Video-based AI for beat-to-beat assessment of cardiac function

Advancing hierarchical neural networks with scale-aware pyramidal feature learning for medical image dense prediction

BACKGROUND AND OBJECTIVE

Hierarchical neural networks are pivotal in medical imaging for multi-scale representation, aiding in tasks such as object detection and segmentation. However, their effectiveness is often limited by the loss of intra-scale information and misalignment of inter-scale features. Our study introduces the Integrated-Scale Pyramidal Interactive Reconfiguration to Enhance feature learning (INSPIRE).

METHODS

INSPIRE focuses on intra-scale semantic enhancement and precise inter-scale spatial alignment, integrated with a novel spatial-semantic back augmentation technique. We evaluated INSPIRE's efficacy using standard hierarchical neural networks, such as UNet and FPN, across multiple medical segmentation challenges including brain tumors and polyps. Additionally, we extended our evaluation to object detection and semantic segmentation in natural images to assess generalizability.

RESULTS

INSPIRE demonstrated superior performance over standard baselines in medical segmentation tasks, showing significant improvements in feature learning and alignment. In identifying brain tumors and polyps, INSPIRE achieved enhanced precision, sensitivity, and specificity compared to traditional models. Further testing in natural images confirmed the adaptability and robustness of our approach.

CONCLUSIONS

INSPIRE effectively enriches semantic clarity and aligns multi-scale features, achieving integrated spatial-semantic coherence. This method seamlessly integrates with existing frameworks used in medical image analysis, thereby promising to significantly enhance the efficacy of computer-aided diagnostics and clinical interventions. Its application could lead to more accurate and efficient imaging processes, essential for improved patient outcomes.

Bridging multi-level gaps: Bidirectional reciprocal cycle framework for text-guided label-efficient segmentation in echocardiography

Text-guided visual understanding is a potential solution for downstream task learning in echocardiography. It can reduce reliance on labeled large datasets and facilitate learning clinical tasks. This is because the text can embed highly condensed clinical information into predictions for visual tasks. The contrastive language-image pretraining (CLIP) based methods extract image-text features by constructing a contrastive learning pre-train process in a sequence of matched text and images. These methods adapt the pre-trained network parameters to improve downstream task performance with text guidance. However, these methods still have the challenge of the multi-level gap between image and text. It mainly stems from spatial-level, contextual-level, and domain-level gaps. It is difficult to deal with medical image-text pairs and dense prediction tasks. Therefore, we propose a bidirectional reciprocal cycle (BRC) framework to bridge the multi-level gaps. First, the BRC constructs pyramid reciprocal alignments of embedded global and local image-text feature representations. This matches complex medical expertise with corresponding phenomena. Second, BRC enforces the forward inference to be consistent with the reverse mapping (i.e., the text → feature is consistent with the feature → text or feature → image). This enforces the perception of the contextual relationship between input data and feature. Third, the BRC can adapt to the specific downstream segmentation task. This embeds complex text information to directly guide downstream tasks with a cross-modal attention mechanism. Compared with 22 existing methods, our BRC can achieve state-of-the-art performance on segmentation tasks (DSC = 95.2%). Extensive experiments on 11048 patients show that our method can significantly improve the accuracy and reduce the reliance on labeled data (DSC increased from 81.5% to 86.6% with text assistance in 1% labeled proportion data).

Deep Learning in Echocardiography for Enhanced Detection of Left Ventricular Function and Wall Motion Abnormalities

Cardiovascular diseases (CVDs) remain a leading cause of mortality worldwide, underscoring the need for advancements in diagnostic methodologies to improve early detection and treatment outcomes. This systematic review examines the integration of advanced deep learning (DL) techniques in echocardiography for detecting cardiovascular abnormalities, adhering to PRISMA 2020 guidelines. Through a comprehensive search across databases like IEEE Xplore, PubMed, and Web of Science, 29 studies were identified and analyzed, focusing on deep convolutional neural networks (DCNNs) and their role in enhancing the diagnostic precision of echocardiographic assessments. The findings highlight DL's capability to improve the accuracy and reproducibility of detecting and classifying echocardiographic data, particularly in measuring left ventricular function and identifying wall motion abnormalities. Despite these advancements, challenges such as data diversity, image quality, and the computational demands of DL models hinder their broader clinical adoption. In conclusion, DL offers significant potential to enhance the diagnostic capabilities of echocardiography. However, successful clinical implementation requires addressing issues related to data quality, computational demands, and system integration.

Semi-Supervised Echocardiography Video Segmentation via Adaptive Spatio-Temporal Tensor Semantic Awareness and Memory Flow

Accurate segmentation of cardiac structures in echocardiography videos is vital for diagnosing heart disease. However, challenges such as speckle noise, low spatial resolution, and incomplete video annotations hinder the accuracy and efficiency of segmentation tasks. Existing video-based segmentation methods mainly utilize optical flow estimation and cross-frame attention to establish pixel-level correlations between frames, which are usually sensitive to noise and have high computational costs. In this paper, we present an innovative echocardiography video segmentation framework that exploits the inherent spatio-temporal correlation of echocardiography video feature tensors. Specifically, we perform adaptive tensor singular value decomposition (t-SVD) on the video semantic feature tensor within a learnable 3D transform domain. By utilizing learnable thresholds, we preserve the principal singular values to reduce redundancy in the high-dimensional spatio-temporal feature tensor and enforce its potential low-rank property. Through this process, we can capture the temporal evolution of the target tissue by effectively utilizing information from limited labeled frames, thus overcoming the constraints of sparse annotations. Furthermore, we introduce a memory flow method that propagates relevant information between adjacent frames based on the multi-scale affinities to precisely resolve frame-to-frame variations of dynamic tissues, thereby improving the accuracy and continuity of segmentation results. Extensive experiments conducted on both public and private datasets validate the superiority of our proposed method over state-of-the-art methods, demonstrating improved performance in echocardiography video segmentation.

CACTUS: An open dataset and framework for automated Cardiac Assessment and Classification of Ultrasound images using deep transfer learning

Cardiac ultrasound (US) scanning is one of the most commonly used techniques in cardiology to diagnose the health of the heart and its proper functioning. During a typical US scan, medical professionals take several images of the heart to be classified based on the cardiac views they contain, with a focus on high-quality images. However, this task is time consuming and error prone. Therefore, it is necessary to consider ways to automate these tasks and assist medical professionals in classifying and assessing cardiac US images. Machine learning (ML) techniques are regarded as a prominent solution due to their success in the development of numerous applications aimed at enhancing the medical field, including addressing the shortage of echography technicians. However, the limited availability of medical data presents a significant barrier to the application of ML in the field of cardiology, particularly regarding US images of the heart. This paper addresses this challenge by introducing the first open graded dataset for Cardiac Assessment and ClassificaTion of UltraSound (CACTUS), which is available online. This dataset contains images obtained from scanning a CAE Blue Phantom and representing various heart views and different quality levels, exceeding the conventional cardiac views typically found in literature. Additionally, the paper introduces a Deep Learning (DL) framework consisting of two main components. The first component is responsible for classifying cardiac US images based on the heart view using a Convolutional Neural Network (CNN) architecture. The second component uses the concept of Transfer Learning (TL) to utilize knowledge from the first component and fine-tune it to create a model for grading and assessing cardiac images. The framework demonstrates high performance in both classification and grading, achieving up to 99.43% accuracy and as low as 0.3067 error, respectively. To showcase its robustness, the framework is further fine-tuned using new images representing additional cardiac views and also compared to several other state-of-the-art architectures. The framework's outcomes and its performance in handling real-time scans were also assessed using a questionnaire answered by cardiac experts.

Self-supervised learning for label-free segmentation in cardiac ultrasound

Segmentation and measurement of cardiac chambers from ultrasound is critical, but laborious and poorly reproducible. Neural networks can assist, but supervised approaches require the same problematic manual annotations. We build a pipeline for self-supervised segmentation combining computer vision, clinical knowledge, and deep learning. We train on 450 echocardiograms and test on 18,423 echocardiograms (including external data), using the resulting segmentations to calculate measurements. Coefficient of determination (r2) between clinically measured and pipeline-predicted measurements (0.55-0.84) are comparable to inter-clinician variation and to supervised learning. Average accuracy for detecting abnormal chambers is 0.85 (0.71-0.97). A subset of test echocardiograms (n = 553) have corresponding cardiac MRIs (the gold standard). Correlation between pipeline and MRI measurements is similar to that of clinical echocardiogram. Finally, the pipeline segments the left ventricle with an average Dice score of 0.89 (95% CI [0.89]). Our results demonstrate a manual-label free, clinically valid, and scalable method for segmentation from ultrasound.

