MixCL: Pixel label matters to contrastive learning

TL;DR

MixCL introduces a novel pre-training framework for medical images that leverages pixel labels alongside image identities, enhancing representation robustness and significantly improving segmentation performance with limited labeled data.

Contribution

The paper proposes MixCL, a new contrastive learning framework that incorporates pixel labels into self-supervised pre-training for medical image analysis.

Findings

Improves Dice coefficient by 5.28% with 5% labeled data on Spleen.

Enhances Dice coefficient by 14.12% with 15% labeled data on BTVC.

Demonstrates robustness of representations in medical image segmentation.

Abstract

Contrastive learning and self-supervised techniques have gained prevalence in computer vision for the past few years. It is essential for medical image analysis, which is often notorious for its lack of annotations. Most existing self-supervised methods applied in natural imaging tasks focus on designing proxy tasks for unlabeled data. For example, contrastive learning is often based on the fact that an image and its transformed version share the same identity. However, pixel annotations contain much valuable information for medical image segmentation, which is largely ignored in contrastive learning. In this work, we propose a novel pre-training framework called Mixed Contrastive Learning (MixCL) that leverages both image identities and pixel labels for better modeling by maintaining identity consistency, label consistency, and reconstruction consistency together. Consequently, thus pre-trained model has more robust representations that characterize medical images. Extensive experiments demonstrate the effectiveness of the proposed method, improving the baseline by 5.28% and 14.12% in Dice coefficient when 5% labeled data of Spleen and 15% of BTVC are used in fine-tuning, respectively.