Key result
CardiacFusion-Net classifies sudden cardiac arrest with ~0.95 AUC, outperforming single-modality and late-fusion models.
Why the study?
Sudden cardiac arrest risk may be expressed differently across anatomy, mechanical function, and electrical activity, but studies containing CMR, echocardiography, and ECG for the same patient are uncommon.
Does combining modality-specific deep-learning representations from cardiac magnetic resonance imaging, echocardiography, and 12-lead ECG via CardiacFusion-Net improve discrimination for sudden cardiac arrest classification compared with single-modality models and conventional fusion baselines in a matched multimodal benchmark?
Observational (n=1,204)
Yes
Does combining modality-specific deep-learning representations from cardiac magnetic resonance imaging, echocardiography, and 12-lead ECG via CardiacFusion-Net improve discrimination for sudden cardiac arrest classification compared with single-modality models and conventional fusion baselines in a matched multimodal benchmark?
In a proof-of-concept matched multimodal benchmark, fusing deep-learning representations from CMR, echocardiography, and 12-lead ECG improved discrimination for sudden cardiac arrest risk over single modalities and conventional fusion, while maintaining robustness when individual modalities were withheld.
Answers are grounded in this paper and cite their sources.
Requires prospective validation before clinical use; leaves open multimodal fusion value for SCA risk stratification.
Risk of sudden cardiac arrest may be expressed differently by anatomy, mechanical function and electrical activity, while studies that contain all three modalities for the same patient are uncommon. We developed CardiacGuard-Net, a modular deep-learning framework that learns modality-specific representations from cardiac magnetic resonance imaging (CMR), echocardiography and twelve-lead electrocardiography (ECG), then combines them for cardiac-arrest classification. The modality-specific models achieved mean Dice 0.953 for CMR segmentation, 3.8% mean absolute error for echocardiographic ejection-fraction estimation and AUC 0.938 for 24-h arrest prediction from ECG. For multimodal evaluation, 1,204 records were assembled into a matched benchmark using age band, sex, ejection-fraction band and diagnostic category because no dataset available to the study contained all three modalities for the same patients. On this benchmark, CardiacFusion-Net achieved AUC 0.947, compared with 0.920 for late fusion. Removing the ECG, CMR or echocardiographic branch reduced AUC by 0.035, 0.026 and 0.016, respectively. A single trained model also retained discrimination when modalities were withheld, with AUCs of 0.890 for ECG alone, 0.905 for MRI+echo, 0.921 for MRI + ECG and 0.931 for echo + ECG. The multimodal results are therefore best interpreted as an internal proof-of-concept benchmark of complementary information, not as patient-level clinical validation.
No takes yet. Share an insight, caveat, or question.
Jarin Prity (2026) conducted an observational in Sudden cardiac arrest risk (n=1,204). CardiacGuard-Net / CardiacFusion-Net (multimodal CMR, echocardiography, and 12-lead ECG deep-learning fusion) vs. Late fusion, early fusion, and single-modality baselines was evaluated on Sudden cardiac arrest classification discrimination (AUC-ROC) on the matched multimodal benchmark (AUC 0.947 vs 0.920). CardiacFusion-Net achieved an AUC of 0.947 for sudden cardiac arrest classification on a matched multimodal benchmark, outperforming late fusion (AUC 0.920) and ECG alone (AUC 0.890).
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: