Generation of Synthetic Electronic Medical Record Text

By Jiaqi Guan, Runzhe Li, Sheng Yu, Xuegong Zhang. 2018

61 citations recorded by the source corpus.

Machine learning (ML) and Natural Language Processing (NLP) have achieved remarkable success in many fields and have brought new opportunities and high expectation in the analyses of medical data. The most common type of medical data is the massive free-text electronic medical records (EMR). It is widely regarded that mining such massive data can bring up important information for improving medical practices as well as for possible new discoveries on complex diseases. However, the free EMR texts are lacking consistent standards, rich of private information, and limited in availability. Also, as they are accumulated from everyday practices, it is often hard to have a balanced number of samples for the types of diseases under study. These problems hinder the development of ML and NLP methods for EMR data analysis. To tackle these problems, we developed a model to generate synthetic text of EMRs called Medical Text Generative Adversarial Network or mtGAN. It is based on the GAN framework and is trained by the REINFORCE algorithm. It takes disease features as inputs and generates synthetic texts as EMRs for the corresponding diseases. We evaluate the model from micro-level, macro-level and application-level on a Chinese EMR text dataset. The results show that the method has a good capacity to fit real data and can generate realistic and diverse EMR samples. This provides a novel way to avoid potential leakage of patient privacy while still supply sufficient well-controlled cohort data for developing downstream ML and NLP methods. It can also be used as a data augmentation method to assist studies based on real EMR data.

Subfields: Artificial Intelligence, General Social Sciences, Natural Language Processing, Medical & Health AI

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In General Social Sciences

Alec Radford

In Natural Language Processing

Citations in the corpus

Release Strategies and the Social Impacts of Language Models (2019, background): “These include biomedical literature analysis [4], generating synthetic test data [27], and generating radiology reports [37] and EEG reports [ 7].”
Natural Language Processing in Electronic Health Records in relation to healthcare decision-making: A systematic review (2023, background): “Utilising synthetic data is a possible solution [158, 159, 160].”
Synthetic data generation for tabular health records: A systematic review (2022, background): “Within healthcare, SDG has been researched for di ff erent modalities including biomedical signals [17], medical images [18], electronic health records (EHR) free-text content [19], time-series smart-home activity data [20] and EHR tabular data [21].”
Synthetic Data for Deep Learning (2019, background): “This abundance of generative models for text has not, however, led to any significant use of synthetic data for training NLP models; this has been done only in very restricted domains such as electronic medical records [216] (we discuss this field in detail in Section 7.”