Contrastive Learning with Temporal Correlated Medical Images: A Case Study using Lung Segmentation in Chest X-Rays

TL;DR

This paper introduces CL-TCI, a contrastive learning framework leveraging temporal correlations in medical images, specifically chest X-rays, to improve lung segmentation performance in semi-supervised and transfer learning scenarios.

Contribution

It adapts state-of-the-art contrastive learning methods MoCo and SimCLR to incorporate temporal correlations in medical images for segmentation tasks.

Findings

CL-TCI outperforms baselines without temporal correlation in multiple datasets.

MoCo-based CL-TCI generally performs better than SimCLR-based version.

Temporal correlation enhances contrastive learning effectiveness in medical image segmentation.

Abstract

Contrastive learning has been proved to be a promising technique for image-level representation learning from unlabeled data. Many existing works have demonstrated improved results by applying contrastive learning in classification and object detection tasks for either natural images or medical images. However, its application to medical image segmentation tasks has been limited. In this work, we use lung segmentation in chest X-rays as a case study and propose a contrastive learning framework with temporal correlated medical images, named CL-TCI, to learn superior encoders for initializing the segmentation network. We adapt CL-TCI from two state-of-the-art contrastive learning methods-MoCo and SimCLR. Experiment results on three chest X-ray datasets show that under two different segmentation backbones, U-Net and Deeplab-V3, CL-TCI can outperform all baselines that do not incorporate any temporal correlation in both semi-supervised learning setting and transfer learning setting with limited annotation. This suggests that information among temporal correlated medical images can indeed improve contrastive learning performance. Between the two variations of CL-TCI, CL-TCI adapted from MoCo outperforms CL-TCI adapted from SimCLR in most settings, indicating that more contrastive samples can benefit the learning process and help the network learn high-quality representations.