MedAug: Contrastive learning leveraging patient metadata improves representations for chest X-ray interpretation

By Y. Vu, Richard Wang, Niranjan Balachandar, Can Liu, Andrew Y. Ng, Pranav Rajpurkar. 2021

98 citations recorded by the source corpus.

Self-supervised contrastive learning between pairs of multiple views of the same image has been shown to successfully leverage unlabeled data to produce meaningful visual representations for both natural and medical images. However, there has been limited work on determining how to select pairs for medical images, where availability of patient metadata can be leveraged to improve representations. In this work, we develop a method to select positive pairs coming from views of possibly different images through the use of patient metadata. We compare strategies for selecting positive pairs for chest X-ray interpretation including requiring them to be from the same patient, imaging study or laterality. We evaluate downstream task performance by fine-tuning the linear layer on 1% of the labeled dataset for pleural effusion classification. Our best performing positive pair selection strategy, which involves using images from the same patient from the same study across all lateralities, achieves a performance increase of 14.4% in mean AUC from the ImageNet pretrained baseline. Our controlled experiments show that the keys to improving downstream performance on disease classification are (1) using patient metadata to appropriately create positive pairs from different images with the same underlying pathologies, and (2) maximizing the number of different images used in query pairing. In addition, we explore leveraging patient metadata to select hard negative pairs for contrastive learning, but do not find improvement over baselines that do not use metadata. Our method is broadly applicable to medical image interpretation and allows flexibility for incorporating medical insights in choosing pairs for contrastive learning.

Subfields: Artificial Intelligence, Epidemiology, Radiology, Nuclear Medicine and Imaging

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 Epidemiology

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

Transformers in Medical Imaging: A Survey (2022, background): “We believe such studies for ViT-based models, along with multi-instance contrastive learning to leverage patient meta data [391], will provide further insights to the community.”
Learning to Exploit Temporal Structure for Biomedical Vision-Language Processing (2023, background): “The closest work exploits patient metadata to select positive or negative examples in unimodal contrastive learning [66,78].”
BiomedCLIP: a multimodal biomedical foundation model pretrained from fifteen million scientific image-text pairs (2023, background): “MedAug [38] leverages patient metadata to select positive image pairs that go beyond augmentations of the same image.”