Why the study?
High-quality ECG data is scarce due to privacy concerns and limited resources, while existing generative models rely on small datasets, lack comprehensive evaluation frameworks, and overlook applications beyond data augmentation.
DiffuSETS provides a novel framework for generating high-fidelity, clinically meaningful 12-lead ECGs from text reports, which could mitigate data scarcity and aid in cardiology education.
May augment scarce ECG datasets for research; leaves open clinical validation and regulatory assessment before any diagnostic use.
The scarcity of high-quality electrocardiogram (ECG) data, driven by privacy concerns and limited medical resources, creates a pressing need for effective ECG signal generation. Existing approaches for generating ECG signals typically rely on small training datasets, lack comprehensive evaluation frameworks, and overlook potential applications beyond data augmentation. To address these challenges, we propose DiffuSETS, a framework capable of generating ECG signals with high semantic alignment and fidelity. DiffuSETS accepts various modalities of clinical text reports and patient-specific information as inputs, enabling the creation of clinically meaningful ECG signals. Additionally, we introduce a comprehensive benchmarking methodology to assess the effectiveness of ECG generative models. Our model achieves excellent results in tests, proving its superiority in the task of ECG generation. Furthermore, we showcase its potential to mitigate data scarcity while exploring applications in cardiology education and medical knowledge discovery.
No takes yet. Share an insight, caveat, or question.
Chen et al. (2025) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: