Preservational Learning Improves Self-supervised Medical Image Models by Reconstructing Diverse Contexts

TL;DR

This paper introduces Preservational Contrastive Representation Learning (PCRL), a method that combines contrastive learning with explicit context reconstruction to enhance self-supervised medical image representations, leading to superior performance in various tasks.

Contribution

It proposes a novel Preservational Learning approach that explicitly reconstructs diverse image contexts, improving information preservation in self-supervised medical image models.

Findings

PCRL outperforms existing self-supervised methods in 5 classification/segmentation tasks.

PCRL surpasses supervised models in several benchmarks.

Explicit context reconstruction enhances representation quality.

Abstract

Preserving maximal information is one of principles of designing self-supervised learning methodologies. To reach this goal, contrastive learning adopts an implicit way which is contrasting image pairs. However, we believe it is not fully optimal to simply use the contrastive estimation for preservation. Moreover, it is necessary and complemental to introduce an explicit solution to preserve more information. From this perspective, we introduce Preservational Learning to reconstruct diverse image contexts in order to preserve more information in learned representations. Together with the contrastive loss, we present Preservational Contrastive Representation Learning (PCRL) for learning self-supervised medical representations. PCRL provides very competitive results under the pretraining-finetuning protocol, outperforming both self-supervised and supervised counterparts in 5 classification/segmentation tasks substantially.