Diagnosing Severe Low-Gradient vs Moderate Aortic Stenosis with Artificial Intelligence Based on Echocardiography Images

Diagnosis of aortic valve stenosis (AS) is performed manually by a physician experienced in echocardiography imaging. A specific subtype of AS, a severe low-gradient AS, is the most challenging one in terms of differentiating it from the moderate AS. In this study, an artificial intelligence (AI)-based model was used to diagnose the severe low-gradient AS in a fully automatic manner. Data from 158 consecutive patients undergoing echocardiography examination to assess AS severity were used. The obtained performance of our fully automatic approach was AUC = 0.719, 95% confidence interval, 0.640-0.798. It is an important step towards a comprehensive and automatic, image-only-based clinical decision support system for determining the presence of AS and its severity, especially in AS subtypes, such as low-gradient AS.

Preprocedural CT angiography and machine learning for mortality prediction after transcatheter aortic valve replacement

Prediction of outcomes following transcatheter aortic valve replacement (TAVR) is challenging. Considering that in aortic stenosis outcomes are governed by both valve degeneration and myocardial adverse remodeling, we aimed to evaluate machine-learning leveraging pre-procedural computed tomography (CT) for the prediction of 1-year mortality following TAVR. The analysis included data of consecutive patients who underwent TAVR at a high-volume center between January 2017 and January 2022 and was externally validated on unseen data from 3 international sites. Machine learning by extreme gradient boosting was trained and tested using clinical variables, CT-derived volumetric measurements including myocardial mass, and quantitative fibrocalcific aortic valve characteristics measured using standardized software. The EuroScore II and a separate machine learning risk score based exclusively on baseline clinical characteristics served as comparators. The derivation cohort included 631 consecutive patients (48 % men, 80 ± 8 years old, EuroSCORE II 6.5 [4.6-10.3] %). Machine learning was externally validated on data of 596 patients (48 % men, 81 ± 8 years old, EuroSCORE II 5.4 [4.7-8.1] %). In external validation, the machine learning prognostic risk score had an area under the receiver operator curve of 0.79 (0.74-0.84) which was superior to the EuroSCORE 0.59 (0.53-0.66), and the machine learning risk based on clinical data alone 0.64 (0.59-0.69), p < 0.001 for difference. Machine-learning integrating clinical data and CT-derived imaging characteristics was found to predict 1-year all-cause mortality following TAVR significantly better than clinical variables or clinical risk scores alone; and can help identify patients at higher prognostic risk prior to the procedure.

Causal recurrent intervention for cross-modal cardiac image segmentation

Cross-modal cardiac image segmentation is essential for cardiac disease analysis. In diagnosis, it enables clinicians to obtain more precise information about cardiac structure or function for potential signs by leveraging specific imaging modalities. For instance, cardiovascular pathologies such as myocardial infarction and congenital heart defects require precise cross-modal characterization to guide clinical decisions. The growing adoption of cross-modal segmentation in clinical research underscores its technical value, yet annotating cardiac images with multiple slices is time-consuming and labor-intensive, making it difficult to meet clinical and deep learning demands. To reduce the need for labels, cross-modal approaches could leverage general knowledge from multiple modalities. However, implementing a cross-modal method remains challenging due to cross-domain confounding. This challenge arises from the intricate effects of modality and view alterations between images, including inconsistent high-dimensional features. The confounding complicates the causality between the observation (image) and the prediction (label), thereby weakening the domain-invariant representation. Existing disentanglement methods face difficulties in addressing the confounding due to the insufficient depiction of the relationship between latent factors. This paper proposes the causal recurrent intervention (CRI) method to overcome the above challenge. It establishes a structural causal model that allows individual domains to maintain causal consistency through interventions. The CRI method integrates diverse high-dimensional variations into a singular causal relationship by embedding image slices into a sequence. This approach further distinguishes stable and dynamic factors from the sequence, subsequently separating the stable factor into modal and view factors and establishing causal connections between them. It then learns the dynamic factor and the view factor from the observation to obtain the label. Experimental results on cross-modal cardiac images of 1697 examples show that the CRI method delivers promising and productive cross-modal cardiac image segmentation performance.

Automated Deep Learning Phenotyping of Tricuspid Regurgitation in Echocardiography

Importance

Accurate assessment of tricuspid regurgitation (TR) is necessary for identification and risk stratification.

Objective

To design a deep learning computer vision workflow for identifying color Doppler echocardiogram videos and characterizing TR severity.

Design, Setting, and Participants

An automated deep learning workflow was developed using 47 312 studies (2 079 898 videos) from Cedars-Sinai Medical Center (CSMC) between 2011 and 2021. Data analysis was performed in 2024. The pipeline was tested on a temporally distinct test set of 2462 studies (108 138 videos) obtained in 2022 at CSMC and a geographically distinct cohort of 5549 studies (278 377 videos) from Stanford Healthcare (SHC). Training and validation cohorts contained data from 31 708 patients at CSMC receiving care between 2011 and 2021. Patients were chosen for parity across TR severity classes, with no exclusion criteria based on other clinical or demographic characteristics. The 2022 CSMC test cohort and SHC test cohorts contained studies from 2170 patients and 5014 patients, respectively.

Exposure

Deep learning computer vision model.

Main Outcomes and Measures

The main outcomes were area under the receiver operating characteristic curve (AUC), sensitivity, and specificity in identifying apical 4-chamber (A4C) videos with color Doppler across the tricuspid valve and AUC in identifying studies with moderate to severe or severe TR.

Results

In the CSMC test dataset, the view classifier demonstrated an AUC of 1.000 (95% CI, 0.999-1.000) and identified at least 1 A4C video with color Doppler across the tricuspid valve in 2410 of 2462 studies with a sensitivity of 0.975 (95% CI, 0.968-0.982) and a specificity of 1.000 (95% CI, 1.000-1.000). In the CSMC test cohort, moderate or severe TR was detected with an AUC of 0.928 (95% CI, 0.913-0.943), and severe TR was detected with an AUC of 0.956 (95% CI, 0.940-0.969). In the SHC cohort, the view classifier correctly identified at least 1 TR color Doppler video in 5268 of the 5549 studies, resulting in an AUC of 0.999 (95% CI, 0.998-0.999), a sensitivity of 0.949 (95% CI, 0.944-0.955), and a specificity of 0.999 (95% CI, 0.999-0.999). The artificial intelligence model detected moderate or severe TR with an AUC of 0.951 (95% CI, 0.938-0.962) and severe TR with an AUC of 0.980 (95% CI, 0.966-0.988).

Conclusions and Relevance

In this study, an automated pipeline was developed to identify clinically significant TR with excellent performance. With open-source code and weights, this project can serve as the foundation for future prospective evaluation of artificial intelligence-assisted workflows in echocardiography.

Artificial Intelligence in Diagnosis of Heart Failure

Heart failure (HF) is a complex and varied condition that affects over 50 million people worldwide. Although there have been significant strides in understanding the underlying mechanisms of HF, several challenges persist, particularly in the accurate diagnosis of HF. These challenges include issues related to its classification, the identification of specific phenotypes, and the assessment of disease severity. Artificial intelligence (AI) algorithms have the potential to transform HF care by enhancing clinical decision-making processes, enabling the early detection of patients at risk for subclinical or worsening HF. By integrating and analyzing vast amounts of data with intricate multidimensional interactions, AI algorithms can provide critical insights that help physicians make more timely and informed decisions. In this review, we explore the challenges in current diagnosis of HF, basic AI concepts and common AI algorithms, and latest AI research in HF diagnosis.

