Fully automatic four-module deep learning pipeline for identifying, segmenting, and Stanford-subtyping aortic dissection (AD) from CT angiography (CTA). A 3D full-resolution nnU-Net first segments the aorta; the segmented boundary is then multi-view projected for AD identification; for AD-positive cases, a second 3D nnU-Net segments the true lumen (TL) and false lumen (FL); finally, a classifier performs Stanford subtyping from multi-view maximum-density projections of the TL/FL. On 386 CTA scans, the pipeline achieved 0.979 accuracy for AD identification, Dice of 0.968 (TL) and 0.971 (FL) for lumen segmentation, and 0.990 accuracy for Stanford subtyping.

Adversarial deep learning model that predicts treatment effects for cardiology patients from structured electronic health record (EHR) data, aiming to forecast expected clinical outcomes of specific treatment choices given a patient's clinical status. Two autoencoders separately learn representations of patient characteristics and of the treatments given; an adversarial loss then encourages these representations to capture the correlational structure between a patient's status and the treatment received, improving downstream outcome prediction over non-adversarial baselines. Evaluated on two private cardiology EHR cohorts from a Chinese hospital, ADTEP modestly outperformed a non-adversarial ablation (DTEP) and classical baselines (logistic regression, SVM) at predicting major adverse cardiac events (MACE) after acute coronary syndrome (AUC 0.662 vs. 0.653/0.648/0.621) and at heart-failure outcome prediction.

Complete motion-analysis workflow for the left atrium (LA) using 3D Cine MRI, combining an online-learning segmentation network with an image-registration network to compute LA displacement vector fields (DVF) and principal strains across the cardiac cycle. Validated on 10 healthy volunteers and 8 cardiovascular disease patients, Aladdin accurately tracks LA wall motion and can identify regional deformation abnormalities that may indicate focal pathology, agreeing well with 2D Cine MRI global function estimates.

Open-source, fully automated framework for end-to-end aortic analysis from CT angiography (CTA), built on top of TotalSegmentator baseline segmentations with additional refinement. AortaExplorer extracts established biomarkers such as diameters across anatomical segments defined by the European Society of Cardiology, and introduces new metrics including aortic tortuosity. Diameter measurements were validated against expert manual readings in more than 10,000 CTA scans from Danish population cohorts, and the tortuosity index's increase with age is consistent with prior literature; the tool reduces per-case analysis time from about 15 minutes to under 5 minutes.

Self-supervised deep learning model for coronary artery segmentation from invasive X-ray coronary angiography (ICA), designed to reduce reliance on large annotated datasets. CM-UNet combines a Contrastive Masked Autoencoder (CMAE) with a UNet backbone: an online encoder-decoder branch reconstructs masked image patches while a momentum branch produces contrastive embeddings, jointly pretraining the network on unannotated angiography images before fine-tuning on a small labeled set. Fine-tuning with only 18 annotated images (instead of 500) led to just a 15.2% drop in Dice score, versus a 46.5% drop for baseline models trained without this self-supervised pretraining -- demonstrating strong label efficiency for coronary segmentation.

Cross-modality cardiac image segmentation model that addresses spatial-temporal confounding -- where the anatomy and imaging-modality elements of cardiac images are intertwined across space and time. DCL performs multi-dimensional causal intervention, modeling causal relationships between images and labels as well as causality along the time and space dimensions, integrating historical optimal interventions to transfer knowledge across temporal contexts. A diffusion mechanism further keeps extracted anatomical elements causally invariant across modalities. On cross-modality cardiac images (MR, CT, and ultrasound), DCL achieved a mean Dice of 0.951, outperforming other advanced segmentation methods.

Domain-Shuffle Temporal Attention Network for coronary vessel extraction from X-ray coronary angiography (XCA), trained entirely on synthetic temporal XCA data without requiring manual vessel annotations. By leveraging synthetic data generation and a domain-shuffle temporal attention mechanism, DOSTA-Net avoids the need for costly expert-labeled real angiography sequences while still learning temporally consistent vessel segmentation across frames of an XCA sequence.

3D convolutional autoencoder that filters reverberation clutter artifacts from transthoracic echocardiography (TTE) video sequences, improving downstream measurements such as speckle-tracking strain. Built on a 3D U-Net-style encoder-decoder with an input-output skip connection to preserve fine structures and attention-gate modules to focus on cluttered regions, the network was trained on synthetic clutter simulated across six ultrasound vendors and generalized well to real in vivo artifactual sequences, substantially reducing the discrepancy between cluttered and clutter-free strain profiles while running in a fraction of a second per sequence.

Machine-learning surrogate model for estimating pulsatile hemodynamic fields (velocity, pressure) in coronary arteries from a steady-state computational fluid dynamics (CFD) prior, avoiding the high computational cost of full pulsatile CFD. The model, a neural field conditioned on hemodynamic boundary conditions, is discretisation-independent and can be parametrised with message-passing or self-attention layers by relaxing point-wise action to permutation-equivariance. Evaluated on 74 stenotic coronary arteries from coronary CT angiography (CCTA) with patient-specific pulsatile CFD as ground truth, the model produced accurate, discretisation-independent estimates of pulsatile velocity and pressure fields.

Lightweight convolutional neural network for binary ECG classification that represents ECG beats as spectrograms (via short-time Fourier transform) rather than raw signals, after denoising and frequency filtration to reduce data volume while preserving diagnostically relevant information. Using the large public PTB-XL dataset, the spectrogram-based CNN reached 99.06% accuracy, outperforming an equivalent raw-signal CNN, while also reducing memory usage and computation by avoiding complex architectures; the authors additionally studied the effect of signal up/down-sampling on classification performance.

CNN-based arrhythmia classifier that fuses the magnitude (scalogram) and phase (phasogram) of the continuous wavelet transform (CWT) of ECG heartbeats, rather than relying on magnitude information alone as most prior 2D-representation approaches do. Several fusion strategies (input-level, intermediate-layer, and output-level fusion) were compared on the public PhysioNet MIT-BIH Arrhythmia database. Despite a simple CNN architecture, the best fusion strategy achieved about 98.5% overall accuracy, 98.5% sensitivity and 95.6% specificity, competitive with more complex state-of-the-art approaches.

Supervised contrastive learning framework for robust, fine-grained echocardiographic view classification across 47 clinically meaningful view types (rather than collapsing views into a few broad categories, as most prior work does). Introduces TTE47, the first publicly available benchmark with 47 fine-grained views independently annotated by three experts, enabling rigorous quantification of inter-observer agreement. A tailored contrastive loss produces a feature space that aligns more strongly with underlying anatomy than with any single annotator's labeling style, and the model outperforms cross-entropy and standard supervised-contrastive baselines on both TTE47 and the public TMED-2 benchmark (the latter without dataset-specific pretraining).

Self-supervised contrastive learning technique for classifying carotid plaques from ultrasound images under label scarcity. In a pretext task, a triplet network takes three augmented views (strong- and weak-augmentation) of each image and promotes their similarity from both feature- and instance-level perspectives to learn effective plaque representations; the resulting encoder is then fine-tuned on labeled ultrasound images for the downstream classification task. FeaCL achieved 83.4% classification accuracy using only 30% of the training data -- a 16.3% improvement over the same network trained without the self-supervised pretext task.

Self-supervised ECG embedding model inspired by BERT/RoBERTa from natural language processing, designed for efficient medical signal analysis. HeartBERT translates ECG signals into an intermediate synthetic 'language' via signal quantization and discretization (Lloyd-Max quantization), then trains a RoBERTa-style encoder from scratch on this text-like representation using the MIT-BIH Arrhythmia Database, PTB-XL, and European ST-T Database. The resulting embeddings are evaluated on two downstream tasks -- sleep-stage classification and heartbeat classification -- using bidirectional LSTM heads, showing particular strength when only small labeled training datasets are available.

Interpretable coronary heart disease (CHD) risk prediction model based on the LightGBM gradient-boosting algorithm, combined with SHAP (SHapley Additive exPlanations) values to make individual risk predictions explainable to clinicians. Trained on the public BRFSS_2015 survey dataset and externally validated on the Framingham and Z-Alizadeh Sani datasets, the model reached 90.60% accuracy and 81.06% AUROC on BRFSS_2015, with SHAP analysis identifying age, smoking status, diabetes, hypertension, and high cholesterol as the most influential risk features. A companion CHD scoring system was built from the model to give clinicians a user-friendly risk-assessment tool.

Self-supervised multi-encoder autoencoder (MEAE) that separates heartbeat-related source signals from noisy photoplethysmogram (PPG) via blind source separation, improving heart-rate detection without requiring any pre-processing or manual data selection. Trained entirely on PPG signals from a large open polysomnography database (with no cleaning or curation), the model is then applied to a noisy real-world PPG dataset collected during daily activities of 9 subjects and a surgical dataset of 4,681 patients; the extracted heartbeat-related source signal significantly improves heart-rate detection accuracy compared with using the raw PPG signal directly.

Deep learning strategy for cost-effective, comprehensive cardiac screening from ECG alone, by transferring domain-specific structural information from cardiac magnetic resonance (CMR) imaging into ECG representations. Combines multimodal contrastive learning with masked data modelling during pretraining on paired ECG-CMR data, then uses only ECG at inference. On 40,044 UK Biobank subjects, the multimodal pretraining improved subject-specific CVD risk prediction by up to 12.19% and cardiac phenotype prediction by up to 27.59% versus ECG-only baselines, with learned ECG representations shown to incorporate information from CMR regions of interest.

Fully automatic, open-source deep learning pipeline for estimating left atrial fibrosis from late gadolinium enhancement (LGE) cardiac MRI, built to remove the operator-dependent steps that limit reproducibility of conventional atrial LGE analysis. A multilabel convolutional neural network delineates the left atrial blood pool, pulmonary veins, and mitral valve; these structures are then used to automatically calculate fibrosis burden via established image-intensity-ratio thresholds, without manual tracing. Validated on a 3D LGE-CMR dataset of 207 scans, the pipeline's automatic segmentation achieved a 91% Dice score against manual tracing, and its fully automatic fibrosis quantification closely matched semi-automatic reference methods. The CNN and pipeline are distributed as part of the open-source CemrgApp platform.

Self-supervised deep learning model that extracts cardiovascular-risk-relevant patterns from multimodal polysomnography (PSG) signals -- EEG, ECG, and respiratory signals -- without relying on manual sleep-stage annotations. Trained on 4,398 participants, the model derives 'projection scores' by contrasting embeddings from individuals with and without cardiovascular disease (CVD) outcomes. Externally validated in an independent cohort of 1,093 participants, ECG-derived projection scores were predictive of prevalent and incident cardiac conditions (particularly CVD mortality), and combining projection scores with the Framingham Risk Score consistently improved prediction (AUC 0.607-0.965 internally, 0.710-0.807 externally across most outcomes).

