Generative Adversarial Networks for Electronic Health Records: A Framework for Exploring and Evaluating Methods for Predicting Drug-Induced Laboratory Test Trajectories

TL;DR

This paper introduces a framework using GANs to generate and evaluate synthetic electronic health record data, specifically laboratory test trajectories, to improve drug impact predictions.

Contribution

It presents an unsupervised evaluation method for GAN-generated health data and demonstrates the benefit of cohort representation learning for better drug impact prediction.

Findings

GANs can generate realistic laboratory test trajectories.

Representation learning enhances predictive accuracy.

Unsupervised evaluation effectively measures synthetic data quality.

Abstract

Generative Adversarial Networks (GANs) represent a promising class of generative networks that combine neural networks with game theory. From generating realistic images and videos to assisting musical creation, GANs are transforming many fields of arts and sciences. However, their application to healthcare has not been fully realized, more specifically in generating electronic health records (EHR) data. In this paper, we propose a framework for exploring the value of GANs in the context of continuous laboratory time series data. We devise an unsupervised evaluation method that measures the predictive power of synthetic laboratory test time series. Further, we show that when it comes to predicting the impact of drug exposure on laboratory test data, incorporating representation learning of the training cohorts prior to training GAN models is beneficial.