Multimodality Imaging of Arrhythmic Risk in Mitral Valve Prolapse

Mitral valve prolapse (MVP) affects 2% to 3% of the general population and is typically benign. However, a subset of patients may develop arrhythmic complications, including sudden cardiac arrest and sudden cardiac death. This review explores the critical role of multimodality imaging in risk stratification for arrhythmic MVP, emphasizing high-risk features such as bileaflet involvement, mitral annular disjunction, the double-peak strain pattern, mechanical dispersion, and myocardial fibrosis. Echocardiography remains the first-line imaging tool for MVP diagnosis, enabling detailed assessment of leaflet morphology, mitral annular disjunction, and mitral regurgitation quantification. Speckle tracking provides insights into abnormal valvular-myocardial mechanics as a potential arrhythmogenic mechanism in MVP. Cardiac magnetic resonance (CMR) offers detailed myocardial tissue characterization through assessment of replacement and interstitial fibrosis using late gadolinium enhancement and T1 mapping/extracellular volume fraction, respectively. Hybrid Positron Emission Tomography/CMR highlights the role of inflammation, which may coexist with fibrosis, in explaining the presence of malignant arrhythmias even with relatively limited fibrosis. The assessment of diffuse fibrosis and inflammation by CMR and Positron Emission Tomography/CMR is particularly valuable in patients without classic imaging risk factors such as mitral annular disjunction, severe mitral regurgitation, or replacement fibrosis. We propose an algorithm integrating clinical, rhythmic, echocardiographic, CMR, and Positron Emission Tomography/CMR parameters for arrhythmic risk stratification and management. Although multimodality imaging is essential for comprehensive risk assessment, most available parameters have not yet been validated in prospective studies nor linked directly to mortality. Consequently, these imaging findings should be interpreted alongside the presence of complex ventricular ectopy, which remains the most robust predictor of mortality in arrhythmic MVP.

Machine learning in point-of-care testing: innovations, challenges, and opportunities

The landscape of diagnostic testing is undergoing a significant transformation, driven by the integration of artificial intelligence (AI) and machine learning (ML) into decentralized, rapid, and accessible sensor platforms for point-of-care testing (POCT). The COVID-19 pandemic has accelerated the shift from centralized laboratory testing but also catalyzed the development of next-generation POCT platforms that leverage ML to enhance the accuracy, sensitivity, and overall efficiency of point-of-care sensors. This Perspective explores how ML is being embedded into various POCT modalities, including lateral flow assays, vertical flow assays, nucleic acid amplification tests, and imaging-based sensors, illustrating their impact through different applications. We also discuss several challenges, such as regulatory hurdles, reliability, and privacy concerns, that must be overcome for the widespread adoption of ML-enhanced POCT in clinical settings and provide a comprehensive overview of the current state of ML-driven POCT technologies, highlighting their potential impact in the future of healthcare.

2D echocardiography video to 3D heart shape reconstruction for clinical application

Transthoracic Echocardiography (TTE) is a crucial tool for assessing cardiac morphology and function quickly and non-invasively without ionising radiation. However, the examination is subject to intra- and inter-user variability and recordings are often limited to 2D imaging and assessments of end-diastolic and end-systolic volumes. We have developed a novel, fully automated machine learning-based framework to generate a personalised 4D (3D plus time) model of the left ventricular (LV) blood pool with high temporal resolution. A 4D shape is reconstructed from specific 2D echocardiographic views employing deep neural networks, pretrained on a synthetic dataset, and fine-tuned in a self-supervised manner using a novel optimisation method for cross-sectional imaging data. No 3D ground truth is needed for model training. The generated digital twins enhance the interpretation of TTE data by providing a versatile tool for automated analysis of LV volume changes, localisation of infarct areas, and identification of new and clinically relevant biomarkers. Experiments are performed on a multicentre dataset that includes TTE exams of 144 patients with normal TTE and 314 patients with acute myocardial infarction (AMI). The novel biomarkers show a high predictive value for survival (area under the curve (AUC) of 0.82 for 1-year all-cause mortality), demonstrating that personalised 3D shape modelling has the potential to improve diagnostic accuracy and risk assessment.

Multiple token rearrangement Transformer network with explicit superpixel constraint for segmentation of echocardiography

Diagnostic cardiologists have considerable clinical demand for precise segmentation of echocardiography to diagnose cardiovascular disease. The paradox is that manual segmentation of echocardiography is a time-consuming and operator-dependent task. Computer-aided segmentation can reduce the workflow greatly. However, it is challenging to segment multi-type echocardiography, which is reflected in differential anatomic structures, artifacts, and blurred borderline. This study proposes the multiple token rearrangement Transformer network (MTRT-Net) embedded in three novel modules to address the corresponding three challenges. First, the depthwise deformable attention module can extract flexible features to adapt to anatomic structures of echocardiography with different ages and diseases. Second, the superpixel supervised module can cluster similar features and keep discriminative features away to make the segmentation regions tend to be an entire body. The artifacts have the influence in separating the complete internal region. Third, the atrous affinity aggregation module can integrate affinity features near the borderline to judge the blurred regions. Overall, the three modules rearrange the relationships of tokens and broaden the diversity of features. Besides, the explicit constraint brought by the superpixel supervised module enhances the performance of fitting ability. This study has 13747 echocardiography to train and test the MTRT-Net. Abundant experiments also validate the performance of MTRT-Net. Therefore, MTRT-Net can assist the diagnostician in segmenting the echocardiography precisely.

Machine Learning to Predict Outcomes of Fetal Cardiac Disease: A Pilot Study

Prediction of outcomes following a prenatal diagnosis of congenital heart disease (CHD) is challenging. Machine learning (ML) algorithms may be used to reduce clinical uncertainty and improve prognostic accuracy. We performed a pilot study to train ML algorithms to predict postnatal outcomes based on clinical data. Specific objectives were to predict (1) in utero or neonatal death, (2) high-acuity neonatal care and (3) favorable outcomes. We included all fetuses with cardiac disease at Sunnybrook Health Sciences Centre, Toronto, Canada, from 2012 to 2021. Prediction models were created using the XgBoost algorithm (tree-based) with fivefold cross-validation. Among 211 cases of fetal cardiac disease, 61 were excluded (39 terminations, 21 lost to follow-up, 1 isolated arrhythmia), leaving a cohort of 150 fetuses. Fifteen (10%) demised (10 neonates) and 65 (48%) of live births required high acuity neonatal care. Of those with clinical follow-up, 60/87 (69%) had a favorable outcome. Prediction models for fetal or neonatal death, high acuity neonatal care and favorable outcome had AUCs of 0.76, 0.84 and 0.73, respectively. The most important predictors for death were the presence of non-cardiac abnormalities combined with more severe CHD. High acuity of postnatal care was predicted by anti Ro antibody and more severe CHD. Favorable outcome was most predicted by no right heart disease combined with genetic abnormalities, and maternal medications. Prediction models using ML provide good discrimination of key prenatal and postnatal outcomes among fetuses with congenital heart disease.

Applications, challenges and future directions of artificial intelligence in cardio-oncology

BACKGROUND

The management of cardiotoxicity related to cancer therapies has emerged as a significant clinical challenge, prompting the rapid growth of cardio-oncology. As cancer treatments become more complex, there is an increasing need to enhance diagnostic and therapeutic strategies for managing their cardiovascular side effects.

OBJECTIVE

This review investigates the potential of artificial intelligence (AI) to revolutionize cardio-oncology by integrating diverse data sources to address the challenges of cardiotoxicity management.

METHODS

We explore applications of AI in cardio-oncology, focusing on its ability to leverage multiple data sources, including electronic health records, electrocardiograms, imaging modalities, wearable sensors, and circulating serum biomarkers.

RESULTS

AI has demonstrated significant potential in improving risk stratification and longitudinal monitoring of cardiotoxicity. By optimizing the use of electrocardiograms, non-invasive imaging, and circulating biomarkers, AI facilitates earlier detection, better prediction of outcomes, and more personalized therapeutic interventions. These advancements are poised to enhance patient outcomes and streamline clinical decision-making.

CONCLUSIONS

AI represents a transformative opportunity in cardio-oncology by advancing diagnostic and therapeutic capabilities. However, successful implementation requires addressing practical challenges such as data integration, model interpretability, and clinician training. Continued collaboration between clinicians and AI developers will be essential to fully integrate AI into routine clinical workflows.