Optimized Hierarchical Fused Fuzzy Deep Reinforcement Learning model for predicting non-sinus (higher-risk) cardiac rhythms from PQRST-analyzed 12-lead ECG data. The three-stage approach combines data preprocessing, reinforcement learning, and fuzzy deep learning to classify sinus vs. non-sinus rhythms. Evaluated on a 12-lead ECG dataset of 10,646 patients, OHFFDRL achieved 94% accuracy, an AUC of 0.91, and was interpreted using SHAP, LIME, calibration curves, adversarial vulnerability analysis, and integrated gradients; TAxis (ventricular repolarization movement range) was found to be the most important distinguishing feature.

Prototype-based neural network for interpretable, uncertainty-aware classification of aortic stenosis (AS) severity from B-mode echocardiography videos. Rather than a black-box prediction, ProtoASNet bases its output on similarity scores between the input video and a set of learned spatio-temporal prototypes (typically highlighting valve calcification and restricted leaflet motion), and uses an abstention loss to flag ambiguous/uncertain cases for expert review. Evaluated on a private clinical dataset and the public TMED-2 dataset, it achieved balanced accuracy of 80.0% (private) and 79.7% (TMED-2), improving to 82.4% when uncertain cases are excluded.

Regression-based convolutional neural network that directly estimates the heart-rate-corrected QT interval (QTc) from a 12-lead ECG, aiming to improve on automated QTc measurements from commercial ECG systems, which often diverge from expert readings. QTcNet was trained on 120,300 algorithm-labeled ECGs from an internal hospital cohort and the public MIMIC-IV database (after correcting for a vendor-specific measurement bias), and evaluated against expert QTc measurements in three independent external cohorts (PTB Diagnostic ECG Database, QTcMS, and ECGRDVQ). It roughly halved the mean absolute error compared with standard ECG analysis software (from 23.4ms to 13.4ms across external validation cohorts), with explainability analysis confirming the model focuses on physiologically plausible QRS-onset and T-offset regions.

Reinforcement-learning-based unsupervised domain adaptation framework for spatio-temporal (2D+time) echocardiography segmentation, extending the authors' earlier RL4Seg work to full-length video sequences. RL4Seg3D uses a sliding-window approach supporting high-resolution, full-sized inputs, and fuses multiple reward mechanisms to improve segmentation reliability without requiring additional expert annotations in the target domain. Trained and evaluated on a large dataset of over 30,000 echocardiography videos, it outperforms baselines and foundation models on overall segmentation accuracy as well as echocardiography-specific metrics including anatomical/temporal validity and mitral-valve-commissure landmark precision, and supports test-time optimization via calibrated uncertainty estimates.

Disentangled representation learning model for cardiac image analysis that factorises 2D medical images (MRI, CT) into a spatial 'anatomy factor' (a semantically meaningful multi-channel map, produced by a U-Net-style anatomy encoder) and a non-spatial 'modality factor' (a latent vector capturing imaging-specific characteristics). This disentangled representation supports semi-supervised segmentation using only a fraction of labeled images (matching fully supervised performance), multi-task learning (e.g. jointly regressing cardiac indices), multimodal pooling of MRI and CT data, and image-to-image synthesis between modalities via latent-space arithmetic (swapping modality factors). SDNet also demonstrates that its modality factor alone can predict the input imaging modality with high accuracy.

Self-supervised learning (SSL) model that identifies chronically stressed mother-fetus dyads from raw maternal abdominal ECG (aECG), which contains both maternal and fetal cardiac signals. Built on a self-supervised representation-learning approach originally developed for ECG-based emotion recognition, the model is pretrained on public ECG datasets and evaluated on a cohort of pregnant women with chronic stress exposure validated by psychological inventory, maternal hair cortisol, and the fetal stress index (FSI). Using maternal ECG alone with the publicly pretrained model, it detected the chronic-stress-exposure group with AUROC 0.982 and predicted psychological stress score (R2 0.943), FSI (R2 0.946), and maternal hair cortisol (R2 0.931).

Transformer-based multi-view multiple-instance learning (MIL) framework for patient-level coronary stenosis classification from multi-view invasive coronary angiography. Rather than requiring expensive view-level stenosis annotations, SegmentMIL is trained end-to-end on real-world clinical data using only patient-level labels already present in hospital systems, and jointly predicts stenosis presence while localizing the affected artery (left/right) and segment. It captures temporal dynamics and dependencies across the multiple angiographic views per patient (which prior view-level models ignore), and outperforms both single-view models and classical MIL baselines on internal and external clinical evaluations.

Real-time deep-learning model that fuses serial 12-lead ECG waveforms with sequential vital signs and routinely available clinical data to predict hospital admission early during emergency department (ED) encounters with cardiac presentations (chest pain, dyspnea, syncope, presyncope). Developed and validated on the public MIMIC-IV, MIMIC-IV-ED, and MIMIC-IV-ECG databases (n=30,421 ED stays with >=1 ECG; n=11,273 with >=2 ECGs), the model improves on baseline tabular (random forest) and ECG-only models by leveraging how a patient's risk evolves with successive ECGs during the visit, addressing a key limitation of single-time-point risk scores.

Self-supervised representation-learning method for 12-lead ECG signals, designed to reduce reliance on large labeled datasets for downstream ECG classification. TSSL exploits two structural properties of ECG data: temporally, it encourages stable representations for the same individual across time while keeping different leads distinguishable; spatially, it enforces consistency in the relationships between signals and their representations across the different leads of a single recording. Evaluated on three public ECG datasets (CPSC2018, Chapman, PTB-XL), TSSL-pretrained models approached the performance of fully supervised training while using only about 10% of the labeled data.

Automatic coronary artery segmentation pipeline for coronary CT angiography (CCTA). A 2D DenseNet classifier first screens out CT slices that don't contain coronary artery, then a 3D-UNet -- enhanced with dense blocks in the encoder for richer feature extraction and residual, feature-rectifying blocks in the decoder -- segments the coronary artery tree in the remaining slices. A Gaussian-weighted merging scheme combines overlapping 3D patch predictions, up-weighting the more reliable predictions near each patch's center. On the authors' in-house CCTA dataset, the method achieved a Dice similarity coefficient of 0.826.

Open-source, user-guided deep learning tool for coronary artery segmentation from invasive coronary angiography (ICA), designed to improve on traditional quantitative coronary angiography (QCA) edge-detection algorithms that typically require manual correction. Rather than segmenting the whole coronary tree indiscriminately, AngioPy lets the user click a handful of ground-truth points along a specific target vessel (including side branches), and predicts a binary mask for that single artery at the chosen cardiac-cycle time-step. Evaluated against an established QCA system on angiograms from the FAME 2 trial, AngioPy achieved an average F1 score of 0.927 (internal) and 0.924 (external validation), with vessel-diameter and lesion minimal-lumen-diameter measurements showing excellent agreement with QCA (r=0.93-0.96).

Deep learning model for multi-class 3D segmentation of the aorta and its thirteen branches from CT angiography, intended to support planning of endovascular aortic interventions. CIS-UNet combines a CNN encoder with a symmetric decoder and a novel Context-aware Shifted Window Self-Attention (CSW-SA) bottleneck block that adapts the Swin transformer's patch-merging mechanism to more efficiently capture global spatial context. Trained and evaluated via 4-fold cross-validation on the first public multi-branch aorta CTA dataset (59 patients), CIS-UNet outperformed the state-of-the-art SwinUNETR baseline, achieving a mean Dice of 0.713 vs. 0.697 and mean surface distance of 2.78mm vs. 3.39mm, while being more computationally efficient.

Deep-learning model that estimates a patient's 10-year risk of major adverse cardiovascular events (MACE) directly from a single routine chest radiograph (CXR), intended as an opportunistic risk-assessment tool when the inputs needed for the standard ASCVD risk calculator (lipids, blood pressure, smoking status, etc.) are missing. A 2D convolutional network takes the CXR image alone as input and outputs a continuous 10-year MACE risk estimate. Developed on 147,801 CXRs from 40,718 participants in the PLCO cancer-screening trial and externally validated in 8,869 outpatients with unknown ASCVD risk and 2,132 with known risk at Mass General Brigham, CXR CVD-Risk identified people at elevated MACE risk (adjusted hazard ratio 1.73 for a >=7.5% predicted risk) and provided added discrimination beyond the traditional ASCVD score.

Two-stage deep learning pipeline that localizes premature ventricular contraction (PVC) beats directly from raw, unsegmented ECG signal, without relying on hand-crafted features or pre-existing R-peak annotations. An encoder-decoder network first localizes the R-peak of every heartbeat (normal or anomalous); the resulting R-peak positions are then passed to CardioIncNet, a 1D InceptionTime-based classifier, which delineates each beat as healthy or PVC. Evaluated with both single-dataset and cross-dataset protocols across three public ECG databases, the pipeline reached F1 scores above 0.99 (single-dataset) and 0.979 (cross-dataset) for R-peak localization, and above 0.96 and 0.85 respectively for PVC beat classification.

Open-source, vendor-agnostic deep learning pipeline that retrospectively measures left ventricular global longitudinal strain (GLS) from routine apical-4-chamber echocardiography B-mode video, without requiring speckle-tracking software or manual tracing. The pipeline reuses EchoNet-Dynamic's LV semantic-segmentation network to trace the LV endocardial border frame-by-frame, then measures the frame-to-frame change in traced myocardial length across the cardiac cycle to derive GLS. In external validation against a large 3D-echocardiography-derived GLS dataset and a prospective two-sonographer, two-vendor repeated-measures study, the automated strain measurement showed lower inter- and intra-measurement variability than human readers and moderate agreement with reference speckle-tracking strain (ICC 0.58), while being robust to image-quality differences and vendor.

Pediatric-specific extension of EchoNet-Dynamic: a video-based deep learning model that segments the left ventricle and estimates ejection fraction (EF) from apical-4-chamber (A4C) and parasternal short-axis (PSAX) pediatric echocardiogram clips. Because adult-trained echo models generalize poorly to children (who vary widely in heart size, rate, and image quality), EchoNet-Peds was trained from scratch on a dedicated pediatric video dataset. It segments the LV with a Dice similarity coefficient of 0.89 in both views, estimates EF with a mean absolute error of 3.66%, and identifies pediatric systolic dysfunction with an AUC of 0.95, significantly outperforming an adult-trained model applied to the same pediatric data.

Automated deep learning workflow that detects and grades tricuspid regurgitation (TR) severity from full transthoracic echocardiography studies. The pipeline first identifies apical-4-chamber (A4C) video clips with color Doppler across the tricuspid valve from a full echo study, then applies a dedicated R(2+1)D video classifier to grade TR severity. Trained on over 2 million echo videos from 47,312 studies at Cedars-Sinai Medical Center and externally validated at Stanford Healthcare, the model identified color-Doppler A4C views with AUC ≥0.999 and detected clinically significant (moderate-or-severe) TR with AUC 0.951 and severe TR with AUC 0.980. Code and trained model weights are released to support prospective evaluation of AI-assisted TR screening.

