, and F1-score. Visualization techniques, such as t-SNE clustering and occlusion sensitivity mapping, were employed to interpret model decisions. The model achieved an average accuracy of 96.2% ± 5.5%, with balanced precision and recall. The Tuberosity Score, derived from the Euclidean distance between embeddings, provided a continuous measure of deformity severity, correlating well with clinical assessments. This AI-based framework offers an objective, high-accuracy classification system for TBD. The Tuberosity Score enhances diagnostic precision, potentially aiding in surgical planning and improving patient outcomes.

Deep learning methods provide enormous promise in automating manually intense tasks such as medical image segmentation and provide workflow assistance to clinical experts. Deep neural networks (DNN) require a significant amount of training examples and a variety of expert opinions to capture the nuances and the context, a challenging proposition in oncological studies (H. Wang et al., Nature, vol. 620, no. 7972, pp. 47-60, Aug 2023). Inter-reader variability among clinical experts is a real-world problem that severely impacts the generalization of DNN reproducibility. This study proposes quantifying the variability in DNN performance using expert opinions and exploring strategies to train the network and adapt between expert opinions. We address the inter-reader variability problem in the context of prostate gland segmentation using a well-studied DNN, the 3D U-Net model. Reference data includes magnetic resonance imaging (MRI, T2-weighted) with prostate glandular anatomy annotations from two expert readers (R#1, n = 342 and R#2, n = 204). 3D U-Net was trained and tested with individual expert examples (R#1 and R#2) and had an average Dice coefficient of 0.825 (CI, [0.81 0.84]) and 0.85 (CI, [0.82 0.88]), respectively. Combined training with a representative cohort proportion (R#1, n = 100 and R#2, n = 150) yielded enhanced model reproducibility across readers, achieving an average test Dice coefficient of 0.863 (CI, [0.85 0.87]) for R#1 and 0.869 (CI, [0.87 0.88]) for R#2. We re-evaluated the model performance across the gland volumes (large, small) and found improved performance for large gland size with an average Dice coefficient to be at 0.846 [CI, 0.82 0.87] and 0.872 [CI, 0.86 0.89] for R#1 and R#2, respectively, estimated using fivefold cross-validation. Performance for small gland sizes diminished with average Dice of 0.8 [0.79, 0.82] and 0.8 [0.79, 0.83] for R#1 and R#2, respectively.

Glioblastoma (GBM) is a highly aggressive brain tumor with poor prognosis. This study aimed to construct and validate a radiomics-based machine learning model for predicting overall survival (OS) in IDH-wildtype GBM after maximal safe surgical resection using magnetic resonance imaging. A total of 582 patients were retrospectively enrolled, comprising 301 in the training cohort, 128 in the internal validation cohort, and 153 in the external validation cohort. Volumes of interest (VOIs) from contrast-enhanced T1-weighted imaging (CE-T1WI) were segmented into three regions: contrast-enhancing tumor, necrotic non-enhancing core, and peritumoral edema using an ResNet-based segmentation network. A total of 4,227 radiomic features were extracted and filtered using LASSO-Cox regression to identify signatures. The prognostic model was constructed using the Mime prediction framework, categorizing patients into high- and low-risk groups based on the median OS. Model performance was assessed using the concordance index (CI) and Kaplan-Meier survival analysis. Independent prognostic factors were identified through multivariable Cox regression analysis, and a nomogram was developed for individualized risk assessment. The Step Cox [backward] + RSF model achieved CIs of 0.89, 0.81, and 0.76 in the training, internal and external validation cohorts. Log-rank tests demonstrated significant survival differences between high- and low-risk groups across all cohorts (P < 0.05). Multivariate Cox analysis identified age (HR: 1.022; 95% CI: 0.979, 1.009, P < 0.05), KPS score (HR: 0.970, 95% CI: 0.960, 0.978, P < 0.05), rad-scores of the necrotic non-enhancing core (HR: 8.164; 95% CI: 2.439, 27.331, P < 0.05), and peritumoral edema (HR: 3.748; 95% CI: 1.212, 11.594, P < 0.05) as independent predictors of OS. A nomogram integrating these predictors provided individualized risk assessment. This deep learning segmentation-based radiomics model demonstrated robust performance in predicting OS in GBM after maximal safe surgical resection. By incorporating radiomic signatures and advanced machine learning algorithms, it offers a non-invasive tool for personalized prognostic assessment and supports clinical decision-making.