Ethical challenges and evolving strategies in the integration of artificial intelligence into clinical practice

Artificial intelligence (AI) has rapidly transformed various sectors, including healthcare, where it holds the potential to transform clinical practice and improve patient outcomes. However, its integration into medical settings brings significant ethical challenges that need careful consideration. This paper examines the current state of AI in healthcare, focusing on five critical ethical concerns: justice and fairness, transparency, patient consent and confidentiality, accountability, and patient-centered and equitable care. These concerns are particularly pressing as AI systems can perpetuate or even exacerbate existing biases, often resulting from non-representative datasets and opaque model development processes. The paper explores how bias, lack of transparency, and challenges in maintaining patient trust can undermine the effectiveness and fairness of AI applications in healthcare. In addition, we review existing frameworks for the regulation and deployment of AI, identifying gaps that limit the widespread adoption of these systems in a just and equitable manner. Our analysis provides recommendations to address these ethical challenges, emphasizing the need for fairness in algorithm design, transparency in model decision-making, and patient-centered approaches to consent and data privacy. By highlighting the importance of continuous ethical scrutiny and collaboration between AI developers, clinicians, and ethicists, we outline pathways for achieving more responsible and inclusive AI implementation in healthcare. These strategies, if adopted, could enhance both the clinical value of AI and the trustworthiness of AI systems among patients and healthcare professionals, ensuring that these technologies serve all populations equitably.

[Precision medicine in the diagnostics and treatment of cardiomyopathies : State of the art]

Cardiomyopathies (CMP) comprise a group of heterogeneous heart muscle diseases that have molecular genetic causes and cannot be adequately explained by other cardiovascular diseases. The diagnosis and treatment of CMPs have made significant progress in recent decades, which is reflected in novel specific cardiomyopathy guidelines of the European Society of Cardiology (ESC) published in 2023. A patient-centered approach combines multimodal diagnostics, such as echocardiography, magnetic resonance imaging, genetic testing and biopsy to enable precise etiological classification and thus personalized treatment. Existing treatment options have been expanded through innovative treatments. Mavacamten treats left ventricular outflow tract (LVOT) obstruction in hypertrophic CMP, while tafamidis and RNA-based treatment, such as patisiran and vutrisiran specifically affect transthyretin-mediated amyloidosis. Advances in gene therapy open up new perspectives. Artificial intelligence (AI) is pivotal in precision medicine where AI-assisted analyses enhance the diagnosis of subclinical diseases, optimize imaging modalities and accelerate the development of new treatment approaches. The ESC guidelines are an important milestone in the care of patients with CMP, while also emphasizing the need for further research and scientific ethical discussions, especially with respect to AI and innovative forms of treatment.

Automated generation of echocardiography reports using artificial intelligence: a novel approach to streamlining cardiovascular diagnostics

Accurate interpretation of echocardiography measurements is essential for diagnosing cardiovascular diseases and guiding clinical management. The emergence of large language models (LLMs) like ChatGPT presents a novel opportunity to automate the generation of echocardiography reports and provide clinical recommendations. This study aimed to evaluate the ability of an LLM (ChatGPT) to 1) generate comprehensive echocardiography reports based solely on provided echocardiographic measurements, and when enriched with clinical information 2) formulate accurate diagnoses, along with appropriate recommendations for further tests, treatment, and follow-up. Echocardiographic data from n = 13 fictional cases (Group 1) and n = 8 clinical cases (Group 2) were input into the LLM. The model's outputs were compared against standard clinical assessments conducted by experienced cardiologists. Using a dedicated scoring system, the LLM's performance was evaluated and stratified based on its accuracy in report generation, diagnostic precision, and the appropriateness of its recommendations. Patterns, frequency and examples of misinterpretations by LLM were analysed. Across all cases, mean total score was 6.86 (SD = 1.12). Group 1 had a mean total score of 6.54 (SD = 1.13) and accuracy of 3.92 (SD = 0.86), while Group 2 scored 7.38 (SD = 0.92) and 4.38 (SD = 0.92), respectively. Recommendations were 2.62 (SD = 0.51) for Group 1 and 3.00 (SD = 0.00) for Group 2, with no significant differences (p = 0.096). Fully acceptable reports were 85.7%, borderline acceptable 14.3%, and none were not acceptable. Of 299 parameters, 5.3% were misinterpreted. The LLM demonstrated a high level of accuracy in generating detailed echocardiography reports, mostly correctly identifying normal and abnormal findings, and making accurate diagnoses across a range of cardiovascular conditions. ChatGPT, as an LLM, shows significant potential in automating the interpretation of echocardiographic data, offering accurate diagnostic insights and clinical recommendations. These findings suggest that LLMs could serve as valuable tools in clinical practice, assisting and streamlining clinical workflow.

SonoGif.com: A Free Online Tool to Remove Protected Health Information From Any Ultrasound Clip

BACKGROUND

The proliferation of online medical education, particularly in point-of-care ultrasound (POCUS), has been limited by challenges in removing protected health information (PHI) from imaging data. These challenges include PHI embedded in both metadata and within the ultrasound images themselves, complicating compliance with HIPAA standards.

OBJECTIVE

To describe the development and functionality of SonoGif.com, a free, browser-based tool designed to facilitate the de-identification and sharing of ultrasound clips without the need for specialized software.

METHODS

SonoGif was developed using JavaScript to run entirely within a web browser, preserving data privacy by ensuring ultrasound clips remain on the user's device during initial processing. DICOM files are parsed using the open-source dicomParser library, while standard video formats are rendered with native HTML5 Canvas APIs. Users can manually annotate images to obscure on-screen PHI. The resulting de-identified frames are transmitted to a secure server, where FFmpeg compiles them into shareable video formats.

RESULTS

Since its public release in 2019, SonoGif has been used to de-identify over 3000 ultrasound clips by users worldwide, including those in low-resource settings. Its accessibility, simplicity, and adherence to privacy regulations have made it a valuable tool for medical educators and clinicians seeking to share ultrasound media for teaching and research.

CONCLUSION

SonoGif is a free web-based application that allows for easy and secure removal of PHI from ultrasound media. It broadens global access to ultrasound education by eliminating technical barriers and enabling safe image sharing across diverse clinical and educational environments. The platform is available at https://sonogif.com with source code accessible on GitHub.

Artificial Intelligence in Nuclear Cardiac Imaging: Novel Advances, Emerging Techniques, and Recent Clinical Trials

Cardiovascular disease (CVD) is a leading cause of death, accounting for over 30% of annual global fatalities. Ischemic heart disease, in turn, is a frontrunner of worldwide CVD mortality. With the burden of coronary disease rapidly growing, understanding the nuances of cardiac imaging and risk prognostication becomes paramount. Myocardial perfusion imaging (MPI) is a frequently utilized and well established testing modality due to its significant clinical impact in disease diagnosis and risk assessment. Recently, nuclear cardiology has witnessed major advancements, driven by innovations in novel imaging technologies and improved understanding of cardiovascular pathophysiology. Applications of artificial intelligence (AI) to MPI have enhanced diagnostic accuracy, risk stratification, and therapeutic decision-making in patients with coronary artery disease (CAD). AI techniques such as machine learning (ML) and deep learning (DL) neural networks offer new interpretations of immense data fields, acquired through cardiovascular imaging modalities such as nuclear medicine (NM). Recently, AI algorithms have been employed to enhance image reconstruction, reduce noise, and assist in the interpretation of complex datasets. The rise of AI in nuclear medicine (AI-NM) has proven itself groundbreaking in the efficiency of image acquisition, post-processing time, diagnostic ability, consistency, and even in risk-stratification and outcome prognostication. To that end, this narrative review will explore these latest advances in AI in nuclear medicine and its rapid transformation of the cardiac diagnostics landscape. This paper will examine the evolution of AI-NM, review novel AI techniques and applications in nuclear cardiac imaging, summarize recent AI-NM clinical trials, and explore the technical and clinical challenges in its implementation of artificial intelligence.

Deep-learning based multibeat echocardiographic cardiac phase detection

BACKGROUND

End-to-end automatic detection of cardiac phase in multibeat echocardiograms is crucial for measuring cardiac parameters in clinical scenarios. However, existing studies face limitations due to the high cost of data annotation and collection, and time-consuming detection processes.

PURPOSE

This study introduces a novel multibeat echocardiographic cardiac phase detection network, EchoPhaseNet, to perform fast and accurate cardiac phase detection of variable-length multibeat echocardiographic sequences, with low annotation costs and limited data.