Self-supervised deep learning system that detects ECG anomalies indirectly, by learning to forecast what a normal ECG signal should look like next. FADE is trained only on normal ECG segments using a novel morphology-inspired loss function; at inference time, a large mismatch between the forecast and the observed signal flags an anomaly, avoiding the need for labeled abnormal-beat datasets. Evaluated on the public MIT-BIH NSR and MIT-BIH Arrhythmia databases, FADE reached an average accuracy of 83.84% for anomaly detection and 85.46% for correctly classifying normal ECG, and the approach can be adapted to new recording contexts via domain adaptation.

Deep learning model that non-invasively estimates cardiac output (CO) from wearable seismocardiography (SCG), a single-lead ECG, and body mass index (BMI), as a potential alternative to invasive right heart catheterization (RHC). Parallel 1D-CNN branches extract features from the SCG and ECG waveforms, which are fused with BMI and passed through a lightweight regression head to predict CO directly. Trained and evaluated via leave-pair-out cross-validation on 73 heart-failure patients (83 RHC encounters) from an open PhysioNet dataset, the model achieved an RMSE of 1.00 L/min (22%) and Pearson correlation of 0.75 versus catheterization-derived CO, with particularly strong performance in low-output states.

Deep channel-attention network for segmenting the full coronary vessel tree from sequential X-ray coronary angiography (XCA) frames, rather than a single static image. An encoder-decoder architecture fuses temporal-spatial feature maps across the XCA sequence via skip connections, then uses channel-attention blocks in the decoder to refine features and separate thin vessel structures from complex, noisy backgrounds; a Dice loss addresses the severe foreground/background class imbalance typical of XCA. The authors report that SVS-net outperforms prior 2D and video-based baselines on both quantitative vessel-segmentation metrics and visual validation.

Open, transparent deep-learning method for measuring left ventricular global longitudinal strain (GLS) from routine 2D echocardiography, built as an alternative to proprietary vendor strain software. Unity-GLS is a multi-image neural network (based on the HigherHRNet-W32 pose-estimation architecture) that identifies the mitral annulus, LV apex, and endocardial curve from a target frame plus six neighbouring frames, across apical 4-, 3-, and 2-chamber views. Validated against multi-expert (11-reader) consensus tracings from 100 echocardiograms in a UK-wide collaborative, Unity-GLS agreed with expert consensus as strongly as individual human experts and two proprietary vendor packages (correlation with consensus: 0.91 vs. 0.73-0.85 for other methods).

Explainable machine learning model that detects and localizes left ventricular (LV) scar in hypertrophic cardiomyopathy (HCM) patients directly from 12-lead ECG, as a faster and cheaper alternative to late-gadolinium-enhancement (LGE) cardiac MRI, the clinical gold standard. XplainScar first uses an HCM-specific ECG segmentation algorithm to extract morphological features (duration, amplitude, slope, energy) from the QRS complex, ST segment and T wave of each lead, then combines unsupervised and self-supervised representation learning to predict scar presence and reveal which ECG features are associated with scar location (basal, mid, or apical LV). Trained on 500 HCM patients from the Johns Hopkins HCM Registry and validated on a held-out cohort of 248 HCM patients from UCSF, it reached 88% precision, 90% sensitivity, 78% specificity and an F1-score of 89% for scar detection on the external test set, analyzing a 10-patient batch of ECGs in under one minute.

Multi-label 12-lead ECG diagnosis model submitted to the PhysioNet/Computing in Cardiology Challenge 2020, built on the same residual 1D CNN family as the authors' earlier Nature Communications model but retrained and validated across the challenge's large, multi-institutional pooled training set (CPSC2018, China 12-Lead ECG Database, St. Petersburg INCART, PTB and PTB-XL, and the Georgia 12-Lead ECG Database). The model uses an unsupervised pretraining stage -- predicting unseen samples of a partially masked ECG signal -- before supervised fine-tuning to jointly detect nine diagnostic classes (atrial fibrillation, first-degree AV block, left and right bundle branch block, normal rhythm, premature atrial/ventricular contraction, and ST-segment depression/elevation). The 2020 Challenge was notable for requiring every team to publicly release both their trained model weights and full training code, making this one of relatively few 12-lead ECG classifiers with an end-to-end reproducible public pipeline.

One of the foundational baseline segmentation networks submitted to the 2017 Automated Cardiac Diagnosis Challenge (ACDC), comparing 2D and 3D convolutional network designs for segmenting the left ventricle cavity, myocardium, and right ventricle cavity from short-axis cine cardiac MRI at end-diastole and end-systole. The accompanying study systematically explored the tradeoffs between 2D and 3D convolutions for this task, finding that, due to the highly anisotropic voxel spacing typical of clinical cine cardiac MRI, 2D networks that treat each slice independently can match or exceed 3D networks while being far cheaper to train. The public code and pretrained weights for the best-performing configuration have served as a widely used, simple baseline for later cardiac MRI segmentation research (including for automatically deriving ventricular volumes and ejection fraction).

nnU-Net-based pipeline for automated detection and sub-classification of acute thoracic aortic dissection (AD) on heterogeneous CT imaging, formulated as a semantic segmentation task rather than direct image classification. The model segments the false lumen (ascending and descending) and the dissection membrane -- along with optional indirect signs such as hemopericardium, aortic wall hematoma, and supra-aortic branch dissection -- and a patient is classified as AD-positive if at least two of the three primary segmented regions exceed a volume threshold tuned via Youden's index; the same pipeline additionally flags Stanford type A dissections. Trained on 157 heterogeneous internal CT studies (not restricted to a single contrast protocol) from Mannheim University Medical Centre and evaluated on an internal held-out test set as well as public external datasets (ImageTBAD and AVT), the model reached an AUROC of 98.7% internally and 97.0% externally, and correctly flagged 93.3% of dissection cases that had not been clinically suspected before imaging. The authors state the trained network will be made publicly available as a non-medical device for further scientific research.

Coronary artery calcium (CAC) scoring model that transfers a CNN trained for calcium scoring on non-contrast CT (NCCT) to coronary CT angiography (CCTA), where iodinated contrast otherwise confounds calcium detection and large annotated CCTA training sets are scarce. The CAC-scoring CNN is split into a feature generator and a classifier; the feature generator is trained on the NCCT source domain and adapted to the CCTA target domain via adversarial learning combined with a maximum-mean-discrepancy loss, while the source-domain classifier is reused unchanged for the target domain. Builds directly on the authors' earlier non-contrast CT calcium-scoring network.

The original U-Net baseline segmentation network introduced alongside the CAMUS (Cardiac Acquisitions for Multi-structure Ultrasound Segmentation) dataset, one of the largest fully open-access, expert-annotated 2D echocardiography benchmarks. The network segments the left ventricle endocardium (LVEndo), left ventricle epicardium/myocardium (LVEpi), and left atrium (LA) from apical 2-chamber and 4-chamber echo views at end-diastole (ED) and end-systole (ES). In the original ten-fold cross-validation benchmark comparing U-Net, U-Net++, Stacked Hourglass, Anatomically Constrained Neural Networks, and classical methods, the U-Net variant (18M parameters) achieved the best overall accuracy, reaching Dice scores of 0.939 (ED) / 0.916 (ES) for LVEndo and 0.954 (ED) / 0.945 (ES) for LVEpi, approaching inter-observer variability. A pretrained checkpoint of this baseline U-Net is distributed via the University of Sherbrooke's vitalab CASTOR project as part of a broader library for building anatomically-constrained cardiac segmentation pipelines.

End-to-end deep learning framework that predicts the procedural outcome of percutaneous coronary intervention (PCI) for chronic total occlusion (CTO) lesions directly from preprocedural coronary CT angiography, aiming to replace slower manual scoring systems (J-CTO, CT-RECTOR, KCCT). The pipeline first segments the coronary artery tree (Patch-UCTNet), detects candidate CTO lesions along the delineated vessel, extracts pathological lesion features with a Swin Transformer, and classifies two outcomes: successful guidewire crossing within 30 minutes and overall PCI success. In the original study, the model completed reconstruction and analysis 85% faster than manual scores (73.7s vs. 418-467s) and was more accurate than the manual CT-RECTOR, KCCT, and J-CTO_CCTA_ scores, reaching an AUROC of 0.97 on the internal test set and 0.96 on an independent external validation cohort (186 patients, 189 CTO lesions).

Two-stage convolutional neural network that automatically detects and anatomically labels coronary artery, thoracic aorta, and cardiac-valve calcifications in low-dose chest CT acquired for lung-cancer screening. A first CNN with a large receptive field (via dilated convolutions) identifies and anatomically labels candidate calcifications; a second CNN filters true positives from the candidates. Trained and evaluated on 1,744 CT scans from the National Lung Screening Trial (NLST), reaching an F1 of 0.89 (soft-filter reconstructions) / 0.84 (sharp-filter reconstructions) for coronary artery calcifications and a linearly-weighted kappa of 0.90-0.91 for per-subject cardiovascular risk categorization versus the manual reference standard.

Dual Bayesian ResNet model for heart murmur detection from multi-location phonocardiogram (PCG) recordings, developed for the George B. Moody PhysioNet Challenge 2022. Each patient's PCG recordings are segmented into overlapping log-mel spectrograms, which are passed through two Bayesian ResNet binary classifiers running simultaneously (present vs. unknown-or-absent, and unknown vs. present-or-absent); the two outputs are aggregated into a patient-level present/unknown/absent murmur classification. An optional second-stage XGBoost model integrates the DBRes output with demographic data and hand-crafted signal features. On the Challenge's official hidden test set, DBRes achieved a weighted accuracy of 0.771 for the murmur-detection task, placing 4th among all teams.

Deep learning pipeline for detection and quantification of abdominal aortic aneurysm (AAA) -- a typically asymptomatic condition often missed incidentally by radiologists -- from abdominal-pelvic CT. A modified 3D U-Net segments the aorta on both contrast and non-contrast CT volumes with a variable number of images, after which an ellipse-fitting post-processing step measures the aortic cross-sectional diameter along the vessel to detect aneurysmal dilation. Trained and validated on 321 abdominal-pelvic CT examinations from Massachusetts General Hospital, the model reached a sensitivity/specificity of 0.91/0.95 on the primary validation set, and 0.85/1.0 on a separate 57-exam generalization test set with different patient demographics and acquisition characteristics; the authors report that DeepAAA exceeded literature-reported radiologist performance for incidental AAA detection.

