ECGAN: Self-supervised generative adversarial network for electrocardiography

TL;DR

ECGAN is a self-supervised generative adversarial network that produces high-quality, conditionally diverse synthetic electrocardiography data, aiding biomedical research while respecting privacy constraints.

Contribution

This work introduces ECGAN, a novel self-supervised GAN for generating realistic, conditionally diverse ECG signals with a new quality assessment framework.

Findings

ECGAN outperforms existing models in ECG sequence generation.

The model effectively conditions on specific arrhythmias.

Synthetic data improves predictive model training.

Abstract

High-quality synthetic data can support the development of effective predictive models for biomedical tasks, especially in rare diseases or when subject to compelling privacy constraints. These limitations, for instance, negatively impact open access to electrocardiography datasets about arrhythmias. This work introduces a self-supervised approach to the generation of synthetic electrocardiography time series which is shown to promote morphological plausibility. Our model (ECGAN) allows conditioning the generative process for specific rhythm abnormalities, enhancing synchronization and diversity across samples with respect to literature models. A dedicated sample quality assessment framework is also defined, leveraging arrhythmia classifiers. The empirical results highlight a substantial improvement against state-of-the-art generative models for sequences and audio synthesis.