Multi-Label Clinical Time-Series Generation via Conditional GAN

Abstract

With wide applications of electronic health records (EHR), deep learning methods have been adopted to analyze EHR data on various tasks such as representation learning, clinical event prediction, and phenotyping. However, due to privacy constraints, limited access to EHR becomes a bottleneck for deep learning research. Recently, generative adversarial networks (GANs) have been successful in generating EHR data. However, there are still challenges in high-quality EHR generation, including generating time-series EHR and uncommon diseases given imbalanced datasets. In this work, we propose a Multi-label Time-series GAN (MTGAN) to generate EHR data and simultaneously improve the quality of uncommon disease generation. The generator of MTGAN uses a gated recurrent unit (GRU) with a smooth conditional matrix to generate sequences and uncommon diseases. The critic gives scores using Wasserstein distance to recognize real samples from synthetic samples by considering both data and temporal features. We also propose a training strategy to calculate temporal features for real data and stabilize GAN training. Furthermore, we design multiple statistical metrics and prediction tasks to evaluate the generated data. Experimental results demonstrate the quality of the synthetic data and the effectiveness of MTGAN in generating realistic sequential EHR data, especially for uncommon diseases.

Generative AI Mitigates Representation Bias and Improves Model Fairness Through Synthetic Health Data

“…However, SMOTE, when used with high-dimensional time-series data, may decrease data variability and introduce correlation between samples [10][11][12]. In response, alternative approaches based on Generative Adversarial Networks (GAN) are gaining ground [13][14][15][16][17]. GANs have shown incredible results in generating realistic images [18], text [19], and speech [20] in addition to improving privacy [21].…”