Early semi-supervised sequence model for estimating cardiovascular and metabolic risk directly from consumer wearable sensor data (heart rate, step count, and activity level), rather than from clinical-grade ECG or imaging. A multi-task long short-term memory (LSTM) network is first pretrained using semi-supervised sequence learning or heuristic pretraining on unlabeled wearable time series, then fine-tuned to jointly predict four self-reported conditions: diabetes, high cholesterol, high blood pressure, and sleep apnea. Trained and validated on 57,675 person-weeks of data from participants in UCSF's Health eHeart study using the Cardiogram app on Fitbit, Apple Watch, or Android Wear devices, DeepHeart outperformed hand-engineered heart-rate-variability biomarkers from the medical literature, reaching AUROCs of 0.845 (diabetes), 0.744 (high cholesterol), 0.809 (high blood pressure), and 0.830 (sleep apnea). The paper was an early demonstration that population-scale, passively-collected wearable heart-rate data could support cardiometabolic risk screening without any dedicated clinical measurement.

Deep learning platform for fully automatic segmentation and phenotyping of coronary intravascular ultrasound (IVUS) pullbacks, packaged with a desktop GUI and CLI. A convolutional encoder-decoder network delineates the internal (lumen) and external elastic lamina borders on each cross-sectional IVUS frame; downstream rule-based analysis derives lumen area, plaque area, plaque burden, automatically flags lesions with plaque burden exceeding 40%, and reports minimum lumen area and maximum plaque burden along the pullback. Also supports end-diastolic gating and manual contour editing. Trained on 305 clinical IVUS pullbacks (270 train / 35 validation) from Philips and Boston Scientific catheters at Emory University; downstream evaluations have applied DeepIVUS to tasks such as automated detection of stent underexpansion.

Self-supervised model that performs single-frame digital-subtraction-angiography-style vessel/background separation directly from a single live (non-subtracted) coronary angiogram frame, then supports fine-tuned coronary vessel segmentation. A U-Net-style network is pretrained via an image-to-image translation objective on 58,128 unannotated angiography DICOM series (3,756 patients), then fine-tuned for vessel segmentation on just 40 expert-annotated frames, reaching a Dice of 0.828 on the held-out fine-tuning set and a new state-of-the-art Dice of 0.755 on the public XCAD benchmark. Intended to help clinicians visualize potential stenosis sites without requiring true two-frame digital subtraction acquisition.

Clinically-informed modification of the ResNet-18 architecture for identifying occlusion myocardial infarction (OMI) -- a severe, often ST-elevation-negative heart attack caused by complete blockage of a coronary artery -- from a single 12-lead ECG. The network first learns lead-specific temporal features via 1xk temporal convolutions, then learns cross-lead spatial concordance/discordance (e.g. reciprocal ST changes) via a 12x1 spatial convolution placed after the residual blocks, with saliency maps highlighting the most relevant leads and waveform regions for explainability. Benchmarked against ResNet-18 and other CNN/random-forest baselines on a multisite real-world clinical dataset of 10,893 ECGs (OMI rate 6.5%), reaching a test AUROC of 0.889 and an average precision of 0.587, outperforming the compared models.

Ensemble classifier combining hand-engineered expert features with a deep convolutional neural network for classifying cardiac rhythm from a single-lead ECG recording into normal sinus rhythm, atrial fibrillation, another rhythm, or too noisy to classify. A large set of expert features (from time-, frequency-, and template-based analysis) is fed into a gradient-boosted tree classifier (AdaBoost), and its output is combined with a separate deep CNN operating directly on the raw waveform; combining both feature families measurably outperformed either alone. ENCASE won 1st place in the PhysioNet/Computing in Cardiology Challenge 2017 (single-lead AF classification) with an overall F1 score of 0.83 on the official hidden test set, and remains a widely cited example of combining classical signal-processing features with deep representations for ECG classification.

End-to-end deep learning model for detecting fetal QRS complexes directly from non-invasive abdominal ECG (aECG) signals, without requiring a separate reference maternal ECG or heavy hand-crafted feature engineering. The model adopts a ResNet architecture built from 1-D octave convolutions (OctConv), which factorize feature maps into high- and low-frequency components to capture multiple temporal frequency scales while reducing memory and compute cost relative to a standard 1-D ResNet; the resulting feature-importance weighting also highlights the signal regions most relevant to each detection. Evaluated on the PhysioNet/CinC Challenge 2013 fetal ECG database (with added Gaussian and motion-artifact noise to mimic real-world recording conditions), the model reached an F1 score of 91.1% while cutting computation by more than 50% for less than a 2% drop in performance versus a non-octave ResNet baseline.

Video-based deep learning system for automated screening of rheumatic heart disease (RHD) from echocardiography, aimed at low-resource settings where RHD -- the most common acquired heart disease in children and young adults worldwide -- is endemic but echocardiography expertise is scarce. A 3D convolutional neural network (C3D) classifies each echo video clip, explicitly modeling temporal information across frames; a supervised meta-classifier then aggregates the per-video predictions from an exam (which may contain many video clips from different views) into a single exam-level RHD diagnosis. Evaluated on 11,646 echocardiography videos from 912 screening exams collected in underserved areas of Brazil and Uganda, the 3D C3D network significantly outperformed a comparison 2D CNN (VGG16) that ignores temporal information, and the learned aggregation model reached 72.77% exam-level diagnostic accuracy, exceeding simple majority voting across a patient's videos.

Self-supervised learning method for carotid plaque segmentation from B-mode ultrasound images, designed to reduce the amount of expert-labeled data needed to train a segmentation network. A level-set-and-least-squares-based deformation procedure synthesizes registration image pairs from unlabeled carotid ultrasound images, and a spatial-transformer-based registration pretext task pretrains a Stacked U-Net (Su-Net) to focus on plaque-contour features before fine-tuning on a small labeled dataset for total plaque area (TPA) segmentation. Evaluated on carotid ultrasound datasets from two different institutions and countries, the method showed robust generalization when trained with only a small number of labeled images.

Heart murmur detection model combining spectrogram-derived deep features with hand-crafted peak-interval (PI) features extracted from phonocardiogram recordings, submitted to the George B. Moody PhysioNet Challenge 2022 (team CAU_UMN) and later extended into the 'SpectroHeart' method. Peak-to-peak interval sequences and their summary statistics are combined with spectrogram representations of the PCG signal, optionally alongside patient demographic data, to classify murmur presence across multiple auscultation locations. The team's Challenge submission placed 5th of all teams on the murmur-detection task.

nnU-Net-based segmentation network that detects and delineates stenotic lesions directly from X-ray coronary angiography frames, developed for the ARCADE (MICCAI 2023) stenosis-detection challenge. A companion model (YOLO-Angio, same team) handles vessel-tree segmentation; StenUNet focuses specifically on pixel-wise localization of stenotic regions. Placed 3rd overall among ARCADE challenge entrants with an F1 score of 0.5348 on the hold-out test set, within 0.0005 of the 2nd-place team.

Dilated U-Net model for fully automatic segmentation of the intima-media complex (IMC) of the common carotid artery on longitudinal B-mode ultrasound images, used to measure carotid intima-media thickness (cIMT) -- a standard imaging biomarker of subclinical atherosclerosis. A far-wall detection step first localizes the region of interest, and the dilated U-Net then segments the near- and far-wall IMC boundaries within it. Trained and evaluated with 5-fold cross-validation on a multicenter database of 2,176 images annotated by two experts, the method reached a mean absolute thickness difference of under 120 micrometres versus the reference annotations -- smaller than the approximately 180-micrometre inter-observer variability -- with a 98.7% fully-automatic success rate (only 1.3% of cases required manual correction).

U-Net-variant segmentation model that identifies and quantifies coronary artery calcium (CAC) directly from routine non-gated, non-contrast chest CT scans -- the kind ordered for lung-cancer screening or unrelated indications rather than a dedicated cardiac scan -- so that the tens of millions of such scans performed annually can be opportunistically screened for cardiovascular risk without any extra imaging. Predicted calcium masks are combined with the CT's Hounsfield units to compute an Agatston-equivalent score. Trained on 446 expert-segmented scans from 98 medical centers across the U.S. Department of Veterans Affairs national health system (capturing substantial heterogeneity in scanners and protocols) and benchmarked against 795 patients with a paired same-year gated CAC study: nongated AI-CAC differentiates zero-vs-nonzero and <100-vs->=100 Agatston categories with 89.4% (F1 0.93) and 87.3% (F1 0.89) accuracy respectively, and its score stratifies 10-year all-cause mortality (CAC 0 vs. >400: 25.4% vs. 60.2%, hazard ratio 3.49) and composite stroke/MI/death risk (33.5% vs. 63.8%, hazard ratio 3.00). In a simulated opportunistic-screening run across 8,052 low-dose CT scans, cardiologists confirmed 99.2% of patients flagged with AI-CAC >400 would benefit from lipid-lowering therapy. Code and trained model weights are both public under an MIT license.

Unified deep-learning model for 12-lead ECG analysis that predicts a broad range of cardiac and non-cardiac discharge diagnoses coded under the ICD-10 classification system, evaluated as a unified screening tool for emergency departments where a single ECG could flag many potential conditions at once rather than one disease at a time. Introduces the MIMIC-IV-ECG-ICD-ED benchmark dataset (derived from MIMIC-IV and MIMIC-IV-ECG) and reports AUROC scores across diverse diagnostic scenarios (all discharge diagnoses vs. emergency-department-only diagnoses, cardiac vs. non-cardiac ICD-10 chapters), suggesting integration into emergency-department clinical decision-support systems.

Multi-outcome AI-ECG risk-estimation platform that turns a single 12-lead ECG into a patient-specific survival curve, using a residual convolutional neural network trained with a discrete-time survival loss to predict not just risk but time-to-event. Beyond all-cause and cardiovascular mortality, separately fine-tuned heads predict future ventricular arrhythmia, complete heart block, atrial fibrillation, atherosclerotic cardiovascular disease, heart failure, and (in follow-up work) hypertension -- all from a single resting ECG, including in ECGs a cardiologist would read as normal. Derived on 1.16 million ECGs from 189,539 Beth Israel Deaconess Medical Center patients and externally validated across transnational cohorts in the USA, Brazil (CODE), and the UK (UK Biobank). A variational autoencoder and genome/phenome-wide association analyses were used to show the model's predictions track biologically plausible features (QRS morphology, LV structure/function, and loci linked to cardiac structure, QT interval, and biological aging). NHS trials of AIRE were planned for late 2025. No public code or model weights have been released.

