Conditional Electrocardiogram Generation Using Hierarchical Variational Autoencoders

2025-03-03

arXiv.org

AI Summary

To address the scarcity of high-quality, expert-annotated 12-lead electrocardiogram (ECG) data for cardiovascular disease diagnosis—due to high acquisition costs, labeling noise, severe class imbalance, and stringent privacy constraints—this work proposes the conditional Nested Variational Autoencoder for ECG (cNVAE-ECG). It is the first open-source, ECG-specific conditional hierarchical VAE architecture, integrating pathology label embedding, multi-scale feature disentanglement, hierarchical variational inference, and time-series reconstruction loss to enable fine-grained, controllable ECG generation. Clinically validated synthetic ECGs exhibit physiological plausibility and pathology-specific fidelity. In downstream transfer learning tasks, cNVAE-ECG achieves a 2% absolute improvement in AUROC over state-of-the-art ECG-GAN baselines, effectively mitigating small-sample limitations and generalization bottlenecks in low-data clinical scenarios.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Cardiovascular diseases (CVDs) are disorders impacting the heart and circulatory system. These disorders are the foremost and continuously escalating cause of mortality worldwide. One of the main tasks when working with CVDs is analyzing and identifying pathologies on a 12-lead electrocardiogram (ECG) with a standard 10-second duration. Using machine learning (ML) in automatic ECG analysis increases CVD diagnostics' availability, speed, and accuracy. However, the most significant difficulty in developing ML models is obtaining a sufficient training dataset. Due to the limitations of medical data usage, such as expensiveness, errors, the ambiguity of labels, imbalance of classes, and privacy issues, utilizing synthetic samples depending on specific pathologies bypasses these restrictions and improves algorithm quality. Existing solutions for the conditional generation of ECG signals are mainly built on Generative Adversarial Networks (GANs), and only a few papers consider the architectures based on Variational Autoencoders (VAEs), showing comparable results in recent works. This paper proposes the publicly available conditional Nouveau VAE model for ECG signal generation (cNVAE-ECG), which produces high-resolution ECGs with multiple pathologies. We provide an extensive comparison of the proposed model on various practical downstream tasks, including transfer learning scenarios showing an area under the receiver operating characteristic (AUROC) increase up to 2% surpassing GAN-like competitors.

Problem

• Generating synthetic ECG signals with specific pathologies
• Overcoming limitations of medical data for ML training
• Improving CVD diagnostics using conditional VAE models

Innovation

• Hierarchical Variational Autoencoders for ECG generation
• Conditional Nouveau VAE model for multi-pathology ECGs
• Improves AUROC by 2% over GANs in downstream tasks