Computer-aided diagnosis systems must make critical decisions from medical images that are often noisy, ambiguous, or conflicting, yet today's models are trained on overly simplistic labels that ignore diagnostic uncertainty. One-hot labels erase inter-rater variability and force models to make overconfident predictions, especially when faced with incomplete or artifact-laden inputs. We address this gap by introducing a novel framework that brings uncertainty back into the label space. Our method leverages neural network training dynamics (NNTD) to assess the inherent difficulty of each training sample. By aggregating and calibrating model predictions during training, we generate uncertainty-aware pseudo-labels that reflect the ambiguity encountered during learning. This label augmentation approach is architecture-agnostic and can be applied to any supervised learning pipeline to enhance uncertainty estimation and robustness. We validate our approach on a challenging echocardiography classification benchmark, demonstrating superior performance over specialized baselines in calibration, selective classification, and multi-view fusion.

Epicardial (EAT) and pericardial adipose tissue (PAT) are increasingly recognized as distinct fat depots with implications for cardiovascular disease. This review discusses their anatomical and physiological characteristics, as well as their pathophysiological roles. EAT, in direct contact with the myocardium, exerts local inflammatory and metabolic effects on the heart, while PAT influences cardiovascular health rather systemically. We sought to discuss the currently used imaging modalities to assess these fat compartments-CT, MRI, and echocardiography-emphasizing their advantages, limitations, and the urgent need for standardization for both scanning and image reconstruction. Advances in image segmentation, particularly deep learning-based approaches, have improved the accuracy and reproducibility of EAT and PAT quantification. This review also explores the role of EAT and PAT as risk factors for cardiovascular outcomes, summarizing conflicting evidence across studies. Finally, we summarize the effects of medical therapy and lifestyle interventions on reducing EAT volume. Understanding and accurately quantifying EAT and PAT is essential for cardiovascular risk stratification and may open new pathways for therapeutic interventions.

The prospective, multicenter German National Cohort (NAKO) provides a unique dataset of cardiac magnetic resonance (CMR) cine images. Effective processing of these images requires a robust segmentation and quality control pipeline. A deep learning model for semantic segmentation, based on the nnU-Net architecture, was applied to full-cycle short-axis cine images from 29,908 baseline participants. The primary objective was to determine data on structure and function for both ventricles (LV, RV), including end-diastolic volumes (EDV), end-systolic volumes (ESV), and LV myocardial mass. Quality control measures included a visual assessment of outliers in morphofunctional parameters, inter- and intra-ventricular phase differences, and time-volume curves (TVC). These were adjudicated using a five-point rating scale, ranging from five (excellent) to one (non-diagnostic), with ratings of three or lower subject to exclusion. The predictive value of outlier criteria for inclusion and exclusion was evaluated using receiver operating characteristics analysis. The segmentation model generated complete data for 29,609 participants (incomplete in 1.0%), of which 5,082 cases (17.0%) underwent visual assessment. Quality assurance yielded a sample of 26,899 (90.8%) participants with excellent or good quality, excluding 1,875 participants due to image quality issues and 835 participants due to segmentation quality issues. TVC was the strongest single discriminator between included and excluded participants (AUC: 0.684). Of the two-category combinations, the pairing of TVC and phases provided the greatest improvement over TVC alone (AUC difference: 0.044; p<0.001). The best performance was observed when all three categories were combined (AUC: 0.748). By extending the quality-controlled sample to include mid-level 'acceptable' quality ratings, a total of 28,413 (96.0%) participants could be included. The implemented pipeline facilitated the automated segmentation of an extensive CMR dataset, integrating quality control measures. This methodology ensures that ensuing quantitative analyses are conducted with a diminished risk of bias.