Multimodal machine learning model that flags patients at risk of transthyretin amyloid cardiomyopathy (ATTR-CM) -- a progressive, underdiagnosed disease with expanding disease-modifying treatment options -- from routinely available ECG waveforms, echocardiographic measurements, demographics, and diagnosis codes for orthopedic manifestations of amyloidosis (e.g. carpal tunnel syndrome, spinal stenosis). Developed on 799 patients with 5-fold cross-validation (AUROC 0.85) and externally validated on 422 patients at a separate site (AUROC 0.82), with consistent accuracy across Hispanic, non-Hispanic Black, and non-Hispanic White patients. In a subsequent nonrandomized, single-system, multisite clinical trial (the Cardiac Amyloidosis Discovery Trial), patients flagged by ATTRACTnet and referred for confirmatory amyloid scintigraphy were positive for ATTR-CM 48% of the time, more than 2.8x the positivity rate of historical (15.3%) and contemporary (17.0%) controls referred by usual clinical judgment (P < .001 for both). This is a proprietary clinical AI program; no public code or model weights have been released.

ECG-guided photoplethysmography (PPG) foundation model pretrained on over 100,000 hours of synchronized PPG-ECG recordings from 58,796 subjects across five clinical and wearable sources, using a CLIP-style contrastive alignment framework so the PPG encoder inherits physiologically grounded structure from paired ECG. Achieves state-of-the-art performance on 13 of 15 conventional physiological-analysis tasks across eight datasets, and shows meaningful discriminative capability (AUC >= 0.70) for 307 ICD-10-coded phenotypes across 16 phecode chapters, including many non-cardiovascular conditions.

LoRA-adapted domain-specialized cardiology text embedding model built on BioLinkBERT (340M parameters), identified as the top performer among 10 encoder- and decoder-style transformer architectures benchmarked head-to-head for cardiology semantic retrieval. LoRA fine-tuning on ~150,000 cardiology-textbook-derived sentence pairs raised its cardiology semantic-separation score from 0.033 (zero-shot) to 0.510, the highest of any evaluated architecture (including decoder models up to 10x larger), while remaining Pareto-optimal for the separation/throughput trade-off at 143.5 embeddings/sec and a 1.51GB memory footprint.

Foundation vision-language model for cardiac MRI that learns pathophysiological visual representations directly from the natural-language radiology reports accompanying each scan, rather than from hand-labeled targets. A Multi-scale Vision Transformer (MViT, Kinetics-400-initialized) video encoder for cine CMR sequences is contrastively pretrained (InfoNCE) against a PubMed-pretrained BERT text encoder over 19,041 multi-institutional CMR studies. The frozen vision encoder transfers with strong performance to left-ventricular ejection-fraction regression (MAE 3.34% on a UK Biobank hold-out of ~4,259-45,623 participants) and detecting HFrEF (LVEF<40%, AUC 0.880), and the paper reports emergent zero-/few-shot performance across 39 cardiac and non-cardiac conditions including cardiac amyloidosis and hypertrophic cardiomyopathy. Code and pretrained MViT encoder weights are both released (Hugging Face, CC BY-NC 4.0).

Multimodal cardiac-sensing foundation model pretrained with generative masked pretraining on ECG, PPG, and paired clinical/machine-generated text reports from roughly 1.7 million individuals across three large-scale critical-care and outpatient ECG datasets. A channel-embedding scheme lets the same model accept any combination of 12-lead ECG, single-lead/wearable ECG, and PPG. The resulting embeddings transfer to diagnostic classification, demographic recognition, vital-sign measurement, clinical-outcome prediction, and ECG question answering. Pretrained weights require a signed academic-access agreement rather than an open download.

Vision Transformer-based classifier that detects abnormal left-ventricular ejection fraction (LVEF < 50%) directly from static, non-gated, non-contrast chest CT scans -- an imaging modality ordered for unrelated indications (lung cancer screening, pulmonary embolism, trauma) in over 80 million US exams a year -- as a form of opportunistic heart-failure screening. Fine-tunes the encoder of the CT-ViT (GenerateCT) framework, with separate spatial (axial-plane) and z-axis (slice-wise) self-attention blocks, on 3D CT volumes paired with echocardiogram-derived LVEF labels from 25,948 Columbia University studies; reaches an AUROC of 0.786 on a held-out test set and 0.762 on external validation at Weill Cornell Medicine, clearly outperforming demographic/diagnosis-code-only baselines (Random Forest, XGBoost, AUROC 0.54-0.61). On a radiologist-comparison subset, the model's weighted F1 (0.80-0.81) exceeded two board-certified thoracic radiologists (0.62-0.80) at a small fraction of the interpretation time. Grad-CAM saliency maps highlighted clinically sensible correlates of reduced LVEF (cardiomegaly, dilated superior vena cava, calcified ascending aorta, pacemaker hardware, pulmonary edema). No public code or model weights have been released.

Multimodal deep-learning framework that estimates and forecasts abnormal laboratory values directly from a 12-lead ECG plus routinely available demographics, biometrics, and vital signs -- reframing dozens of blood tests as binary classification targets predictable from a test that is already fast, non-invasive, and nearly universal in acute care. A structured state-space (S4) encoder processes the raw ECG waveform and is late-fused with an MLP encoder over the tabular metadata; the same architecture is trained both to estimate the closest lab value within 60 minutes of the ECG ('abnormality prediction') and to forecast whether a value will become abnormal 30/60/120 minutes into the future ('abnormality forecasting'). Trained and evaluated on 385,480 linked ECG-lab-value samples from 127,994 MIMIC-IV patients, the model reaches AUROC > 0.7 for 24 distinct lab abnormalities in the prediction setting and 24 in the forecasting setting, spanning cardiac, renal, hematological, metabolic, immunological, and coagulation categories -- with NT-proBNP elevation the best-predicted marker (AUROC 0.90), followed by hemoglobin, albumin, and hematocrit derangements (AUROC > 0.82). Code for dataset construction, training, and evaluation is public under an MIT license; no pretrained model weights are released.

Generalist reconstruction foundation model for accelerating cardiac MRI (CMR) acquisition, designed to recover diagnostic-quality images from highly undersampled (8x-24x) multi-coil k-space data across heterogeneous scanners, field strengths, and cardiovascular diseases. Combines a CLIP-ViT-based module for semantic/contextual understanding of the anatomy being imaged with a physics-informed data-consistency reconstruction network, trained on MMCMR-427K -- the largest public multimodal CMR k-space database to date. Intended as an upstream substrate that feeds downstream segmentation, phenotyping, and diagnosis models (e.g. automated cardiac-phenotype extraction via nnU-Net) rather than replacing them. Released by the CMRxRecon-challenge consortium; code and the underlying database are public for academic, non-commercial use, but no separately downloadable pretrained checkpoint is provided.

Fully automated pipeline for interpreting coronary angiograms that chains four purpose-built neural networks: (1) angiographic projection-angle identification, (2) left/right coronary artery detection, (3) arterial segment localization, and (4) stenosis-severity estimation. Trained on 13,843 angiographic studies (195,195 videos) from 11,972 adult patients at UCSF (2008-2019), with projection-angle and LCA/RCA-detection tasks each reaching precision/sensitivity/F1 at or above 90%. For predicting obstructive coronary artery disease (>=70% stenosis), CathAI reaches an AUC of 0.862 internally, 0.869 on external angiograms from the University of Ottawa Heart Institute, and 0.775 after retraining on quantitative-coronary-angiography labels from the Montreal Heart Institute core lab. No public code or model weights have been released.

Multi-view foundation model for coronary angiography trained with video-text contrastive learning on 203,808 angiography videos from 28,117 patients across 32,473 studies at the Montreal Heart Institute, externally validated on 4,249 studies from UCSF. Integrates multiple angiographic projections with attention-based pooling for study-level assessment spanning diagnostic, prognostic, and disease-progression tasks: significant-stenosis detection (AUROC 0.888 internal / 0.89 external), stenosis-percentage estimation (MAE 13.6% vs. 19.0% for clinical reports), chronic total occlusion, intracoronary thrombus, and coronary calcification detection. Transfer learning further enables one-year MACE prediction (AUROC 0.79) and LVEF estimation (MAE 7.3%) from the same angiography embeddings, with a mean in-hospital inference time of 4.2 seconds.

AI-driven pipeline for quantitative coronary-stenosis assessment from routine DICOM coronary angiography videos, combining vessel tracking with a video Swin3D transformer trained and validated on 182,418 angiography videos spanning 5 years at the Montreal Heart Institute. Achieves a mean absolute error of 20.15% and a classification AUROC of 0.8294 for stenosis-percentage prediction against cardiologist assessment, with lower inter-rater variability than two expert interventional cardiologists, and can be fine-tuned to quantitative coronary angiography (QCA) data for even lower error (MAE 7.75%).

Supervised EfficientNetV2-based 12-lead ECG model trained on over 1 million ECGs from the Montreal Heart Institute to predict 77 cardiac conditions derived from American Heart Association recommendations, plus fine-tuned digital-biomarker heads for reduced LVEF, 5-year atrial-fibrillation risk, and long-QT-syndrome (LQTS) detection/genotyping. Validated on 881,403 ECGs across 11 geographically diverse cohorts (4 public, 7 private health systems), achieving AUROCs above 0.98 for the 77-condition interpretation task while being 60x smaller and 29x faster at inference than its self-supervised DeepECG-SSL counterpart, with up to 9.7x lower CO2 emissions on equivalent tasks.

Self-supervised EfficientNetV2-based 12-lead ECG foundation model pretrained via contrastive learning and masked-lead modeling on 1.9 million ECGs (Montreal Heart Institute plus CODE-15% and MIMIC-IV), then fine-tuned for the same 77-condition ECG interpretation task and digital-biomarker extraction as DeepECG-SL. Outperforms the supervised counterpart on label-scarce digital-biomarker tasks, with the largest gains on LQTS genotype classification (AUROC 0.931 vs. 0.850, n=127 ECGs) and 5-year atrial-fibrillation risk (AUROC 0.742 vs. 0.734, n=132,050 ECGs), and outperforms ECG-FM and ECGFounder on shared external diagnostic classes.

Deep learning framework, developed in collaboration with the MONAI community, for automatic segmentation of tricuspid valve leaflets from transthoracic 3D echocardiograms in children with hypoplastic left heart syndrome (HLHS) and other forms of single-ventricle congenital heart disease, integrated into 3D Slicer via MONAILabel for interactive clinical/research use. Addresses a modality (pediatric 3D echocardiography) and population (single-ventricle congenital heart disease) largely absent from adult-focused cardiac AI models.

Open-weight video-based deep neural network that predicts reduced right-ventricular systolic function (RVSF) directly from routine left and right coronary angiogram videos, enabling real-time RV-dysfunction screening in the catheterization lab when echocardiography is unavailable. Built on an X3D-M spatiotemporal video architecture (Kinetics-400 pretrained) that aggregates per-video probabilities into a study-level normal-vs-reduced RVSF classification, with Grad-CAM/Guided-Backpropagation explainability confirming attention to RV-specific coronary motion rather than left-ventricular signal. Trained on 8,053 angiographic studies from 6,923 Montreal Heart Institute patients (2017-2023), externally validated at UCSF, and prospectively deployed at MHI via the PACS-AI platform, where AI assistance improved reader accuracy from 72.1% to 77.6% for cardiologists and 43.5% to 64.0% for medical students.

