Search Results (79)

Precise segmentation of cardiac substructures in magnetic resonance imaging is pivotal for diagnosis and treatment planning but remains impeded by anatomical scale heterogeneity and the scarcity of high-quality pixel-level annotations. Existing deep learning paradigms often struggle to simultaneously resolve the global geometry of ventricular cavities and the fine-grained boundaries of the myocardium, particularly in low-data regimes. To address these challenges, we propose SAS-SemiUNet++, a holistic semi-supervised segmentation framework. This architecture incorporates two novel mechanisms: (1) The Scale-Aware Semantic Recalibration (SASR) unit, which functions as a dynamic semantic gate to adaptively adjust receptive fields, mimicking a radiologist’s variable-focus mechanism to capture multi-scale anatomical details, and (2) Stochastic Consistency Regularization (SCR), a dual-path perturbation strategy that enforces geometric invariance on unlabeled data, thereby mitigating overfitting to noisy pseudo-labels. Comprehensive evaluations on the ACDC benchmark demonstrate that SAS-SemiUNet++ significantly outperforms state-of-the-art methods, achieving superior segmentation accuracy and boundary fidelity, particularly in reducing the 95% Hausdorff distance. This study presents a data-efficient and robust solution for cardiac image analysis, offering potential for scalable clinical deployment. Full article

The performance of deep-learning-based segmentation models is strongly dependent on the quality of the input data, which is frequently heterogeneous or degraded in real-world medical imaging scenarios. This work presents a data-centric algorithmic pipeline designed to improve cardiac MRI segmentation accuracy through systematic image enhancement and automatic slice-quality filtering. The proposed method is formalized as deterministic algorithm that combines image processing and supervised learning components. The approach integrates a contrast- and structure-preserving enhancement stage, based on bilateral filtering and adaptive histogram equalization, with a quality-aware selection algorithm. Slice quality is assessed using anatomical attributes extracted via YOLOv11s-based localization and a supervised classification model trained to identify diagnostically reliable images. When applied to transformer-based segmentation architectures such as ViTUNeT, the pipeline yields consistent improvements across all evaluation metrics without increasing model complexity or training cost. These findings emphasize the importance of algorithmic data curation as an effective strategy for enhancing robustness and stability in deep-learning segmentation pipelines and demonstrate the broader applicability of the proposed approach to computer-vision tasks involving heterogeneous or low-quality image datasets. Full article

Background: Cardiovascular magnetic resonance (CMR) is the clinical gold standard for assessing cardiac anatomy and function. However, the manual segmentation of cardiac structures and myocardial infarction (MI) is time-consuming, prone to inter-observer variability, and often depends on contrast-enhanced imaging. Although deep learning (DL) has enabled substantial automation, challenges remain in generalizability, particularly for MI detection from non-contrast cine CMR. Objective: This study proposes a comprehensive DL-based framework for automatic segmentation of cardiac structures and myocardial infarction using contrast-free cine CMR. Methods: The framework integrates multiple convolutional neural network (CNN) architectures for cardiac structure segmentation with an attention-based deep learning model for MI localization. Post-processing refinement using stacked autoencoders and active contour modeling is applied to improve anatomical consistency. Segmentation performance is evaluated using overlap-based and boundary-based metrics, including the Dice Similarity Coefficient (DSC), Mean Contour Distance (MCD), and Hausdorff Distance (HD). Results: The best-performing model achieved Dice scores of 0.93 ± 0.05 for the left ventricular (LV) cavity, 0.89 ± 0.04 for the LV myocardium, and 0.91 ± 0.06 for the right ventricular (RV) cavity, with consistently low boundary errors across all structures. Myocardial infarction segmentation achieved a Dice score of 0.80 ± 0.02 with high recall, demonstrating reliable infarct localization without the use of contrast agents. Conclusions: By enabling accurate cardiac structure and myocardial infarction segmentation from contrast-free cine CMR, the proposed framework supports broader clinical applicability, particularly for patients with contraindications to gadolinium-based contrast agents and in emergency or resource-limited settings. This approach facilitates scalable, contrast-independent cardiac assessment. Full article

