Objectives: The aim of this study was to synthesize current evidence on artificial intelligence (AI) adoption in cardiovascular imaging across low- and middle-income countries (LMICs), highlighting diagnostic performance, implementation barriers, and potential solutions. Methods: We conducted a systematic review of PubMed, Embase, Cochrane Library, Web of Science, and Scopus for studies evaluating AI-based echocardiography, cardiac CT, or cardiac MRI in LMICs. Articles were screened according to PRISMA guidelines, and data on diagnostic outcomes, challenges, and enabling factors were extracted and narratively synthesized. Results: Twelve studies met the inclusion criteria. AI-driven methods frequently surpassed 90% accuracy in detecting coronary artery disease, rheumatic heart disease, and left ventricular hypertrophy, often enabling task shifting to non-expert operators. Challenges included limited dataset diversity, operator dependence, infrastructure constraints, and ethical considerations. Insights from high-income countries, such as automated segmentation and accelerated imaging, suggest potential for broader AI integration in cardiac MRI and CT. Conclusions: AI holds promise for enhancing cardiovascular care in LMICs by improving diagnostic accuracy and workforce efficiency. However, multi-center data sharing, targeted training, reliable infrastructure, and robust governance are essential for sustainable adoption. This review underscores AI’s capacity to bridge resource gaps in LMICs, offering practical pathways for future research, clinical practice, and policy development in global cardiovascular imaging.

Background. Severe left ventricular hypertrophy (SLVH) is a high-risk structural cardiac abnormality associated with increased risk of heart failure. It is typically assessed using echocardiography or cardiac magnetic resonance imaging, but these modalities are limited by cost, accessibility, and workflow burden. We introduce a deep learning framework that classifies SLVH directly from chest radiographs, without intermediate anatomical estimation models or demographic inputs. A key contribution of this work lies in interpretability. We quantify how clinically relevant attributes are encoded within internal representations, enabling transparent model evaluation and integration into AI-assisted workflows. Methods. We construct class-balanced subsets from the CheXchoNet dataset with equal numbers of SLVH-positive and negative cases while preserving the original train, validation, and test proportions. ResNet-18 is fine-tuned from ImageNet weights, and a Vision Transformer (ViT) encoder is pretrained via masked autoencoding with a trainable classification head. No anatomical or demographic inputs are used during training. We apply Mutual Information Neural Estimation (MINE) to quantify dependence between learned features and five attributes: age, sex, interventricular septal diameter (IVSDd), posterior wall diameter (LVPWDd), and internal diameter (LVIDd). Results. ViT achieves an AUROC of 0.82 [95% CI: 0.78–0.85] and an AUPRC of 0.80 [95% CI: 0.76–0.85], indicating strong performance in SLVH detection from chest radiographs. MINE reveals clinically coherent attribute encoding in learned features: age > sex > IVSDd > LVPWDd > LVIDd. Conclusions. This study shows that SLVH can be accurately classified from chest radiographs alone. The framework combines diagnostic performance with quantitative interpretability, supporting reliable deployment in triage and decision support.

Psoriasis (PsO) severity scoring is important for clinical trials but is hindered by inter-rater variability and the burden of in person clinical evaluation. Remote imaging using patient captured mobile photos offers scalability but introduces challenges, such as variation in lighting, background, and device quality that are often imperceptible to humans but can impact model performance. These factors, along with inconsistencies in dermatologist annotations, reduce the reliability of automated severity scoring. We propose a framework to automatically flag problematic training images that introduce spurious correlations which degrade model generalization, using a gradient based interpretability approach. By tracing the gradients of misclassified validation images, we detect training samples where model errors align with inconsistently rated examples or are affected by subtle, nonclinical artifacts. We apply this method to a ConvNeXT based weakly supervised model designed to classify PsO severity from phone images. Removing 8.2% of flagged images improves model AUC-ROC by 5% (85% to 90%) on a held out test set. Commonly, multiple annotators and an adjudication process ensure annotation accuracy, which is expensive and time consuming. Our method detects training images with annotation inconsistencies, potentially removing the need for manual review. When applied to a subset of training data rated by two dermatologists, the method identifies over 90% of cases with inter-rater disagreement by reviewing only the top 30% of samples. This improves automated scoring for remote assessments, ensuring robustness despite data collection variability.

Purpose: To examine the principal technical challenges of four-dimensional cinematic rendering (4D-CR) in cardiac imaging and identify priorities for reproducible clinical implementation. Methods: This narrative review searched PubMed, Scopus, and Google Scholar from January 2016 through mid-2026 using terms related to cinematic rendering, global illumination, four-dimensional imaging, cardiac CT, congenital heart disease, and volume rendering. Reference lists and relevant computer graphics literature were also examined. Results: Extending photorealistic cinematic rendering across cardiac phases introduces computational and rendering-time demands, motion artifact and inter-phase misregistration, non-standardized illumination parameters, and challenges in anatomical and temporal validation. Additional barriers include integration into picture archiving and communication systems, clinical workflow, and clinician training. Emerging approaches include AI-assisted rendering optimization, real-time processing, and neural scene representations such as Gaussian splatting, which may decouple photorealistic playback from repeated path tracing. Current evidence remains dominated by technical reports and small case series rather than systematic comparative validation. Conclusion: Broader clinical use of 4D-CR requires standardized technical reporting, systematic anatomical and temporal validation, clinically integrated storage and display, and structured training. Technical advances offer potential solutions, but convincing rendered appearance alone does not establish diagnostic fidelity.