Multimodal large language model for ECG medical-report generation and cardiology conversational question-answering. An ECG-CoCa encoder (contrastive ECG-report pretraining in the style of OpenCLIP) is paired with a LLaVA-style vision-language architecture and an LLM backbone, fine-tuned on a purpose-built 45k-example ECG-instruction dataset (19k diagnosis examples + 25k multi-turn dialogue examples) built from five public 12-lead ECG datasets. Produces free-text diagnostic reports and supports zero-shot ECG-report retrieval classification.

Self-supervised ECG representation-learning method that adapts Joint-Embedding Predictive Architecture (JEPA) -- originally developed for images -- to 1D electrocardiogram signals. A Vision Transformer encoder (ViT-XS/S/B) is pretrained to predict masked temporal segments of the ECG directly in latent feature space, using a masking strategy tailored to time-series, on more than 1 million ECGs pooled from MIMIC-IV-ECG, CODE-15%, PTB-XL, Chapman-Shaoxing, CPSC2018/Extra, Georgia, PTB, and St-Petersburg-INCART. After fine-tuning on PTB-XL, the ViT-S/JEPA model reaches 0.945 AUC on the all-statements diagnostic task, exceeding prior self-supervised ECG baselines including CPC and ST-MEM. Developed at the Zuse Institute Berlin.

Systematic study of domain specialization for large language models in electrocardiography, comparing supervised fine-tuning (QLoRA) against retrieval-augmented generation (RAG) as two paths to inject ECG/cardiology knowledge into open-weight Llama 3.1 models (8B and 70B). Question-answer and multiple-choice pairs were generated from ECG/cardiology literature and used both for fine-tuning and for a multi-layered evaluation (multiple-choice accuracy, text-similarity metrics, LLM-as-a-judge, and blinded human-cardiologist review). The fine-tuned Llama 3.1 70B ranked first overall, exceeding the RAG variants and Claude Sonnet 3.7 on in-distribution multiple-choice and text-similarity metrics, though RAG and Claude generalized better to semantically complex, out-of-distribution questions. Developed by AI4Health at the University of Oldenburg with Charite Berlin; the finetuning/RAG/evaluation code is public, but per the paper's data-availability statement neither the training corpus nor the fine-tuned weights are released (German copyright law, section 60d UrhG).

Explainable Inception-style 1D CNN for multi-label arrhythmia detection from 12-lead ECGs, integrating Grad-CAM visualization to highlight the waveform segments driving each prediction. Trained on MIMIC-IV-ECG and externally validated on PTB-XL across atrial fibrillation, sinus tachycardia, conduction disturbances (RBBB/LBBB/LAFB), long QT, Wolff-Parkinson-White pattern, and paced-rhythm detection, with all metrics exceeding 90% internally and strong generalization on external validation.

Multi-instance contrastive self-supervised learning framework for echocardiography video representation learning: two distinct videos from the same patient exam are treated as positive pairs (rather than augmentations of a single clip), and a frame-reordering pretext task on temporally shuffled frames encourages the 3D-CNN backbone to learn temporally coherent representations. After self-supervised pretraining on unlabeled echocardiograms, the backbone is efficiently fine-tuned for cardiac disease classification (severe aortic stenosis, left ventricular hypertrophy) using very few labeled examples, outperforming SimCLR, standard multi-instance SimCLR, and Kinetics-400-initialized baselines across a range of training-data ratios.

Framework for training an echocardiography left-ventricle segmentation model purely by data-free knowledge distillation: a ConvLSTM-based student network learns to reproduce the masks produced by an EchoNet-Dynamic (DeepLabV3-ResNet50) teacher on entirely synthetic echo videos, with no real labeled data or even real videos required. Achieves state-of-the-art results identifying end-diastolic/end-systolic frames, reaching segmentation quality close to real-data training with substantially fewer weights; also introduces a human-annotation-free evaluation method using a large auxiliary model.

Predicts cardiac MRI (CMR) tissue-characterization findings -- wall-motion abnormalities and myocardial scar -- directly from standard transthoracic echocardiography videos (A4C/A2C/PLAX views), using a factorized 3D R2+1D convolutional network. Trained and validated on a single-institution cohort of more than 1,400 patients with paired echo and CMR studies within 30 days of each other. Released weights and inference code cover the two binarized outcomes (wall motion, scar); the continuous CMR tissue markers (native T1, T2, ECV) evaluated in the paper are not part of the public release.

Largest federated cardiac CT analysis to date (n=8,104 scans) across a real-world federation of German university hospitals, addressing partially-labeled data across sites via a two-step semi-supervised knowledge-distillation strategy: task-specific CNNs first predict on unlabeled data per label type, then a SWIN-UNETR transformer learns from these predictions with label-specific heads. Learns a single federated model that simultaneously predicts TAVI-relevant landmarks (aortic hinge points, coronary ostia, membranous septum) and calcification from cardiac CT, improving generalizability over UNet-based baselines on downstream tasks.

First multimodal LLM to unify ECG time series, 12-lead ECG images, and text for grounded, clinician-aligned ECG interpretation. A dual-encoder framework (ECG-CoCa time-series encoder plus a LLaVA-style vision-language backbone) extracts complementary time-series and image features with cross-modal alignment, trained on knowledge-guided instruction data (ECG-Grounding, linking diagnoses to measurable waveform parameters such as QRS/PR intervals) plus the 1.15-million-conversation ECG-Instruct corpus. Introduces the "Grounded ECG Understanding" benchmark and improves predictive performance, explainability, and grounding over prior ECG-language models such as ECG-Chat and PULSE.

Multi-scale ECG-language pretraining model that aligns 12-lead ECG signals with clinical text reports at three granularities -- token, beat, and rhythm level -- rather than a single global embedding. First fine-tunes a cardiology-specialized text encoder to improve understanding of ECG report language, then trains an ECG-FM-initialized ECG encoder against it with hierarchical contrastive supervision. Outperforms prior ECG-language and self-supervised baselines including MERL, ST-MEM, and HeartLang on zero-shot classification, linear probing, and ECG report generation, with especially large gains at low label fractions. Developed at the University of Hong Kong (HKU-MedAI).

Multi-instance-learning (MIL) model for detecting >=50% coronary stenosis directly from curved multiplanar reformation (CMR) images generated during routine coronary CT angiography (CCTA) reads, without requiring slice-level annotations. A multi-range Hounsfield-unit preprocessing pipeline (Sobel edge detection across five attenuation windows) highlights plaque and vessel-wall structures, which a VGG16-based encoder with positional encoding and multi-head attention aggregates across each patient's 'bag' of up to 36 CMR slices per artery to give an interpretable, attention-weighted patient-level prediction. Trained and five-fold cross-validated on 900 real-world CCTA cases (776 LAD / 694 RCA / 600 LCX) from Sahlgrenska University Hospital, reaching AUCs of 0.91-0.92 across the three major coronary arteries. Code (preprocessing + MIL training pipeline) is public; the clinical CMR dataset and trained weights are not released.

LoRA-adapted domain-specialized cardiology text embedding model built on MPNet-base (109M parameters), identified as Pareto-optimal for balanced accuracy/throughput deployment among 10 encoder- and decoder-style architectures benchmarked for cardiology semantic retrieval. LoRA fine-tuning on ~150,000 cardiology-textbook-derived sentence pairs raised its cardiology semantic-separation score from 0.175 (zero-shot) to 0.386, while delivering 228.8 embeddings/sec at a sub-1GB (0.73GB) memory footprint, making it suitable for consumer-GPU and general-purpose medical NLP deployment where full BioLinkBERT-level accuracy is not required.

Adapts the Segment Anything Model (SAM) to echocardiography video segmentation by giving it a space-time memory that carries both spatial and temporal cues, so that only the first frame of a video needs an external point prompt and every subsequent frame is segmented from a propagated memory prompt instead. A memory reinforcement mechanism uses each frame's predicted mask to suppress speckle-noise features before they are written back into memory, addressing a key failure mode of naively adapting video object segmentation (e.g. XMem) to noisy ultrasound. Built on SAMUS (an ultrasound-adapted SAM) with a frozen SAM backbone and only the image-encoder adapter layers trained. On the semi-supervised CAMUS and EchoNet-Dynamic benchmarks (only end-diastole/end-systole frames labeled), MemSAM reaches 93.3% and 92.8% mean Dice respectively, outperforming UNet, SwinUNet, H2Former, and prior medical-SAM adaptations (MedSAM, MSA, SAMed, SonoSAM, SAMUS) with far fewer prompts, and derives LVEF (via Simpson's biplane method of disks) with a Pearson correlation of 78.9% against ground truth on CAMUS. Training/inference code is public (MIT license); only the starting SAM ViT-B checkpoint is linked for download, not a separately released fine-tuned MemSAM checkpoint.

Ensemble of 3D convolutional neural networks that detects severe aortic stenosis directly from single-view 2D parasternal long-axis (PLAX) transthoracic echocardiogram videos, without requiring Doppler imaging. Representations are first pretrained with patient-level contrastive self-supervised learning on PLAX clips, then fine-tuned for binary AS classification; the ensemble is externally validated across a temporally-distinct cohort and geographically-distinct cohorts in California and New England. No pretrained weights are released; only the training/evaluation pipeline is public.

Patch-based masked-training framework for robust arrhythmia detection from digitized, multi-layout ECG images (e.g. 3x4, 2x6, 12x1 printed/scanned layouts), designed to handle the asynchronous lead timing and partial signal blackout that digitization introduces. An adaptive variable block-count masking strategy focuses model attention on key patches with cross-lead dependencies. Evaluated on PTB-XL digitized into multiple synthetic layouts and externally validated on 400 real digitized ECG images from Chaoyang Hospital, outperforming classical imputation baselines and the CNN foundation model ECGFounder.

Self-supervised foundation model for photoplethysmography (PPG) that explicitly targets the low-signal-quality problem endemic to real-world wearable and bedside PPG, rather than assuming clean input. A Siamese ResNet (SimSiam-style) encoder is pretrained to produce similar representations for temporally-adjacent high- and low-quality PPG segments presumed to share the same underlying physiological state, with curriculum learning that gradually increases the artifact-level gap between paired segments. Pretrained on over 36 million 30-second PPG pairs drawn from more than 24,100 UCSF ICU patients, then fine-tuned and evaluated on six downstream cardiovascular-monitoring datasets/tasks (heart rate, respiratory rate, blood pressure, and atrial-fibrillation detection), where it exceeds prior state-of-the-art by roughly 75% on respiratory-rate estimation and 5% on AF detection, with performance scaling with backbone depth (ResNet152 best). Code and the pretraining dataset are not public; the pretraining data is available upon request under a UCSF-Emory data use agreement.