Three-dimensional (3D) anatomic modeling derived from high-resolution medical imaging, such as computed tomography (CT) and magnetic resonance imaging (MRI), has been increasingly adopted in preclinical testing and device development. This white paper describes a cardiac-specific workflow that integrates 3D printing and silicone molding for support device development and procedural simulation. Patient-derived computed tomography angiography data were segmented using FDA-cleared medical modeling software to isolate the left ventricular anatomy and were further processed in computer-aided design (CAD) to ensure accurate physiological wall thickness and structural fidelity. Material jetting 3D printing was performed on a Stratasys J750 using material distributions designed to mimic the mechanical properties of myocardium, thereby approximating myocardial compliance. In parallel, stereolithography apparatus molds were designed from the left ventricle CAD model to cast transparent, pliable left ventricular models in Sorta-Clear™ 18 silicone. The 3D-printed models preserved intricate morphological detail and were suitable for mechanical manipulation and device deployment studies, whereas silicone models offered tunable mechanical properties, transparency for visualization, and durability for repeated use. Together, these complementary modalities provided rapid manufacturing capability and application-relevant physical representation. Case-specific parameters, strengths, and limitations of both models in enhancing patient care and device testing are highlighted, with relevance to heart failure applications. Current knowledge gaps, workflow and integration challenges, and future opportunities are identified, positioning this work as a reference framework for continued innovation in anatomic modeling. Within the collaborative framework of Mayo Clinic’s Anatomic Modeling Unit and Simulation Center, this integrated modeling workflow demonstrates the value of multidisciplinary collaboration between engineers and clinicians. Clinically, these patient-specific left ventricular models may enable pre-procedural device sizing and positioning and may support simulation of mechanical circulatory support (MCS) deployment while identifying possible anatomic constraints prior to intervention. This workflow has direct applicability in advanced heart failure patients undergoing MCS support, such as the Impella axillary MCS device or the durable LVAD, with potential to reduce procedural uncertainty while reducing complications and improving peri-procedural outcomes. Additionally, these models also serve as high-accuracy educational tools, enabling trainees and multidisciplinary care teams to visualize and possibly rehearse procedural steps while gaining hands-on experience in a risk-free environment. Full article

Artificial intelligence (AI) is rapidly reshaping adult cardiac surgery, enabling more accurate diagnostics, personalized risk assessment, advanced surgical planning, and proactive postoperative care. Preoperatively, deep-learning interpretation of ECGs, automated CT/MRI segmentation, and video-based echocardiography improve early disease detection and refine risk stratification beyond conventional tools such as EuroSCORE II and the STS calculator. AI-driven 3D reconstruction, virtual simulation, and augmented-reality platforms enhance planning for structural heart and aortic procedures by optimizing device selection and anticipating complications. Intraoperatively, AI augments robotic precision, stabilizes instrument motion, identifies anatomy through computer vision, and predicts hemodynamic instability via real-time waveform analytics. Integration of the Hypotension Prediction Index into perioperative pathways has already demonstrated reductions in ventilation duration and improved hemodynamic control. Postoperatively, machine-learning early-warning systems and physiologic waveform models predict acute kidney injury, low-cardiac-output syndrome, respiratory failure, and sepsis hours before clinical deterioration, while emerging closed-loop control and remote monitoring tools extend individualized management into the recovery phase. Despite these advances, current evidence is limited by retrospective study designs, heterogeneous datasets, variable transparency, and regulatory and workflow barriers. Nonetheless, rapid progress in multimodal foundation models, digital twins, hybrid OR ecosystems, and semi-autonomous robotics signals a transition toward increasingly precise, predictive, and personalized cardiac surgical care. With rigorous validation and thoughtful implementation, AI has the potential to substantially improve safety, decision-making, and outcomes across the entire cardiac surgical continuum. Full article