The clinical benefits of neoadjuvant chemoimmunotherapy (NACI) are demonstrated in patients with bladder cancer (BCa); however, more than half fail to achieve a pathological complete response (pCR). This study utilizes multi-center cohorts of 2322 patients with pathologically diagnosed BCa, collected between January 1, 2014, and December 31, 2023, to explore the correlation between tumor budding (TB) status and NACI response and disease prognosis. A deep learning model is developed to noninvasively evaluate TB status based on CT images. The deep learning model accurately predicts the TB status, with area under the curve values of 0.932 (95% confidence interval: 0.898-0.965) in the training cohort, 0.944 (0.897-0.991) in the internal validation cohort, 0.882 (0.832-0.933) in external validation cohort 1, 0.944 (0.908-0.981) in the external validation cohort 2, and 0.854 (0.739-0.970) in the NACI validation cohort. Patients predicted to have a high TB status exhibit a worse prognosis (p < 0.05) and a lower pCR rate of 25.9% (7/20) than those predicted to have a low TB status (pCR rate: 73.9% [17/23]; p < 0.001). Hence, this model may be a reliable, noninvasive tool for predicting TB status, aiding clinicians in prognosis assessment and NACI strategy formulation.

The accurate classification of lung nodules is critical to achieving personalized lung cancer treatment and prognosis prediction. The treatment options for lung cancer and the prognosis of patients are closely related to the type of lung nodules, but there are many types of lung nodules, and the distinctions between certain types are subtle, making accurate classification based on traditional medical imaging technology and doctor experience challenging. In this study, a novel method was used to analyze quantitative features in CT images using CT radiomics to reveal the characteristics of pulmonary nodules, and then feature fusion was used to integrate radiomics features and deep learning features to improve the accuracy of classification. This paper proposes a fusion feature pulmonary nodule classification method that fuses radiomics features with deep learning neural network features, aiming to automatically classify different types of pulmonary nodules (such as Malignancy, Calcification, Spiculation, Lobulation, Margin, and Texture). By introducing the Discriminant Correlation Analysis feature fusion algorithm, the method maximizes the complementarity between the two types of features and the differences between different classes. This ensures interaction between the information, effectively utilizing the complementary characteristics of the features. The LIDC-IDRI dataset is used for training, and the fusion feature model has been validated for its advantages and effectiveness in classifying multiple types of pulmonary nodules. The experimental results show that the fusion feature model outperforms the single-feature model in all classification tasks. The AUCs for the tasks of classifying Calcification, Lobulation, Margin, Spiculation, Texture, and Malignancy reached 0.9663, 0.8113, 0.8815, 0.8140, 0.9010, and 0.9316, respectively. In tasks such as nodule calcification and texture classification, the fusion feature model significantly improved the recognition ability of minority classes. The fusion of radiomics features and deep learning neural network features can effectively enhance the overall performance of pulmonary nodule classification models while also improving the recognition of minority classes when there is a significant class imbalance.

The assessment of lymphovascular invasion (LVI) is crucial in the management of rectal cancer; However, accurately evaluating LVI preoperatively using imaging remains challenging. Recent advances in radiomics have created opportunities for developing more accurate diagnostic tools. This study aimed to develop and validate a deep learning model for predicting LVI in rectal cancer patients using preoperative MR imaging. These cases were randomly divided into a training cohort (n = 233) and an validation cohort (n = 101) at a ratio of 7:3. Based on the pathological reports, the patients were classified into positive and negative groups according to their LVI status. Based on the preoperative MRI T2WI axial images, the regions of interest (ROI) were defined from the tumor itself and the edges of the tumor extending outward by 5 pixels, 10 pixels, 15 pixels, and 20 pixels. The 2D and 3D deep learning features were extracted using the DenseNet121 architecture, and the deep learning models were constructed, including a total of ten models: GTV (the tumor itself), GPTV5 (the tumor itself and the tumor extending outward by 5 pixels), GPTV10, GPTV15, and GPTV20. To assess model performance, we utilized the area under the curve (AUC) and conducted DeLong test to compare different models, aiming to identify the optimal model for predicting LVI in rectal cancer. In the 2D deep learning model group, the 2D GPTV10 model demonstrated superior performance with an AUC of 0.891 (95% confidence interval [CI] 0.850-0.933) in the training cohort and an AUC of 0.841 (95% CI 0.767-0.915) in the validation cohort. The difference in AUC between this model and other 2D models was not statistically significant based on DeLong test (p > 0.05); In the group of 3D deep learning models, the 3D GPTV10 model had the highest AUC, with a training cohort AUC of 0.961 (95% CI 0.940-0.982) and a validation cohort AUC of 0.928 (95% CI 0.881-0.976). DeLong test demonstrated that the performance of the 3D GPTV10 model surpassed other 3D models as well as the 2D GPTV10 model (p < 0.05). The study developed a deep learning model, namely 3D GPTV10, utilizing preoperative MRI data to accurately predict the presence of LVI in rectal cancer patients. By training on the tumor itself and its surrounding margin 10 pixels as the region of interest, this model achieved superior performance compared to other deep learning models. These findings have significant implications for clinicians in formulating personalized treatment plans for rectal cancer patients.