Self- and weakly-supervised pipeline for left-ventricle segmentation across the full cardiac cycle in apical-4-chamber echocardiography videos. A video segmentation network (2D super-image or 3D U-Net encoder) is first pretrained with a self-supervised temporal-masking objective on largely unannotated echo frames, then fine-tuned with weak supervision from the sparse end-diastole/end-systole frame labels that most echo datasets provide. Achieves 93.3% Dice on EchoNet-Dynamic, outperforming nnU-Net and non-SSL baselines, and generalizes to the external CAMUS dataset. Developed by the BioMedIA group at MBZUAI.

ECG foundation model designed to remain accurate when leads are missing or signals are noisy. A 1D ConvNeXt V2 encoder is trained with a dual-mode self-distillation objective (separate lead-missing and noise "teachers") alongside contrastive alignment to detailed diagnostic-criteria text reports retrieved via a lightweight, LLM-free "Cardiac Feature Retrieval" module. Consistently ranks best or second-best across PTB-XL diagnostic tasks and MIT-BIH arrhythmia classification under original, noisy, lead-missing, and combined-corruption conditions. Developed by FPT Software AI Center and the University of Arkansas.

Multimodal cardiac MRI foundation model that fuses 3D+T cine CMR (short-axis and long-axis views) with tabular patient health records (demographics, metabolic, and lifestyle factors) from 42,000 UK Biobank participants. Two-stage self-supervised pretraining -- masked-image reconstruction, then imaging-tabular contrastive alignment -- produces representations that transfer to whole-heart segmentation, cardiac phenotype/physiological-feature regression, and cardiac/metabolic disease classification within one unified framework.

Foundation model that encodes single-lead (lead I) ECGs with information from paired transthoracic echocardiography reports, aimed at label-efficient screening for structural heart disease (SHD) on wearable and portable single-lead ECG devices. A 7-layer 1D-CNN ECG encoder and a RoBERTa-based text encoder are contrastively pretrained (CLIP-style) on 194,551 ECG-echo report pairs from 77,378 adults in the Yale New Haven Health System, then the ECG encoder is fine-tuned on a temporally-distinct cohort to detect reduced LVEF, diastolic dysfunction, and a composite SHD label. Matches a randomly-initialized CNN at full training-data volume but substantially outperforms it in label-scarce regimes (e.g. with only 0.5% of labeled data).

Multi-agent LLM framework, deployed as a Software-as-a-Medical-Device on AWS, that assists cardiologists reading 24-hour Holter/patch ECG monitoring studies. Three fine-tuned LLM agents divide the diagnostic workflow the way a cardiologist would: a table-to-text agent (Llama-3.1-8B) extracts findings from tabular arrhythmia metrics, an image-to-text agent (LLaVA-v1.5-13B) extracts findings from ECG tracing images, and a findings-to-interpretation agent (Llama-3.1-8B) synthesizes both against clinical guidelines with a fact-checking step. Each agent is instruction-tuned on cardiologist-adjudicated reports from 2,000+ real-world patients and further steered at inference with in-context demonstrations matched to the patient's age, sex and arrhythmia class. In blinded cardiologist ratings across eight clinical/security metrics (1-5 scale), ZODIAC outperformed GPT-4o, Gemini-Pro, Llama-3.1-405B, Mixtral-8x22B, and medical-specialist LLMs (BioGPT, Meditron, Med42) on every metric while using under 30B total parameters, and has been integrated into commercial ECG monitoring devices. This is a proprietary product; no public code or model weights have been released.

ECG foundation model built on the xLSTM (extended LSTM) architecture: a bidirectional stack of nine alternating scalar- and matrix-memory LSTM blocks that scales linearly with sequence length, unlike the quadratic cost of transformer-based ECG models. Pretrained with SimDINOv2, a coding-rate-regularized self-distillation (DINO) objective adapted from computer vision to ECG time series, on roughly 8 million recordings from CODE, INCART, and Chapman-Shaoxing-Ningbo. Introduced alongside BenchECG, a standardized 8-dataset/10-task benchmark, on which xECG achieves the best average rank of any publicly available ECG foundation model, with particular strength on long-context tasks (30-minute ambulatory arrhythmia classification, multi-hour sleep-apnea segmentation) where transformer-based models are computationally limited.

Open-source, reproducible pipeline for representing 12-lead ECGs as explicit graphs -- nodes per lead-timepatch, with edges encoding established inter-lead spatial relationships (fully-connected limb- and chest-lead subgraphs bridged via leads I, aVF, V4, and V5) -- and classifying them with a Graph Convolutional Network, paired with GNNExplainer to surface which leads and lead-pairs drove each prediction. Evaluated on PTB-XL for five-class diagnostic superclass classification (AUC 0.86) and, with the same architecture, on anteroseptal-vs-inferior myocardial-infarction localization (AUC 0.92), externally validated on the population-based SHIP cohort (AUC 0.87). Explainability analysis showed the GNN's lead attention recovers standard ECG diagnostic criteria (e.g. V1-V3 for anteroseptal MI, II/III/aVF for inferior MI). Developed at University Medical Center Gottingen; code and an example trained checkpoint are released under CC BY-NC 4.0.

Fully automated deep learning workflow for characterizing cardiac mechanics from balanced steady-state free-precession (bSSFP) cine cardiac MRI. It decouples two convolutional networks—a segmentation net (CarSON) and a 3D motion-estimation net (CarMEN)—to derive left- and right-ventricular volumes plus global and regional myocardial strain and strain rate without manual tracing. Trained and validated on healthy and cardiovascular-disease subjects and shown to be robust across MRI vendors, with excellent intra-scanner repeatability for strain. Developed at Massachusetts General Hospital and the Harvard-MIT Division of Health Sciences and Technology.

Video-based deep learning model that grades aortic regurgitation (AR) severity—none/trace, mild, moderate, or severe—from color Doppler echocardiography. View-specific R(2+1)D 3D-CNNs analyze five standard transthoracic views (PLAX, PLAX-AV, A3C, A3C-AV, A5C) and their outputs are combined by a maximum-severity rule at the study level. Trained on ~47,600 color Doppler videos from Cedars-Sinai and externally validated at Stanford Healthcare, reaching AUCs of 0.95 for at-least-moderate AR and 0.97 for severe AR internally. Developed by the Ouyang lab at Cedars-Sinai Medical Center.

Open-source pipeline that classifies aortic stenosis (AS) severity from transthoracic echocardiography by combining structural and functional information. Video-based R(2+1)D convolutional networks read six B-mode and color Doppler views while a segmentation model measures peak aortic-jet velocity, and an ensemble integrates these into a final severity prediction. Trained on 210,193 images from Kaiser Permanente Northern California and validated across held-out, temporally distinct, and external Stanford and Cedars-Sinai cohorts, reaching AUCs up to 0.96–0.99 for severe AS. Developed by the Ouyang lab.

Video-based deep learning model that estimates 14 common blood biomarkers and laboratory values—including hemoglobin (anemia), B-type natriuretic peptide (BNP), troponin I, and blood urea nitrogen (BUN)—directly from apical-4-chamber echocardiogram videos. Built on a spatiotemporal convolutional network (R(2+1)D-style) with residual connections that produces beat-by-beat estimates for both regression and abnormality classification. Trained on over 70,000 echocardiograms from Stanford Healthcare and externally validated at Cedars-Sinai, reaching AUCs around 0.80–0.86 for detecting anemia and elevated BNP. Developed by the Ouyang and Zou labs at Stanford University and Cedars-Sinai.

Deep learning model for automated phenotyping of mitral stenosis (MS) from echocardiography. Uses video-based R(2+1)D convolutional networks on color Doppler and B-mode views to identify and grade mitral stenosis severity, following the multi-view valvular-assessment approach of the EchoNet family. Trained and validated on large clinical echocardiography cohorts from Kaiser Permanente Northern California with external testing at Stanford Healthcare and Cedars-Sinai. Developed by the Ouyang lab.

CNN-based image classifier that screens 12-lead ECG images for a signature of transthyretin amyloid cardiomyopathy (ATTR-CM), producing a study-level probability score. Trained on a private Yale New Haven Health System cohort of nuclear-imaging-confirmed ATTR-CM cases and matched controls. Used alongside a companion echocardiography model to track pre-clinical ATTR-CM progression years before it would otherwise be confirmed by nuclear amyloid imaging. Developed by Yale's CarDS Lab and distributed as a packaged research-use executable rather than downloadable weights.

EfficientNet-B3 CNN that detects hypertrophic cardiomyopathy directly from images of printed or scanned 12-lead ECGs, rather than from raw digital waveforms, enabling screening from a photo of a paper tracing. Initialized via self-supervised contrastive pretraining on patient identity, then fine-tuned at Yale New Haven Hospital on over 124,000 ECGs from about 67,000 patients, with HCM status confirmed by cardiac MRI or echocardiography. Externally validated on ECG images from MIMIC-IV, Amsterdam UMC, and UK Biobank. Developed by Yale's CarDS Lab.

Screens 12-lead ECGs for Chagas cardiomyopathy by first pretraining a feature extractor to predict blood-biomarker levels from MIMIC-IV-ECG data, then fine-tuning on Brazilian CODE-15%, SaMi-Trop, and PTB-XL recordings; the final model is a 5-model ensemble. Submitted to the George B. Moody PhysioNet Challenge 2025 (Detection of Chagas Disease from the ECG), where it placed 5th on the official leaderboard. Developed by a team from Akershus University Hospital and the University of Oslo.

Automated pipeline that segments both heart ventricles and tracks their motion throughout the cardiac cycle from short-axis cine cardiac MRI, producing 3D bi-ventricular models with per-vertex wall-thickness and curvature measurements over time. Built on a shape-refined multi-task fully convolutional network, followed by non-rigid registration and mesh-based motion tracking. Trained on roughly 400 manually annotated pulmonary hypertension patients as part of Imperial College London's UK Digital Heart Project, and underlies the related 4Dsurvival cardiac-motion survival-prediction study.

1D residual neural network that screens 10-second, 12-lead ECG tracings for six common abnormalities: first- and second-degree AV block patterns, right and left bundle branch block, sinus bradycardia, atrial fibrillation, and sinus tachycardia. Reported F1 scores above 80% and specificity over 99% when benchmarked against cardiology residents. Developed at Universidade Federal de Minas Gerais and trained on the large Brazilian CODE-15% ECG dataset.

Single-lead ECG foundation model pretrained with clinically-guided contrastive learning: rather than relying on hand-labeled tasks, it uses routinely collected clinical metadata and risk scores from 161,000 MIMIC-IV-ECG patients as the training signal. Released in three sizes - Small (~448K parameters), Medium (30.7M), and Large (~296M) - and benchmarked against other ECG foundation models like ECGFounder across 18 tasks and 7 held-out datasets. Developed by Nokia Bell Labs.