MATERIALS AND METHODS

Five echocardiographic datasets were used in this study, including a small-scale private dataset, Echo-DT (DrumTower), a medium-scale publicly available dataset, PhaseDetection, and three additional publicly available datasets: EchoNet-Dynamic, CAMUS, and EchoNet-Dynamic-MultiBeat. EchoPhaseNet and four other deep learning-based cardiac phase detection methods were trained and internally validated on the Echo-DT and PhaseDetection datasets (with sample ratios for training, validation, and testing set at 60%:20%:20% and 80%:0%:20%, respectively), and then externally validated on the other three datasets. Model performance was evaluated using GradCAM for qualitative visualization and absolute frame difference (aFD) for quantitative accuracy, with statistical significance assessed using Tukey's test and Benjamini-Hochberg correction, considering corrected p-values < $<$ 0.05 as significant.

RESULTS

The annotation costs and accuracy of end-diastolic (ED) and end-systolic (ES) detection using EchoPhaseNet were compared with those of four other comparison methods. EchoPhaseNet achieves effective specific phase detections using only ED/ES labels, reducing annotation costs and making it applicable to a wider range of detection scenarios compared to all the comparison methods. On the Echo-DT dataset, EchoPhaseNet's mean aFD values for ED and ES detection in the A4C view samples were 1.08 and 1.04, respectively, significantly outperforming three comparison methods in ED detection accuracy (p-values < $<$ 0.01) and comparable to the remaining one (p-values > $>$ 0.05). On the PhaseDetection dataset, EchoPhaseNet's mean aFD values for ED and ES detection were 1.67 and 2.19, respectively, comparable to the detection accuracies of all four comparison methods (p-values > $>$ 0.05). In addition, EchoPhaseNet showed strong generalization ability on multiple external validation datasets. After training on the small-scale Echo-DT dataset, EchoPhaseNet significantly outperformed the four comparison methods (p-values < $<$ 0.01) in ED detection, achieving mean aFD values of 1.67 and 1.11 on the EchoNet-Dynamic and EchoNet-Dynamic-MultiBeat datasets, respectively. After training on the PhaseDetection dataset, EchoPhaseNet significantly outperformed the four compared methods (p-values < $<$ 0.01) in ES detection on the EchoNet-Dynamic dataset, achieving mean aFD value of 2.58. EchoPhaseNet's inference time for a single 32-frame sequence segment is substantially lower than that of the four compared methods, not exceeding 8 ms on an RTX 4080 GPU using the PyTorch deep learning framework.

CONCLUSIONS

EchoPhaseNet exhibits clear advantages over existing studies in data annotation and collection costs, as well as detection speed, and is applicable to a wider range of detection scenarios. It demonstrates good practicality and promising prospects for clinical multibeat echocardiographic cardiac phase detection.

Artificial intelligence automation of echocardiographic measurements

Background

Accurate measurement of echocardiographic parameters is crucial for the diagnosis of cardiovascular disease and tracking of change over time, however manual assessment is time-consuming and can be imprecise. Artificial intelligence (AI) has the potential to reduce clinician burden by automating the time-intensive task of comprehensive measurement of echocardiographic parameters.

Methods

We developed and validated open-sourced deep learning semantic segmentation models for the automated measurement of 18 anatomic and Doppler measurements in echocardiography. The outputs of segmentation models were compared to sonographer measurements from two institutions to access accuracy and precision.

Results

We utilized 877,983 echocardiographic measurements from 155,215 studies from Cedars-Sinai Medical Center (CSMC) to develop EchoNet-Measurements, an open-source deep learning model for echocardiographic annotation. The models demonstrated a good correlation when compared with sonographer measurements from held-out data from CSMC and an independent external validation dataset from Stanford Healthcare (SHC). Measurements across all nine B-mode and nine Doppler measurements had high accuracy (an overall R2 of 0.967 (0.965 - 0.970) in the held-out CSMC dataset and 0.987 (0.984 - 0.989) in the SHC dataset). When evaluated end-to-end on a temporally distinct 2,103 studies at CSMC, EchoNet-Measurements performed well an overall R2 of 0.981 (0.976 - 0.984). Performance was consistent across patient characteristics including sex and atrial fibrillation status.

Conclusion

EchoNet-Measurement achieves high accuracy in automated echocardiographic measurement that is comparable to expert sonographers. This open-source model provides the foundation for future developments in AI applied to echocardiography.

Automated Aortic Regurgitation Detection and Quantification: A Deep Learning Approach Using Multi-View Echocardiography

Background

Accurate evaluation of aortic regurgitation (AR) severity is necessary for early detection and chronic disease management. AR is most commonly assessed by Doppler echocardiography, however limitations remain given variable image quality and need to integrate information from multiple views. This study developed and validated a deep learning model for automated AR severity assessment from multi-view color Doppler videos.

Methods

We developed a video-based convolutional neural network (R2+1D) to classify AR severity using color Doppler echocardiography videos from five standard views: parasternal long-axis (PLAX), PLAX-aortic valve focus, apical three-chamber (A3C), A3C-aortic valve focus, and apical five-chamber (A5C). The model was trained on 47,638 videos from 32,396 studies (23,240 unique patients) from Cedars-Sinai Medical Center (CSMC) and externally validated on 3369 videos from 1504 studies (1493 unique patients) from Stanford Healthcare Center (SHC).

Results

Combining assessments from multiple views, the EchoNet-AR model achieved excellent identification of both at least moderate AR (AUC 0.95, [95% CI 0.94-0.96]) and severe AR (AUC 0.97, [95% CI 0.96 - 0.98]). This performance was consistent in the external SHC validation cohort for both at least moderate AR (AUC 0.92, [95% CI 0.88-0.96]) and severe AR (AUC 0.94, [95% CI 0.89-0.98]). Subgroup analysis showed robust model performance across varying image quality, valve morphologies, and patient demographics. Saliency map visualizations demonstrated that the model focused on the proximal flow convergence zone and vena contracta, appropriately narrowing on hemodynamically significant regions.

Conclusion

The EchoNet-AR model accurately classifies AR severity and synthesizes information across multiple echocardiographic views with robust generalizability in an external cohort. The model shows potential as an automated clinical decision support tool for AR assessment, however clinical interpretation remains essential, particularly in complex cases with multiple valve pathologies or altered hemodynamics.

AI Learning for Pediatric Right Ventricular Assessment: Development and Validation Across Multiple Centers

Background

Congenital and acquired heart disease affects ∼1% of children globally, with right ventricular (RV) dysfunction being a common and complex issue due to conditions like congenital heart disease (CHD), pulmonary hypertension (PH), and prematurity. Accurate RV assessment is challenging due to its unique geometry, interventricular interactions, and morphological variability in pediatric patients. Fractional area change (FAC), a key echocardiographic measure, correlates strongly with disease severity, aiding in timely intervention and prognosis. AI learning shows the potential to automate and standardize RV assessments, overcoming traditional limitations and improving early diagnosis and management of pediatric cardiovascular disorders.

Methods

Using 24,984 echocardiograms from 3,993 pediatric patients across four tertiary care centers (one in North America, three in Asia), we developed and validated an AI framework for automated RV assessment. The framework employs multi-task learning to perform ventricular segmentation, beat-by-beat quantification of RV FAC, and identification of cardiac abnormalities like PH. It was also extended to enhance left ventricular (LV) functional assessment.

Findings

Our AI system achieved Dice similarity coefficients of 0.86 (apical-four-chamber, A4C) and 0.88 (parasternal-short-axis, PSAX) for RV segmentation, matching expert annotations. It demonstrated robust RV functional assessment, with AUCs of 0.95 (U.S. cohort) and 0.97 (Asian cohort). For PH classification, diagnostic accuracies were 0.95 (U.S.) and 0.94 (Asian), confirming consistent performance across populations. When extended to LV assessment, the framework significantly improved LV ejection fraction (EF) prediction in both U.S. and Asian cohorts.

Interpretation

This validated AI framework enables reliable, automated ventricular function analysis, matching expert-level performance. By enhancing clinical workflows and standardizing pediatric cardiac assessments, it has the potential to improve care management for pediatric cardiovascular disorders, particularly in resource-limited settings.

Funding

This work was supported by the U.S. NIH 1R41HL160362-01 to XBL and K23HL150279 to AT.

