Learning Generalized Medical Image Representations through Image-Graph Contrastive Pretraining
This work addresses the annotation burden for medical experts in deep learning applications, with incremental improvements in medical image understanding.
The paper tackled the problem of reducing expert annotation needs in medical image interpretation by developing an Image-Graph Contrastive Learning framework that pairs chest X-rays with knowledge graphs from radiology notes, achieving superior performance to existing methods in 1% linear evaluation and few-shot settings on the CheXpert dataset.
Medical image interpretation using deep learning has shown promise but often requires extensive expert-annotated datasets. To reduce this annotation burden, we develop an Image-Graph Contrastive Learning framework that pairs chest X-rays with structured report knowledge graphs automatically extracted from radiology notes. Our approach uniquely encodes the disconnected graph components via a relational graph convolution network and transformer attention. In experiments on the CheXpert dataset, this novel graph encoding strategy enabled the framework to outperform existing methods that use image-text contrastive learning in 1% linear evaluation and few-shot settings, while achieving comparable performance to radiologists. By exploiting unlabeled paired images and text, our framework demonstrates the potential of structured clinical insights to enhance contrastive learning for medical images. This work points toward reducing demands on medical experts for annotations, improving diagnostic precision, and advancing patient care through robust medical image understanding.