This study aimed to develop a noninvasive Machine Learning (ML) model to identify clinically significant prostate cancer (csPCa) according to Gleason Score (GS) based on biparametric MRI (bpMRI) radiomic features and clinical information. Methods: This retrospective study included 86 adult Hispanic men (60 ± 8.2 years, median prostate-specific antigen density (PSA-D) 0.15 ng/mL²) with PCa who underwent prebiopsy 3T MRI followed by targeted MRI–ultrasound fusion and systematic biopsy. Two observers performed 2D segmentation of lesions in T2WI/ADC images. We classified csPCa (GS ≥ 7) vs. non-csPCa (GS = 6). Univariate statistical tests were performed for different parameters, including prostate volume (PV), PSA-D, PI-RADS, and radiomic features. Multivariate models were built using the automatic feature selection algorithm Recursive Feature Elimination (RFE) and different classifiers. A stratified split separated the train/test (80%) and validation (20%) sets. Results: Radiomic features derived from T2WI/ADC are associated with GS in patients with PCa. The best model found was multivariate, including image (T2WI/ADC) and clinical (PV and PSA-D) information. The validation area under the curve (AUC) was 0.80 for differentiating csPCa from non-csPCa, exhibiting better performance than PI-RADS (AUC: 0.71) and PSA-D (AUC: 0.78). Conclusion: Our multivariate ML model outperforms PI-RADS v2.1 and established clinical indicators like PSA-D in classifying csPCa accurately. This underscores MRI-derived radiomics’ (T2WI/ADC) potential as a robust biomarker for assessing PCa aggressiveness in Hispanic patients.

Radiotherapy (RT) is an essential part of breast cancer (BC) treatments. Unfortunately, heart exposure to radiation can also impair the long-term survival of patients. Our study aimed to quantify the oncological benefit and the cardiovascular (CV) risk associated with modern RT in a real-world cohort of BC patients. Our descriptive study enrolled BC patients who received adjuvant RT. Ten-year overall survival (OS) was estimated using Predict^® version 2.1 (National Health Service, London, UK). The basal risk of CV events was estimated using the American Heart Association (ACC/AHA) CV score. Treatment volumes and mean cardiac doses were obtained from RT treatment plan records. The increased risk of CV events due to RT was estimated using a model proposed by Darby. The risk of acute myocardial infarction or stroke mortality was estimated using HeartScore^® (European Society of Cardiology, Brussels, Belgium). A total of 256 BC patients were included in the study. The average age of patients was 57 years old (range: 25–91); 49.6% had left BC. The mean cardiac dose was 166 cGy (interquartile range (IQR) 94–273); the estimated hazard ratio (HR) for CV disease was HR 1.12 (confidence interval (CI) 1.04–1.24). The estimated baseline 10-year CV risk was 5.6% (0.2 to 51.2); CV risk increased by 0.9% (range 0.02–35.47%) after RT. The absolute risk of 10-year mortality from CV disease was 2.5% (0.1–9); RT was associated with an estimated 4.9% survival benefit (3.73–6.07) against BC death and a 0.23% (0.17–0.29) estimated increase in CV mortality. Modern RT decreased 10-year BC mortality by 4% but increased CV mortality by 0.2% in this cohort. Our findings encourage the implementation of personalized adjuvant RT treatments that balance risks and benefits to improve long-term BC patient survival.

Generative image models have achieved remarkable progress in both natural and medical imaging. In the medical context, these techniques offer a potential solution to data scarcity-especially for low-prevalence anomalies that impair the performance of AI-driven diagnostic and segmentation tools. However, questions remain regarding the fidelity and clinical utility of synthetic images, since poor generation quality can undermine model generalizability and trust. In this study, we evaluate the effectiveness of state-of-the-art generative models-Generative Adversarial Networks (GANs) and Diffusion Models (DMs)-for synthesizing chest X-rays conditioned on four abnormalities: Atelectasis (AT), Lung Opacity (LO), Pleural Effusion (PE), and Enlarged Cardiac Silhouette (ECS). Using a benchmark composed of real images from the MIMIC-CXR dataset and synthetic images from both GANs and DMs, we conducted a reader study with three radiologists of varied experience. Participants were asked to distinguish real from synthetic images and assess the consistency between visual features and the target abnormality. Our results show that while DMs generate more visually realistic images overall, GANs can report better accuracy for specific conditions, such as absence of ECS. We further identify visual cues radiologists use to detect synthetic images, offering insights into the perceptual gaps in current models. These findings underscore the complementary strengths of GANs and DMs and point to the need for further refinement to ensure generative models can reliably augment training datasets for AI diagnostic systems.