Vision-language model that jointly embeds a cardiac MRI study, treated as video, with the impression section of its clinical report. Combines a video encoder over cine/LGE frame sequences with a Bio+ClinicalBERT text encoder using CLIP-style contrastive training. Supports zero-shot and few-shot classification of cardiomyopathies, amyloidosis, and LV dysfunction, plus image/report retrieval and structured report drafting. Trained on a private, single-institution corpus of roughly 11,000-14,000 CMR study-report pairs from Cleveland Clinic and Case Western.

Domain-specialized text embedding model for clinical cardiology, built by fine-tuning the Qwen3-Embedding-8B language model with LoRA adapters via contrastive learning on cardiology textbook sentences. Reaches 99.60% top-1 accuracy on cardiology-specific semantic retrieval, nearly 16 points above the prior MedTE baseline. The training corpus draws on roughly 150,000 sentences from seven copyrighted textbooks and is not public, though the resulting model weights are freely downloadable.

Foundation model for cine cardiac MRI, self-supervised via masked autoencoding on nearly 75,000 UK Biobank scans. Uses a Vision Transformer with a convolutional stem, unified across long-axis and short-axis views. Fine-tuned checkpoints are released for ventricle and myocardium segmentation, ejection-fraction regression, cardiovascular disease classification, and landmark localization across several public benchmark datasets (ACDC, M&Ms, M&Ms2, EMIDEC, and others).

Multi-task model that jointly assesses signal quality and detects atrial fibrillation from wrist-worn wearable photoplethysmography (PPG), pretrained on roughly one million simulated unlabeled signals before fine-tuning on labeled wearable data. Uses a 1D CNN with separate output heads for signal quality and arrhythmia detection. Developed by Stanford's Ashley Lab.

Fully automatic pipeline that localizes the heart, segments coronary calcium, and produces an Agatston-style coronary artery calcium score from gated and non-gated chest/cardiac CT. A three-stage 3D CNN performs each step in sequence. Validated across the Framingham, NLST, PROMISE, and ROMICAT-II cohorts, where the resulting calcium score predicted cardiovascular events with hazard ratios up to 4.3. Developed by the Harvard AIM Lab.

Reconstructs digital 12-lead ECG waveforms from scanned or photographed paper printouts, using an nnU-Net image segmentation model to trace the signal pixels followed by a Hough-transform-based reconstruction pipeline. This is a digitization tool rather than a diagnostic model - it recovers a usable signal from a paper record rather than producing a diagnosis. Won the PhysioNet/Computing in Cardiology Challenge 2024; developed by a team at the University of Oxford.

Dual-encoder single-lead ECG classifier for atrial fibrillation detection that fuses a raw-signal branch with a spectrogram branch via axial attention and a Transformer. Originally developed as a graduate-course project at TU Darmstadt for the 2017 PhysioNet/CinC Challenge, and later extended in a 2023 follow-up study. Released in four sizes up to 130M parameters (S/M/L/XL), alongside a simpler CNN+LSTM variant.

Open ECG foundation model with 90.9M parameters, built on a wav2vec 2.0-style Transformer and pretrained on 1.25-1.5 million ECGs using a hybrid contrastive-and-generative self-supervised objective. Base pretrained weights and MIMIC-IV-ECG-finetuned downstream checkpoints are both released. Developed on the fairseq_signals framework by the University of Toronto / Vector Institute's Wang lab.

1D residual neural network that predicts a patient's age directly from a 12-lead ECG; the gap between this predicted 'ECG age' and true chronological age is used as a biomarker of cardiovascular risk and mortality. Trained on the CODE-15% Brazilian ECG dataset, with the original R² of 0.71 later reproduced (R² = 0.70) in an independent German validation cohort. Developed by researchers at Uppsala University and UFMG.

Multi-task 12-lead ECG model with output heads for incident atrial-fibrillation risk (as a survival curve), incident mortality risk, prevalent AF classification, sex classification, and age regression. Built on a 1D CNN over the raw waveform, and developed by the Broad Institute's ML4H group as a successor to their ECG-AI model published in Circulation. Trained on ECGs from UK Biobank and Massachusetts General Hospital, neither of which is publicly released.

Large-scale ECG foundation model pretrained on more than 10 million recordings spanning 150 label categories from the Harvard-Emory ECG Database. Built as a general-purpose feature extractor that can be fine-tuned for arrhythmia detection, demographic inference, and event prediction, and externally validated on MIMIC-IV-ECG and PTB-XL. Also used as the pretrained backbone for downstream clinical models such as Pocket-K, a hyperkalemia detector. Developed by Peking University and Harvard-Emory researchers.

Multimodal ECG model that pairs a 1D ConvNeXtV2 signal encoder with a BioLinkBERT text encoder, trained with a joint contrastive-and-captioning objective using LLM-generated descriptions of ECG demographics and waveform patterns in place of raw clinical reports. Validated on arrhythmia diagnosis and ECG-based subject identification, reaching an AUROC of 0.938 fine-tuned and 0.812 zero-shot on PTB-XL diagnostic classification. Developed at Rice University.

Vision-language foundation model fine-tuned from CLIP on more than one million private echocardiogram video-report pairs, enabling zero-shot cardiac function assessment, device identification, and image/text retrieval without task-specific training. Combines a ConvNeXt-Base video encoder with a GPT-2-style text encoder under contrastive pretraining. Training data is private, but model weights and code are public. Developed by Cedars-Sinai's Ouyang lab.

General-purpose vision foundation model for echocardiography, pretrained with a masked autoencoder combined with a periodic contrastive loss designed around the cyclical nature of cardiac motion. Validated on chamber segmentation, view classification, and disease detection, with its largest advantage over non-pretrained baselines and natural-image models like SAM appearing in low-label settings. Pretrained on roughly 290,000 echo clips from a mix of internal and public sources. Developed by Massachusetts General Hospital and Harvard Medical School.

Echocardiography foundation model trained with a latent-predictive (V-JEPA2-style) self-supervised objective rather than pixel reconstruction, pretrained on 18 million echocardiograms from 300,000 patients drawn from the public MIMIC-IV-ECHO dataset plus a private multi-site archive - reportedly the largest echo pretraining corpus assembled to date. With a frozen backbone and only lightweight added layers, it outperforms prior echo foundation models by roughly 20% on ejection-fraction estimation and 17% on right-ventricular pressure estimation, reaches strong view-classification accuracy using just 1% of labels, and transfers zero-shot to pediatric echo better than fully fine-tuned baselines. Developed by the University of Toronto's Bo Wang Lab.

Predicts a patient's age from echocardiogram videos across four standard views (PLAX, A2C, A4C, and subcostal), trained on a private multi-site cohort of over 2.6 million videos from more than 166,000 studies across roughly 90,000 patients. The gap between this AI-predicted age and true chronological age is studied as a marker of accelerated or delayed cardiovascular aging and its relationship to all-cause mortality. Uses a 3D CNN (R(2+1)D) with a separate pretrained model per view. Developed by Cedars-Sinai Medical Center and Stanford's Ouyang lab.

End-to-end pipeline for apical-4-chamber echocardiogram videos that segments the left ventricle, estimates ejection fraction on a beat-to-beat basis, and classifies cardiomyopathy with reduced ejection fraction. Combines a DeepLabV3-ResNet50 segmentation model with a 3D CNN (R2+1D/R3D/MC3) initialized on the Kinetics-400 video dataset. Trained on the public EchoNet-Dynamic dataset released alongside it, and one of the most widely reused open echocardiography models since its 2020 Nature publication. Developed by Stanford University.

Measures interventricular septum thickness, LV internal diameter, and posterior wall thickness from PLAX echocardiogram videos, then classifies the underlying cause of left ventricular hypertrophy as either cardiac amyloidosis or hypertrophic cardiomyopathy. Combines an atrous-convolution 2D CNN for wall-thickness segmentation with a 3D residual CNN for etiology classification. Trained on 28,201 videos across Stanford, Cedars-Sinai, and the Unity Imaging Collaborative. Developed by Stanford University.

Fully automated pipeline that scans a complete transthoracic echocardiogram study, identifies the apical-4-chamber color-Doppler clips showing the mitral valve, and grades mitral regurgitation severity at the study level. Combines a view/valve-presence classifier with a spatiotemporal CNN for severity classification. Trained on a private Cedars-Sinai cohort of 58,614 studies and externally validated on 915 studies from Stanford Healthcare.

Automates standard echocardiographic measurements from video, pairing a measurement model with a companion segmentation component. Developed by Stanford and Cedars-Sinai's Ouyang lab; public documentation on the exact measurements covered, training data, and validation performance is limited compared to other EchoNet-family models.

Detects 12 categories of echocardiogram-confirmed structural heart disease from 12-lead ECG waveforms combined with demographic and clinical covariates. Uses the same architecture as the original, larger EchoNext model but is trained entirely on the public EchoNext-Mini dataset - 100,000 de-identified ECGs from Columbia University Irving Medical Center released on PhysioNet - making it one of the more fully reproducible models of its kind, with public weights, a Docker image, and inference code.

Vision-language foundation model that interprets an entire transthoracic echocardiogram study rather than a single view or video: it classifies the view type of every clip, applies view-informed attention across the full study, and generates or retrieves comprehensive study-level interpretations in English or Italian. Pretrained on a private Cedars-Sinai corpus of 12 million echo video-report pairs. Developed by the Smidt Heart Institute and Stanford's Ouyang lab.

Distills knowledge from EchoCLIP, a vision-language echocardiography model, into ECG embeddings, aiming to improve how well ECG signals alone can predict echo-derived measures of cardiac function. Combines a 1D ECG encoder with a BioBERT text encoder under a probabilistic cross-modal embedding objective that captures uncertainty. Published at MICCAI 2025 by the University of Toronto's McIntosh Lab.

GPT-style decoder-only Transformer pretrained via next-token prediction on tokenized single-lead ECG time series, producing an interpretable general-purpose model that can be fine-tuned for tasks like arrhythmia screening and beat detection. Individual attention heads are shown to respond to physiologically meaningful features such as the P-wave, and token embeddings cluster by position in the cardiac cycle. A companion PPG-pretrained model (PPG-PT) is released in the same repository. Developed at Imperial College London.

GPT-style decoder-only Transformer pretrained via next-token prediction on tokenized PPG time series, the companion model to ECG-PT (HeartGPT) in the same repository. Individual attention heads respond to physiologically meaningful waveform features such as the dicrotic notch, and the model can be fine-tuned for wearable-based cardiac screening tasks. Developed at Imperial College London.

Treats ECGs as a language: a QRS-Tokenizer converts raw waveforms into discrete heartbeat 'words' from a learned 8,192-entry vocabulary, and a spatio-temporal transformer (ST-ECGFormer) is pretrained via masked-sentence modeling over these tokens. Evaluated for robust, competitive performance across six public ECG datasets and published at ICLR 2025. Developed by Peking University's digital health group, pretrained on MIMIC-IV-ECG.