Background: Real-time phase-contrast magnetic resonance (RT-PCMR) imaging allows free-breathing assessment of blood flow across cardiac valves and vessels. However, the feasibility of free-breathing RT-PCMR on a mid-field (0.55T) MRI system has yet to be established. Aim: The primary objective of this study was to implement a RT-PCMR sequence using a dual-density golden-angle spiral readout with SENSE-based compressed sensing (CS) reconstruction on a 0.55T MRI system. The secondary objective was to evaluate the feasibility of this approach in an adult cohort comprising healthy volunteers and patients with cardiovascular disease. Materials and Methods: Data from 33 participants were included in the flow quantification analysis (healthy volunteers: n = 17, 9 females, mean age 30.4 ± 14.6 years; patients: n = 16, 11 females, mean age 45.9 ± 17.4 years), with breath-held (BH) segmented Cartesian PCMR used as the reference standard. Results: In volunteers, RT-PCMR showed good agreement for net flow, peak flow rate, and pulmonary–systemic flow ratio (Qp/Qs), without significant bias (p > 0.05) and slightly underestimated peak velocity [7.9% in the aorta and 8.6% in the main pulmonary artery (MPA)]. In patients, RT-PCMR slightly underestimated peak flow rate (aorta, 6.2%; MPA; 4.6%) and peak velocity (aorta,12.7%; MPA, 10.4%). A sub-analysis of six patients scanned at both 0.55T and 3T showed close agreement between field strengths. Conclusions: These results demonstrate the feasibility of our RT-PCMR sequence on a commercial 0.55T system. Full article

Deep learning models based on supervised learning rely heavily on large annotated datasets and particularly in the context of medical image segmentation, the requirement for pixel-level annotations makes the labeling process labor-intensive, time-consuming and expensive. To overcome these limitations, self-supervised learning (SSL) has emerged as a promising alternative that learns generalizable representations from unlabeled data; however, existing SSL frameworks often employ highly parameterized encoders that are computationally expensive and may lack robustness in label-scarce settings. In this work, we propose a scattering-based SSL framework that integrates Wavelet Scattering Networks (WSNs) and Parametric Scattering Networks (PSNs) into a Bootstrap Your Own Latent (BYOL) pretraining pipeline. By replacing the initial stages of the BYOL encoder with fixed or learnable scattering-based front-ends, the proposed method reduces the number of learnable parameters while embedding translation-invariant and small deformation-stable representations into the SSL pipeline. The pretrained encoders are transferred to a U-Net and fine-tuned for cardiac image segmentation on two datasets with different imaging modalities, namely, cardiac cine MRI (ACDC) and cardiac CT (CHD), under varying amounts of labeled data. Experimental results show that scattering-based SSL pretraining consistently improves segmentation performance over random initialization and ImageNet pretraining in low-label regimes, with particularly pronounced gains when only a few labeled patients are available. Notably, the PSN variant achieves improvements of 4.66% and 2.11% in average Dice score over standard BYOL with only 5 and 10 labeled patients, respectively, on the ACDC dataset. These results demonstrate that integrating mathematically grounded scattering representations into SSL pipelines provides a robust and data-efficient initialization strategy for cardiac image segmentation, particularly under limited annotation and domain shift. Full article

Background/Objectives: Cardiac magnetic resonance (CMR) imaging is a key tool for diagnosing cardiovascular disease, but its analysis remains time-consuming and dependent on expert interpretation, which can limit throughput and reproducibility. To address these challenges, we aim to develop an automated solution that streamlines CMR post-processing, enabling consistent, rapid, and quantitative assessment of cardiac structures and myocardial pathology. Methods: We introduce SPOT-Cardio, an AI-powered imaging analysis toolbox based on a 2D breath-held late gadolinium enhancement (LGE) imaging technology: SPOT. This acquisition combines BR- and BL-LGE images in a single scan, allowing simultaneous capture of high-contrast scar information and detailed cardiac anatomy. Using the resulting CMR images, deep learning models (based on 2D U-Net or MedFormer) were trained to segment cardiac structures and myocardial scars. The trained models and associated image-processing algorithms were then integrated into the open-source medInria platform and specifically within its cardiac-focused MUSICardio application. Results: SPOT-Cardio enables automatic segmentation of cardiac structures and myocardial scars, performs landmark-based regional localization, and extracts key biomarkers such as scar volume, extent, and transmurality. The resulting quantitative measures are presented in standardized bullseye plots accompanied by detailed clinical reports. Conclusions: With a one-click workflow and intuitive visualization, SPOT-Cardio reduces manual workload and supports more accessible and consistent cardiovascular assessment. By integrating advanced image acquisition with AI-based automation, it provides a practical and efficient solution for streamlined and quantitative CMR analysis. Full article

