Multi-modal Masked Siamese Network Improves Chest X-Ray Representation Learning

TL;DR

This paper introduces a novel self-supervised learning approach that incorporates Electronic Health Records (EHR) data into a Masked Siamese Network to improve chest X-ray image representations, outperforming existing methods.

Contribution

It is the first to integrate EHR data into self-supervised pretraining for chest X-ray analysis using a Masked Siamese Network, enhancing representation quality.

Findings

Significant improvement over vanilla MSN and baselines in linear evaluation.

Effective use of demographic, scan metadata, and inpatient data.

Validated on three public chest X-ray datasets with two ViT backbones.

Abstract

Self-supervised learning methods for medical images primarily rely on the imaging modality during pretraining. While such approaches deliver promising results, they do not leverage associated patient or scan information collected within Electronic Health Records (EHR). Here, we propose to incorporate EHR data during self-supervised pretraining with a Masked Siamese Network (MSN) to enhance the quality of chest X-ray representations. We investigate three types of EHR data, including demographic, scan metadata, and inpatient stay information. We evaluate our approach on three publicly available chest X-ray datasets, MIMIC-CXR, CheXpert, and NIH-14, using two vision transformer (ViT) backbones, specifically ViT-Tiny and ViT-Small. In assessing the quality of the representations via linear evaluation, our proposed method demonstrates significant improvement compared to vanilla MSN and state-of-the-art self-supervised learning baselines. Our work highlights the potential of EHR-enhanced self-supervised pre-training for medical imaging. The code is publicly available at: https://github.com/nyuad-cai/CXR-EHR-MSN