What is the clinical evidence from this study?

Study design: Other. Population: Normal ECG (n=90589). Intervention: Generative Adversarial Networks (GANs) vs. Real ECG data. Primary outcome: Maximum Mean Discrepancy (MMD) and Dynamic Time Warping (DTW).

September 19, 2019Open Access

Synthesis of Realistic ECG using Generative Adversarial Networks

Key Result

A Generative Adversarial Network with bidirectional LSTMs and convolution-pooling layers successfully generated realistic synthetic ECG data while protecting the privacy of the original training set.

Structured PICO

Population

90,589 normal ECG records from the MIT-BIH Arrhythmia Database used to train and test generative adversarial networks.

Intervention

Generative Adversarial Networks (GANs) including LSTM, CNN, and BiLSTM architectures

Outcome

Quality of synthetic data assessed by Maximum Mean Discrepancy (MMD) and Dynamic Time Warping (DTW)

GAN architectures can successfully generate synthetic time series signals, such as ECGs, that are structurally similar to real data while protecting patient privacy.

Limitations

The training phase of GANs is highly unstable, with many experiments failing to converge.
Due to the computational complexity of DTW, Euclidean distance was used for privacy evaluation instead.
Further research is required to stabilize GANs and ensure they reduce the privacy risk associated with sharing sensitive medical time series data.

Abstract

Access to medical data is highly restricted due to its sensitive nature, preventing communities from using this data for research or clinical training. Common methods of de-identification implemented to enable the sharing of data are sometimes inadequate to protect the individuals contained in the data. For our research, we investigate the ability of generative adversarial networks (GANs) to produce realistic medical time series data which can be used without concerns over privacy. The aim is to generate synthetic ECG signals representative of normal ECG waveforms. GANs have been used successfully to generate good quality synthetic time series and have been shown to prevent re-identification of individual records. In this work, a range of GAN architectures are developed to generate synthetic sine waves and synthetic ECG. Two evaluation metrics are then used to quantitatively assess how suitable the synthetic data is for real world applications such as clinical training and data analysis. Finally, we discuss the privacy concerns associated with sharing synthetic data produced by GANs and test their ability to withstand a simple membership inference attack. For the first time we both quantitatively and qualitatively demonstrate that GAN architecture can successfully generate time series signals that are not only structurally similar to the training sets but also diverse in nature across generated samples. We also report on their ability to withstand a simple membership inference attack, protecting the privacy of the training set.

Mark Helpful

Bookmark

Relay

View Full Paper