Enhancing Left Ventricular Segmentation in Echocardiograms Through GAN-Based Synthetic Data Augmentation and MultiResUNet Architecture

Background: Accurate segmentation of the left ventricle in echocardiograms is crucial for the diagnosis and monitoring of cardiovascular diseases. However, this process is hindered by the limited availability of high-quality annotated datasets and the inherent complexities of echocardiogram images. Traditional methods often struggle to generalize across varying image qualities and conditions, necessitating a more robust solution. Objectives: This study aims to enhance left ventricular segmentation in echocardiograms by developing a framework that integrates Generative Adversarial Networks (GANs) for synthetic data augmentation with a MultiResUNet architecture, providing a more accurate and reliable segmentation method. Methods: We propose a GAN-based framework that generates synthetic echocardiogram images and their corresponding segmentation masks, augmenting the available training data. The synthetic data, along with real echocardiograms from the EchoNet-Dynamic dataset, were used to train the MultiResUNet architecture. MultiResUNet incorporates multi-resolution blocks, residual connections, and attention mechanisms to effectively capture fine details at multiple scales. Additional enhancements include atrous spatial pyramid pooling (ASPP) and scaled exponential linear units (SELUs) to further improve segmentation accuracy. Results: The proposed approach significantly outperforms existing methods, achieving a Dice Similarity Coefficient of 95.68% and an Intersection over Union (IoU) of 91.62%. This represents improvements of 2.58% in Dice and 4.84% in IoU over previous segmentation techniques, demonstrating the effectiveness of GAN-based augmentation in overcoming data scarcity and improving segmentation performance. Conclusions: The integration of GAN-generated synthetic data and the MultiResUNet architecture provides a robust and accurate solution for left ventricular segmentation in echocardiograms. This approach has the potential to enhance clinical decision-making in cardiovascular medicine by improving the accuracy of automated diagnostic tools, even in the presence of limited and complex training data.

Predicting Cardiac Magnetic Resonance-Derived Ejection Fraction from Echocardiogram Via Deep Learning Approach in Tetralogy of Fallot

Systolic function assessment is essential in children with congenital heart disease. Traditional methods of echocardiographic left ventricular ejection fraction (LVEF) estimation might overestimate systolic function compared to the gold standard of cardiac magnetic resonance imaging (CMR), especially in Tetralogy of Fallot (TOF). Deep learning technologies such as EchoNet-Dynamic offer more consistent cardiac evaluations and can potentially accurately predict LVEF using echocardiographic videos. The EchoNet-Dynamic/EchoNet-Peds models predict LVEF using echocardiograms with expert-measured LVEF as the ground truth. Using a transfer learning approach, we fine-tuned this model to predict LVEF with CMR-derived LVEF as ground truth and TOF echocardiograms as input images. For echocardiograms in the PSAX view, the model predicted CMR LVEF with an R2 of 0.79 and an MAE of 4.41. For the A4C view, the model predicted CMR LVEF with an R2 of 0.53 and an MAE of 6.4. Plotted ROC curves indicate that both tuned models differentiated well between normal and reduced LVEF. This study shows the potential of Convolutional Neural Network (CNN) models in transforming the field of cardiac imaging interpretation via a hybrid approach using the CMR labels and echocardiogram videos offering advancements over conventional methods.

Guideline for the Use of Natriuretic Peptides in the Early Diagnosis and Management of Heart Failure in Primary Care (Joint Consensus Report by the Eurasian Society of Heart Failure and the Turkish Association of Family Medicine)

Heart failure (HF) is a complex clinical condition associated with significant morbidity and mortality. Early diagnosis and effective management at the primary care level are essential for improving patient outcomes and reducing the burden on the healthcare systems. The Eurasian Society of HF and the Turkish Association of Family Medicine developed a guideline that underscores the critical role of natriuretic peptides (NPs) in the early detection, diagnosis, and management of HF. NPs, particularly the N-terminal pro-B-type NP, are a reliable biomarker for identifying HF, guiding treatment decisions, and monitoring disease progression. This guideline emphasizes the importance of measuring the levels of these peptides in primary care so as to detect individuals at risk, confirm the diagnosis of HF in symptomatic patients, and evaluate the treatment response. The recommended thresholds for NP levels account for variations arising from factors such as age, gender, and the presence of other health conditions. B-type natriuretic peptides (BNP) levels ≥ 35 pg/ml or N-terminus-proBNP levels ≥ 125 pg/ml are used to confirm the likelihood of HF in symptomatic patients, enabling timely diagnosis and appropriate intervention. Incorporating NP testing into routine clinical practice enables timely referrals and ensures appropriate management at all stages of HF. Beyond diagnosis, the measurement of NPs provides valuable information about treatment effectiveness and prognosis, allowing clinicians to individualize the treatment. By integrating NP testing into primary care, healthcare providers can facilitate early detection, optimize treatment strategies, and improve the quality of life for patients with or at risk of HF. Thus, this guideline highlights the essential role of primary care physicians in addressing the growing challenges of HF through the effective and evidence-based use of NPs.

Toward Intelligent Head Impulse Test: A Goggle-Free Approach Using a Monocular Infrared Camera

OBJECTIVES

To assess vestibular function, video head impulse test (vHIT) is taken as the gold standard by evaluating the vestibulo-ocular reflex (VOR). However, vHIT requires the patient to wear a specialized head-mounted goggle equipment that needs to be calibrated before each use. For this, we proposed an intelligent head impulse test (iHIT) setting with a monocular infrared camera instead of the head-mounted goggle and contributed correspondingly a video classification approach with deep learning to vestibular function determination.

METHODS

Within the iHIT framework, a monocular infrared camera was set in front of the patient to capture test videos, based on which a dataset DiHIT of HIT video clips was set up. We then proposed a two-stage multi-modal video classification network, trained on the dataset DiHIT, that took as input the eye motion and head motion data extracted from the facial keypoints via HIT clips and outputted the identification of the semicircular canal (SCC) being tested (SCC identification) and determination of VOR abnormality (SCC qualitation).

RESULTS

Experiments on this dataset DiHIT showed that it achieved the accuracy of 100% in prediction of SCC identification. Furthermore, it attained predictive accuracies of 84.1% in horizontal and 79.0% in vertical SCC qualitation.

CONCLUSIONS

Compared with existing video-based HIT, iHIT eliminates goggles, does not require equipment calibration, and achieves complete automation. Furthermore, iHIT will bring more benefits to users due to its low cost and ease of operation. Codes and use case pipeline are available at: https://github.com/dec1st2023/iHIT.

LEVEL OF EVIDENCE

3 Laryngoscope, 135:1161-1168, 2025.

WFUMB Commentary Paper on Artificial intelligence in Medical Ultrasound Imaging

Artificial intelligence (AI) is defined as the theory and development of computer systems able to perform tasks normally associated with human intelligence. At present, AI has been widely used in a variety of ultrasound tasks, including in point-of-care ultrasound, echocardiography, and various diseases of different organs. However, the characteristics of ultrasound, compared to other imaging modalities, such as computed tomography (CT) and magnetic resonance imaging (MRI), poses significant additional challenges to AI. Application of AI can not only reduce variability during ultrasound image acquisition, but can standardize these interpretations and identify patterns that escape the human eye and brain. These advances have enabled greater innovations in ultrasound AI applications that can be applied to a variety of clinical settings and disease states. Therefore, The World Federation of Ultrasound in Medicine and Biology (WFUMB) is addressing the topic with a brief and practical overview of current and potential future AI applications in medical ultrasound, as well as discuss some current limitations and future challenges to AI implementation.

Intraoperative Real-Time IDH Diagnosis for Glioma Based on Automatic Analysis of Contrast-Enhanced Ultrasound Video

OBJECTIVE

Isocitrate dehydrogenase (IDH) is the most important molecular marker of glioma and is highly correlated to the diagnosis, treatment, and prognosis of patients. We proposed a real-time diagnosis method for IDH status differentiation based on automatic analysis of intraoperative contrast-enhanced ultrasound (CEUS) video.

METHODS

Inspired by the Time Intensity Curve (TIC) analysis of CEUS utilized in clinical practice, this paper proposed an automatic CEUS video analysis method called ATAN (Automatic TIC Analysis Network). Based on tumor identification, ATAN automatically selected ROIs (region of interest) inside and outside glioma. ATAN ensures the integrity of dynamic features of perfusion changes at critical locations, resulting in optimal diagnostic performance. The transfer learning mechanism was also introduced by using two auxiliary CEUS datasets to solve the small sample problem of intraoperative glioma data.

RESULTS

Through pretraining on 258 patients on two auxiliary cohorts, ATAN produced the IDH diagnosis with accuracy and AUC of 0.9 and 0.91 respectively on the main cohort of 60 glioma patients (mean age, 50 years ± 14, 28 men) Compared with other existing IDH status differentiation methods, ATAN is a real-time IDH diagnosis method without the need of tumor samples.