Aortic stenosis (AS), a prevalent and serious heart valve disorder, requires early detection but remains difficult to diagnose in routine practice. Although echocardiography with Doppler imaging is the clinical standard, these assessments are typically limited to trained specialists. Point-of-care ultrasound (POCUS) offers an accessible alternative for AS screening but is restricted to basic 2D B-mode imaging, often lacking the analysis Doppler provides. Our project introduces MultiASNet, a multimodal machine learning framework designed to enhance AS screening with POCUS by combining 2D B-mode videos with structured data from echocardiography reports, including Doppler parameters. Using contrastive learning, MultiASNet aligns video features with report features in tabular form from the same patient to improve interpretive quality. To address misalignment where a single report corresponds to multiple video views, some irrelevant to AS diagnosis, we use cross-attention in a transformer-based video and tabular network to assign less importance to irrelevant report data. The model integrates structured data only during training, enabling independent use with B-mode videos during inference for broader accessibility. MultiASNet also incorporates sample selection to counteract label noise from observer variability, yielding improved accuracy on two datasets. We achieved balanced accuracy scores of 93.0% on a private dataset and 83.9% on the public TMED-2 dataset for AS detection. For severity classification, balanced accuracy scores were 80.4% and 59.4% on the private and public datasets, respectively. This model facilitates reliable AS screening in non-specialist settings, bridging the gap left by Doppler data while reducing noise-related errors. Our code is publicly available at github.com/DeepRCL/MultiASNet.

The contractile motion of the heart is strongly determined by the distribution of the fibers that constitute cardiac tissue. Strain analysis informed with the orientation of fibers allows to describe several pathologies that are typically associated with impaired mechanics of the myocardium, such as cardiovascular disease. Several methods have been developed to estimate strain-derived metrics from traditional imaging techniques. However, the physical models underlying these methods do not include fiber mechanics, restricting their capacity to accurately explain cardiac function. In this work, we introduce WarpPINN-fibers, a physics-informed neural network framework to accurately obtain cardiac motion and strains enhanced by fiber information. We train our neural network to satisfy a hyper-elastic model and promote fiber contraction with the goal to predict the deformation field of the heart from cine magnetic resonance images. For this purpose, we build a loss function composed of three terms: a data-similarity loss between the reference and the warped template images, a regularizer enforcing near-incompressibility of cardiac tissue and a fiber-stretch penalization that controls strain in the direction of synthetically produced fibers. We show that our neural network improves the former WarpPINN model and effectively controls fiber stretch in a synthetic phantom experiment. Then, we demonstrate that WarpPINN-fibers outperforms alternative methodologies in landmark-tracking and strain curve prediction for a cine-MRI benchmark with a cohort of 15 healthy volunteers. We expect that our method will enable a more precise quantification of cardiac strains through accurate deformation fields that are consistent with fiber physiology, without requiring imaging techniques more sophisticated than MRI.

Automatic quantification of intramyocardial motion and strain from tagging MRI remains an important but challenging task. We propose a method using implicit neural representations (INRs), conditioned on learned latent codes, to predict continuous left ventricular (LV) displacement -- without requiring inference-time optimisation. Evaluated on 452 UK Biobank test cases, our method achieved the best tracking accuracy (2.14 mm RMSE) and the lowest combined error in global circumferential (2.86%) and radial (6.42%) strain compared to three deep learning baselines. In addition, our method is $\sim$380$\times$ faster than the most accurate baseline. These results highlight the suitability of INR-based models for accurate and scalable analysis of myocardial strain in large CMR datasets.