Background: Medical image segmentation is essential for accurate diagnosis and treatment planning. The U-Net architecture is widely regarded as the gold standard, yet its large size and high computational demand pose significant challenges for practical deployment. Methods: Real data (MRI images) from hospital patients were used in this study. We proposed a novel lightweight architecture tailored specifically for myocardium (cardiac muscle) segmentation. Results: We presented results comparable to state-of-the-art methods in terms of IoU and Dice coefficients. Nonetheless, the results achieved are much more favorable from the perspective of AI’s sustainable development. The proposed architecture ensured the following average results: IOU = 0.7889 and Dice = 0.8780 using 263 k parameters and a total of 6.24 G FLOPs. Conclusions: The proposed schema can potentially be used to support radiologists in improving the diagnostic process. The presented approach is efficient and fast. Most promisingly, the reduction in the model’s complexity is significant compared to the state-of-the-art methods. Full article

The rapid development of Generative Adversarial Networks (GANs) has transformed medical image processing, enabling realistic image synthesis, augmentation, and restoration. This study presents a comparative evaluation of three representative GAN architectures, Pix2Pix, SPADE GAN, and Wasserstein GAN (WGAN), across multiple medical imaging tasks, including segmentation, image synthesis, and enhancement. Experiments were conducted on three benchmark datasets: ACDC (cardiac MRI), Brain Tumor MRI, and CHAOS (abdominal MRI). Model performance was assessed using Fréchet Inception Distance (FID), Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), Dice coefficient, and segmentation accuracy. Results show that SPADE-inpainting achieved the best image fidelity (PSNR ≈ 36 dB, SSIM > 0.97, Dice ≈ 0.94, FID < 0.01), while Pix2Pix delivered the highest segmentation accuracy (Dice ≈ 0.90 on ACDC). WGAN provided stable enhancement and strong visual sharpness on smaller datasets such as Brain Tumor MRI. The findings confirm that no single GAN architecture universally excels across all tasks; performance depends on data complexity and task objectives. Overall, GANs demonstrate strong potential for medical image augmentation and synthesis, though their clinical utility remains dependent on anatomical fidelity and dataset diversity. Full article

Introduction: Changes in left ventricular (LV) wall thickness serve as important diagnostic and prognostic indicators in various cardiovascular diseases. To date, no automated software exists for the measurement of myocardial segmental wall thickness in cardiac MRI (CMR), which leads to reliance on manual caliper measurements that carry risks of inaccuracy. Aims: This paper aims to present a new automated segmental wall thickness measurement software, OptiLayer, developed to address this issue and to compare it with the conventional manual measurement method. Methods: In our pilot study, the algorithm of the OptiLayer software was tested on 50 HEALTHY individuals, and 50 excessively trabeculated noncompaction (LVET) subjects with preserved LV function, whose morphology makes it more challenging to measure left ventricular wall thickness, although often occurring with myocardial thinning. Measurements were performed by two independent investigators who assessed LV wall thicknesses in 16 segments, both manually using the Medis Suite QMass program and automatically with the new OptiLayer method, which enables high-density sampling across the distance between the epicardial and endocardial contours. Results: The results showed that the segmental wall thickness measurement values of the OptiLayer algorithm were significantly higher than those of the manual caliper. In comparisons of the HEALTHY and LVET subgroups, OptiLayer measurements demonstrated differences at several points than manual measurements. Between the investigators, manual measurements showed low intraclass correlations (ICC below 0.6 on average), while measurements with OptiLayer gave excellent agreement (ICC above 0.9 in 75% of segments). Conclusions: Our study suggests that OptiLayer, a new automated wall thickness measurement software based on high-precision anatomical segmentation, offers a faster, more accurate, and more reproducible alternative to manual measurements. Full article