CONCLUSION

ATAN is an effective automatic analysis model of CEUS, with the help of this model, real-time intraoperative diagnosis of IDH with high accuracy can be achieved. Compared with other state-of-the-art deep learning methods, the accuracy of the ATAN model is 15% higher on average.

Role of Artificial Intelligence in Congenital Heart Disease and Interventions

Artificial intelligence has promising impact on patients with congenital heart disease, a vulnerable population with life-long health care needs and, often, a substantially higher risk of death than the general population. This review explores the role artificial intelligence has had on cardiac imaging, electrophysiology, interventional procedures, and intensive care monitoring as it relates to children and adults with congenital heart disease. Machine learning and deep learning algorithms have enhanced not only imaging segmentation and processing but also diagnostic accuracy namely reducing interobserver variability. This has a meaningful impact in complex congenital heart disease improving anatomic diagnosis, assessment of cardiac function, and predicting long-term outcomes. Image processing has benefited procedural planning for interventional cardiology, allowing for a higher quality and density of information to be extracted from the same imaging modalities. In electrophysiology, deep learning models have enhanced the diagnostic potential of electrocardiograms, detecting subtle yet meaningful variation in signals that enable early diagnosis of cardiac dysfunction, risk stratification of mortality, and more accurate diagnosis and prediction of arrhythmias. In the congenital heart disease population, this has the potential for meaningful prolongation of life. Postoperative care in the cardiac intensive care unit is a data-rich environment that is often overwhelming. Detection of subtle data trends in this environment for early detection of morbidity is a ripe avenue for artificial intelligence algorithms to be used. Examples like early detection of catheter-induced thrombosis have already been published. Despite their great promise, artificial intelligence algorithms are still limited by hurdles such as data standardization, algorithm validation, drift, and explainability.

The Honest Enterprise Research Broker: Facilitating Ethical, Efficient, and Secure Access to Health Data for Research

BACKGROUND

 Health systems generate and store vast amounts of clinical data, requiring structured processes to ensure that secondary use of the data is available to researchers in an efficient, ethical, and secure manner.

OBJECTIVE

 We describe a process to provide data to health system researchers at a large, multi-hospital health system that balances efficiency with security and ethics.

METHODS

 The Honest Enterprise Research Broker (HERB) Committee has enacted a systematic process to deliver investigators requesting data, using pre-written SQL code, when possible, to increase efficiency, providing a suite of self-service tools for cohort size estimation, and assessing the security and privacy of data leaving the institution. We evaluated the number of extracts per year, the average time to delivery of the data extract, and user satisfaction with the process.

RESULTS

 From 2018 to 2023, the HERB Committee completed 487 data extracts. The number of requests increased from 51 in 2018 to 121 in 2023. Even as the number of extracts increased, the number of hours per extract decreased from 12.5 in 2018 to 9 in 2023. User satisfaction surveys found a high degree of satisfaction with the process.

CONCLUSION

 Through a process of continuous improvement, the HERB Committee has developed an expedient process to support the research needs of a large academic multi-hospital health system.

Leveraging Comprehensive Echo Data to Power Artificial Intelligence Models for Handheld Cardiac Ultrasound

Objective

To develop a fully end-to-end deep learning framework capable of estimating left ventricular ejection fraction (LVEF), estimating patient age, and classifying patient sex from echocardiographic videos, including videos collected using handheld cardiac ultrasound (HCU).

Patients and Methods

Deep learning models were trained using retrospective transthoracic echocardiography (TTE) data collected in Mayo Clinic Rochester and surrounding Mayo Clinic Health System sites (training: 6432 studies and internal validation: 1369 studies). Models were then evaluated using retrospective TTE data from the 3 Mayo Clinic sites (Rochester, n=1970; Arizona, n=1367; Florida, n=1562) before being applied to a prospective dataset of handheld ultrasound and TTE videos collected from 625 patients. Study data were collected between January 1, 2018 and February 29, 2024.

Results

Models showed strong performance on the retrospective TTE datasets (LVEF regression: root mean squared error (RMSE)=6.83%, 6.53%, and 6.95% for Rochester, Arizona, and Florida cohorts, respectively; classification of LVEF ≤40% versus LVEF > 40%: area under curve (AUC)=0.962, 0.967, and 0.980 for Rochester, Arizona, and Florida, respectively; age: RMSE=9.44% for Rochester; sex: AUC=0.882 for Rochester), and performed comparably for prospective HCU versus TTE data (LVEF regression: RMSE=6.37% for HCU vs 5.57% for TTE; LVEF classification: AUC=0.974 vs 0.981; age: RMSE=10.35% vs 9.32%; sex: AUC=0.896 vs 0.933).

Conclusion

Robust TTE datasets can be used to effectively power HCU deep learning models, which in turn demonstrates focused diagnostic images can be obtained with handheld devices.

M4S-Net: a motion-enhanced shape-aware semi-supervised network for echocardiography sequence segmentation

Sequence segmentation of echocardiograms is of great significance for the diagnosis and treatment of cardiovascular diseases. However, the low quality of ultrasound imaging and the complexity of cardiac motion pose great challenges to it. In addition, the difficulty and cost of labeling echocardiography sequences limit the performance of supervised learning methods. In this paper, we proposed a Motion-enhanced Shape-aware Semi-supervised Sequence Segmentation Network named M4S-Net. First, multi-level shape priors are used to enhance the model's shape representation capabilities, overcoming the low image quality and improving single-frame segmentation. Then, a motion-enhanced optimization module utilizes optical flows to assist segmentation in a geometric sense, which robustly responds to the complex motions and ensures the temporal consistency of sequence segmentation. A hybrid loss function is devised to maximize the effectiveness of each module and further improve the temporal stability of predicted masks. Furthermore, the parameter-sharing strategy allows it to perform sequence segmentation in a semi-supervised manner. Massive experiments on both public and in-house datasets show that M4S-Net outperforms the state-of-the-art methods in both spatial and temporal segmentation performance. A downstream apical rocking recognition task based on M4S-Net also achieves an AUC of 0.944, which significantly exceeds specialized physicians.

AI-echocardiography: Current status and future direction

Echocardiography, which provides detailed evaluations of cardiac structure and pathology, is central to cardiac imaging. Traditionally, the assessment of disease severity, treatment effectiveness, and prognosis prediction relied on detailed parameters obtained by trained sonographers and the expertise of specialists, which can limit access and availability. Recent advancements in deep learning and large-scale computing have enabled the automatic acquisition of parameters in a short time using vast amounts of historical training data. These technologies have been shown to predict the presence of diseases and future cardiovascular events with or without relying on quantitative parameters. Additionally, with the advent of large-scale language models, zero-shot prediction that does not require human labeling and automatic echocardiography report generation are also expected. The field of AI-enhanced echocardiography is poised for further development, with the potential for more widespread use in routine clinical practice. This review discusses the capabilities of deep learning models developed using echocardiography, their limitations, current applications, and research utilizing generative artificial intelligence technologies.

Artificial Intelligence Applications in Cardio-Oncology: A Comprehensive Review

PURPOSE OF REVIEW

This review explores the role of artificial intelligence (AI) in cardio-oncology, focusing on its latest application across problems in diagnosis, prognosis, risk stratification, and management of cardiovascular (CV) complications in cancer patients. It also highlights multi-omics analysis, explainable AI, and real-time decision-making, while addressing challenges like data heterogeneity and ethical concerns.

RECENT FINDINGS

AI can advance cardio-oncology by leveraging imaging, electronic health records (EHRs), electrocardiograms (ECG), and multi-omics data for early cardiotoxicity detection, stratification and long-term risk prediction. Novel AI-ECG models and imaging techniques improve diagnostic accuracy, while multi-omics analysis identifies biomarkers for personalized treatment. However, significant barriers, including data heterogeneity, lack of transparency, and regulatory challenges, hinder widespread adoption. AI significantly enhances early detection and intervention in cardio-oncology. Future efforts should address the impact of AI technologies on clinical outcomes, and ethical challenges, to enable broader clinical adoption and improve patient care.