Despite the Coronary Artery Reporting and Data System (CAD-RADS) providing a standardized approach, radiologists continue to favor free-text reports. This preference creates significant challenges for data extraction and analysis in longitudinal studies, potentially limiting large-scale research and quality assessment initiatives. To evaluate the ability of the generative pre-trained transformer (GPT)-4o model to convert real-world coronary computed tomography angiography (CCTA) free-text reports into structured data and automatically identify CAD-RADS categories and P categories. This retrospective study analyzed CCTA reports from January 2024 and July 2024. A subset of 25 reports was used for prompt engineering to instruct the large language models (LLMs) in extracting CAD-RADS categories, P categories, and the presence of myocardial bridges and noncalcified plaques. Reports were processed using the GPT-4o API (application programming interface) and custom Python scripts. The ground truth was established by radiologists based on the CAD-RADS 2.0 guidelines. Model performance was assessed using accuracy, sensitivity, specificity, and F1-score. Intrarater reliability was assessed using Cohen κ coefficient. Among 999 patients (median age 66 y, range 58-74; 650 males), CAD-RADS categorization showed accuracy of 0.98-1.00 (95% CI 0.9730-1.0000), sensitivity of 0.95-1.00 (95% CI 0.9191-1.0000), specificity of 0.98-1.00 (95% CI 0.9669-1.0000), and F1-score of 0.96-1.00 (95% CI 0.9253-1.0000). P categories demonstrated accuracy of 0.97-1.00 (95% CI 0.9569-0.9990), sensitivity from 0.90 to 1.00 (95% CI 0.8085-1.0000), specificity from 0.97 to 1.00 (95% CI 0.9533-1.0000), and F1-score from 0.91 to 0.99 (95% CI 0.8377-0.9967). Myocardial bridge detection achieved an accuracy of 0.98 (95% CI 0.9680-0.9870), and noncalcified coronary plaques detection showed an accuracy of 0.98 (95% CI 0.9680-0.9870). Cohen κ values for all classifications exceeded 0.98. The GPT-4o model efficiently and accurately converts CCTA free-text reports into structured data, excelling in CAD-RADS classification, plaque burden assessment, and detection of myocardial bridges and calcified plaques.

Automatic quantification of intramyocardial motion and strain from tagging MRI remains an important but challenging task. We propose a method using implicit neural representations (INRs), conditioned on learned latent codes, to predict continuous left ventricular (LV) displacement -- without requiring inference-time optimisation. Evaluated on 452 UK Biobank test cases, our method achieved the best tracking accuracy (2.14 mm RMSE) and the lowest combined error in global circumferential (2.86%) and radial (6.42%) strain compared to three deep learning baselines. In addition, our method is $\sim$380$\times$ faster than the most accurate baseline. These results highlight the suitability of INR-based models for accurate and scalable analysis of myocardial strain in large CMR datasets.

Automatic quantification of intramyocardial motion and strain from tagging MRI remains an important but challenging task. We propose a method using implicit neural representations (INRs), conditioned on learned latent codes, to predict continuous left ventricular (LV) displacement -- without requiring inference-time optimisation. Evaluated on 452 UK Biobank test cases, our method achieved the best tracking accuracy (2.14 mm RMSE) and the lowest combined error in global circumferential (2.86%) and radial (6.42%) strain compared to three deep learning baselines. In addition, our method is $\sim$380$\times$ faster than the most accurate baseline. These results highlight the suitability of INR-based models for accurate and scalable analysis of myocardial strain in large CMR datasets.

Video-based AI systems are increasingly adopted in safety-critical domains such as autonomous driving and healthcare. However, interpreting their decisions remains challenging due to the inherent spatiotemporal complexity of video data and the opacity of deep learning models. Existing explanation techniques often suffer from limited temporal coherence, insufficient robustness, and a lack of actionable causal insights. Current counterfactual explanation methods typically do not incorporate guidance from the target model, reducing semantic fidelity and practical utility. We introduce Latent Diffusion for Video Counterfactual Explanations (LD-ViCE), a novel framework designed to explain the behavior of video-based AI models. Compared to previous approaches, LD-ViCE reduces the computational costs of generating explanations by operating in latent space using a state-of-the-art diffusion model, while producing realistic and interpretable counterfactuals through an additional refinement step. Our experiments demonstrate the effectiveness of LD-ViCE across three diverse video datasets, including EchoNet-Dynamic (cardiac ultrasound), FERV39k (facial expression), and Something-Something V2 (action recognition). LD-ViCE outperforms a recent state-of-the-art method, achieving an increase in R2 score of up to 68% while reducing inference time by half. Qualitative analysis confirms that LD-ViCE generates semantically meaningful and temporally coherent explanations, offering valuable insights into the target model behavior. LD-ViCE represents a valuable step toward the trustworthy deployment of AI in safety-critical domains.