Recent advances in cardiovascular imaging have revolutionized the assessment and management of abdominal aortic aneurysm (AAA) through the integration of sophisticated haemodynamic biomarkers. This comprehensive review evaluates the clinical utility and mechanistic significance of multiple biomarkers in AAA pathogenesis, progression, and treatment outcomes. Advanced cardiac imaging modalities, including four-dimensional magnetic resonance imaging (4D MRI), computational fluid dynamics (CFD), and specialized echocardiography, enable precise quantification of critical haemodynamic parameters. Wall shear stress (WSS) emerges as a fundamental biomarker, with values below 0.4 Pa indicating pathological conditions and increased risk for aneurysm progression. Time-averaged wall shear stress (TAWSS), typically maintaining values above 1.5 Pa in healthy arterial segments, provides crucial information about sustained haemodynamic forces affecting the vessel wall. The oscillatory shear index (OSI), ranging from 0 (unidirectional flow) to 0.5 (purely oscillatory flow), quantifies directional changes in WSS during cardiac cycles. In AAA, elevated OSI values between 0.3 and 0.4 correlate with disturbed flow patterns and accelerated disease progression. The relative residence time (RRT), combining TAWSS and OSI, identifies regions prone to thrombosis, with values exceeding 2–3 Pa⁻¹ indicating increased risk. The endothelial cell activation potential (ECAP), calculated as OSI/TAWSS, serves as an integrated metric for endothelial dysfunction risk, with values above 0.2–0.3 Pa⁻¹ suggesting increased inflammatory activity. Additional biomarkers include the volumetric perivascular characterization index (VPCI), which assesses vessel wall inflammation through perivascular tissue analysis, and pulse wave velocity (PWV), measuring arterial stiffness. Central aortic systolic pressure and the aortic augmentation index provide essential information about cardiovascular load and arterial compliance. Novel parameters such as particle residence time, flow stagnation, and recirculation zones offer detailed insights into local haemodynamics and potential complications. Implementation challenges include the need for specialized equipment, standardized protocols, and expertise in data interpretation. However, the potential for improved patient outcomes through more precise risk stratification and personalized treatment planning justifies continued development and validation of these advanced assessment tools. Full article

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. Full article

Applying pre-trained medical deep learning segmentation models to out-of-domain images often yields predictions of insufficient quality. In this study, we propose using a robust generalizing descriptor, along with augmentation, to enable domain-generalized pre-training and test-time adaptation, thereby achieving high-quality segmentation in unseen domains. In this study, five different publicly available datasets, including 3D CT and MRI images, are used to evaluate segmentation performance in out-of-domain scenarios. The settings include abdominal, spine, and cardiac imaging. Domain-generalized pre-training on source data is used to obtain the best initial performance in the target domain. We introduce a combination of the generalizing SSC descriptor and GIN intensity augmentation for optimal generalization. Segmentation results are subsequently optimized at test time, where we propose adapting the pre-trained models for every unseen scan using a consistency scheme with the augmentation–descriptor combination. The proposed generalized pre-training and subsequent test-time adaptation improve model performance significantly in CT to MRI cross-domain prediction for abdominal (+46.2 and +28.2 Dice), spine (+72.9), and cardiac (+14.2 and +55.7 Dice) scenarios (p < 0.001). Our method enables the optimal, independent use of source and target data, successfully bridging domain gaps with a compact and efficient methodology. Full article

Background/Objectives In this study we approach the problem of medical image segmentation by introducing a new loss function envelope that is derived from the Top-k loss strategy. We exploit the fact that, for semantic segmentation, the training loss is computed at two levels, more specifically at pixel level and at image level. Quite often, the envisaged problem has particularities that include noisy annotation at pixel level and limited data, but with accurate annotations at image level. Methods To address the mentioned issues, the Top-k strategy at image level and respectively the “Bottom all but $\sigma$” strategy at pixel level are assumed. To deal with the discontinuities of the differentials faced in the automatic learning, a derivative smoothing procedure is introduced. Results The method is thoroughly and successfully tested (in conjunction with a variety of backbone models) for several medical image segmentation tasks performed onto a variety of image acquisition types and human body regions. We present the burned skin area segmentation in standard color images, the segmentation of fetal abdominal structures in ultrasound images and ventricles and myocardium segmentation in cardiac MRI images, in all cases yielding performance improvements. Conclusions The proposed novel mechanism enhances model training by selectively emphasizing certain loss values by the use of two complementary strategies. The major benefits of the approach are clear in challenging scenarios, where the segmentation problem is inherently difficult or where the quality of pixel-level annotations is degraded by noise or inconsistencies. The proposed approach performs equally well in both convolutional neural networks (CNNs) and vision transformer (ViT) architectures. Full article