Silk Fibroin Hydrogel for Pulse Waveform Precise and Continuous Perception

Precise and continuous monitoring of blood pressure and cardiac function is of great importance for early diagnosis and timely treatment of cardiovascular diseases. The common tests rely on on-site diagnosis and bulky equipments, hindering early diagnosis. The emerging hydrogels have gained considerable attention in skin bioelectronics by virtue of the similarities to biological tissues and versatility in mechanical, electrical, and biofunctional engineering. However, hydrogels should overcome intrinsic issues such as poor mechanical strength, easy dehydration and freezing, weak adhesiveness and self-recovery, severely limiting their precision and reliability in practical applications. Here, silk fibroin hydrogels are developed as resistive sensors for pulse waveform monitoring. The silk fibroin hydrogel is simultaneously transparent, extremely stretchable, extra tough, adhesive, printable, and environmentally endurable. The silk fibroin hydrogel is also conductive with high sensitivity, short self-healing time, highly repeatable and reliable response, meeting the requirements for wearable sensors for continuous monitoring. The sensors with silk fibroin hydrogel present high-quality and stable waveforms of radical and brachial pulses with high precision and rich features, providing physiological signals of blood pressure and cardiac function. The sensors are promising for personalized health management, daily monitoring and timely diagnosis.

An annotated heterogeneous ultrasound database

Ultrasound is a primary diagnostic tool commonly used to evaluate internal body structures, including organs, blood vessels, the musculoskeletal system, and fetal development. Due to challenges such as operator dependence, noise, limited field of view, difficulty in imaging through bone and air, and variability across different systems, diagnosing abnormalities in ultrasound images is particularly challenging for less experienced clinicians. The development of artificial intelligence (AI) technology could assist in the diagnosis of ultrasound images. However, many databases are created using a single device type and collection site, limiting the generalizability of machine learning models. Therefore, we have collected a large, publicly accessible ultrasound challenge database that is intended to significantly enhance the performance of AI-assisted ultrasound diagnosis. This database is derived from publicly available data on the Internet and comprises a total of 1,833 distinct ultrasound data. It includes 13 different ultrasound image anomalies, and all data have been anonymized. Our data-sharing program aims to support benchmark testing of ultrasound disease diagnosis in multi-center environments.

EFNet: estimation of left ventricular ejection fraction from cardiac ultrasound videos using deep learning

The ejection fraction (EF) is a vital metric for assessing cardiovascular function through cardiac ultrasound. Manual evaluation is time-consuming and exhibits high variability among observers. Deep-learning techniques offer precise and autonomous EF predictions, yet these methods often lack explainability. Accurate heart failure prediction using cardiac ultrasound is challenging due to operator dependency and inconsistent video quality, resulting in significant interobserver variability. To address this, we developed a method integrating convolutional neural networks (CNN) and transformer models for direct EF estimation from ultrasound video scans. This article introduces a Residual Transformer Module (RTM) that extends a 3D ResNet-based network to analyze (2D + t) spatiotemporal cardiac ultrasound video scans. The proposed method, EFNet, utilizes cardiac ultrasound video images for end-to-end EF value prediction. Performance evaluation on the EchoNet-Dynamic dataset yielded a mean absolute error (MAE) of 3.7 and an R2 score of 0.82. Experimental results demonstrate that EFNet outperforms state-of-the-art techniques, providing accurate EF predictions.

Revolutionizing Cardiology: The Role of Artificial Intelligence in Echocardiography

Background: Artificial intelligence (AI) in echocardiography represents a transformative advancement in cardiology, addressing longstanding challenges in cardiac diagnostics. Echocardiography has traditionally been limited by operator-dependent variability and subjective interpretation, which impact diagnostic reliability. This study evaluates the role of AI, particularly machine learning (ML), in enhancing the accuracy and consistency of echocardiographic image analysis and its potential to complement clinical expertise. Methods: A comprehensive review of existing literature was conducted to analyze the integration of AI into echocardiography. Key AI functionalities, such as image acquisition, standard view classification, cardiac chamber segmentation, structural quantification, and functional assessment, were assessed. Comparisons with traditional imaging modalities like computed tomography (CT), nuclear imaging, and magnetic resonance imaging (MRI) were also explored. Results: AI algorithms demonstrated expert-level accuracy in diagnosing conditions such as cardiomyopathies while reducing operator variability and enhancing diagnostic consistency. The application of ML was particularly effective in automating image analysis and minimizing human error, addressing the limitations of subjective operator expertise. Conclusions: The integration of AI into echocardiography marks a pivotal shift in cardiovascular diagnostics, offering enhanced accuracy, consistency, and reliability. By addressing operator variability and improving diagnostic performance, AI has the potential to elevate patient care and herald a new era in cardiology.

Artificial Intelligence in Clinics: Enhancing Cardiology Practice

In recent years, every aspect of the society has rapidly transformed because of the emergence of artificial intelligence (AI) technologies. AI excels not only in image and voice recognition and analysis but also in achieving near-natural conversations through the development of large language models. These technological innovations are steadily being integrated into healthcare settings and can significantly change the way physicians work in clinics in the near future. Patient interviews will predominantly be performed by AI. Physicians will discuss the findings of traditional tests like electrocardiograms and chest X-rays with AI, providing beyond-human interpretation. Additionally, AI is changing areas that have seen little development for a long time, such as auscultation and phonocardiography, and the recognition and quantification of previously challenging observations like the gait analysis. Although barriers to real-world implementation exist, in the near future, a majority of physicians will collaborate with AIs supporting various aspects of clinical practice, consequently enabling more accurate and appropriate diagnosis and treatment of cardiovascular diseases, including ischemic and valvular heart diseases, arrhythmias, and heart failure. This review focuses on AI application in the field of cardiology, specifically on how it can improve the workflow in clinical settings. We examine various examples of AI integration in cardiology to demonstrate how these technologies can lead to more accurate and efficient patient care. Understanding the advancements in AI can lead to more appropriate and streamlined medical practices, which will ultimately benefit both healthcare providers and patients.

Deep Learning Applications in 12-lead Electrocardiogram and Echocardiogram

Artificial intelligence (AI), empowered by advances in deep learning technology, has demonstrated its capabilities in the medical field to automate tedious tasks that are otherwise performed by humans or to detect or predict diseases with higher accuracy compared with experts. Given the ability to take complex multidimensional data as input, AI models have primarily been applied to complex medical imaging and time-series data. Another prominent strength of AI applications is its large scalability. The field of cardiovascular medicine uses various noninvasive and accessible metrics that produce a large amount of complex multidimensional data, such as electrocardiograms (ECGs) and echocardiograms. AI models can increase the utility of such modalities. Simple automation of conventional tasks using AI models provides significant opportunities for cost reduction and capacity expansion. The ability to improve disease detection or prediction at scale may provide novel opportunities for disease screening, enabling early intervention in asymptomatic patients. For example, AI-enabled pipelines can accurately identify cardiomyopathies and congenital heart diseases from a single ECG or echocardiogram recording. The detection of these diseases using the conventional approach usually requires complicated diagnostic strategies or expensive tests. Therefore, underdiagnosis is a huge problem. Using AI models to screen these diseases will provide opportunities for reducing missed cases. The utility of AI models in the medical field is not limited to the development of clinically useful models. Recent research has shown the promise of AI models in mechanism research by combining them with genetic and structural analyses. In this review, we provide an update on the current achievements of the innovative AI application for ECG and echocardiogram and provide insights into the future direction of AI in cardiovascular care and research settings.

AI in Echocardiography: State-of-the-art Automated Measurement Techniques and Clinical Applications

The artificial intelligence (AI) technology in automated measurements has seen remarkable advancements across various vendors, thereby offering new opportunities in echocardiography. Fully automated software particularly has the potential to elevate the analysis and the interpretation of medical images to a new level compared to previous algorithms. Tasks that traditionally required significant time, such as ventricular and atrial volume measurements and Doppler tracing, can now be performed swiftly through AI's automated phase setting and waveform tracing capabilities. The benefits of AI-driven systems include high-precision and reliable measurements, significant time savings, and enhanced workflow efficiency. By automating routine tasks, AI can reduce the burden on clinicians, allowing them to gather additional information, perform additional tests, and improve patient care. While many studies confirm the accuracy and the reproducibility of AI-driven techniques, it is crucial for clinicians to verify AI-generated measurements and ensure high-quality imaging and Doppler waveforms to fully take advantage of the benefits from these technologies. This review discusses the current state of AI-driven automated measurements in echocardiography, their impact on clinical practice, and the strategies required for the effective integration of AI into clinical workflows.