Two Directions for Clinical Data Generation with Large Language Models: Data-to-Label and Label-to-Data

TL;DR

This paper explores how large language models can generate synthetic clinical data to improve detection of Alzheimer's signs in electronic health records, demonstrating that augmented datasets enhance system performance while preserving privacy.

Contribution

It introduces two novel methods for clinical data generation using LLMs—data-to-label and label-to-data—and shows their effectiveness in improving AD detection systems.

Findings

Synthetic data improves AD detection accuracy.

Label-to-data method maintains privacy without sacrificing quality.

Augmented datasets outperform using only expert-annotated data.

Abstract

Large language models (LLMs) can generate natural language texts for various domains and tasks, but their potential for clinical text mining, a domain with scarce, sensitive, and imbalanced medical data, is underexplored. We investigate whether LLMs can augment clinical data for detecting Alzheimer's Disease (AD)-related signs and symptoms from electronic health records (EHRs), a challenging task that requires high expertise. We create a novel pragmatic taxonomy for AD sign and symptom progression based on expert knowledge, which guides LLMs to generate synthetic data following two different directions: "data-to-label", which labels sentences from a public EHR collection with AD-related signs and symptoms; and "label-to-data", which generates sentences with AD-related signs and symptoms based on the label definition. We train a system to detect AD-related signs and symptoms from EHRs, using three datasets: (1) a gold dataset annotated by human experts on longitudinal EHRs of AD patients; (2) a silver dataset created by the data-to-label method; and (3) a bronze dataset created by the label-to-data method. We find that using the silver and bronze datasets improves the system performance, outperforming the system using only the gold dataset. This shows that LLMs can generate synthetic clinical data for a complex task by incorporating expert knowledge, and our label-to-data method can produce datasets that are free of sensitive information, while maintaining acceptable quality.