Joint semi-supervised and contrastive learning enables domain generalization and multi-domain segmentation

TL;DR

SegCLR is a novel semi-supervised and contrastive learning framework that enhances domain generalization and multi-domain segmentation in medical imaging, performing well with limited or no target domain data.

Contribution

We introduce SegCLR, a versatile framework combining supervised and contrastive learning for robust multi-domain segmentation, including zero-shot domain adaptation capabilities.

Findings

SegCLR achieves comparable results to supervised models in unsupervised domain adaptation.

The framework's performance is minimally affected by the amount of unlabeled target data.

SegCLR enables effective segmentation across multiple domains without target domain data.

Abstract

Despite their effectiveness, current deep learning models face challenges with images coming from different domains with varying appearance and content. We introduce SegCLR, a versatile framework designed to segment images across different domains, employing supervised and contrastive learning simultaneously to effectively learn from both labeled and unlabeled data. We demonstrate the superior performance of SegCLR through a comprehensive evaluation involving three diverse clinical datasets of 3D retinal Optical Coherence Tomography (OCT) images, for the slice-wise segmentation of fluids with various network configurations and verification across 10 different network initializations. In an unsupervised domain adaptation context, SegCLR achieves results on par with a supervised upper-bound model trained on the intended target domain. Notably, we discover that the segmentation performance of SegCLR framework is marginally impacted by the abundance of unlabeled data from the target domain, thereby we also propose an effective domain generalization extension of SegCLR, known also as zero-shot domain adaptation, which eliminates the need for any target domain information. This shows that our proposed addition of contrastive loss in standard supervised training for segmentation leads to superior models, inherently more generalizable to both in- and out-of-domain test data. We additionally propose a pragmatic solution for SegCLR deployment in realistic scenarios with multiple domains containing labeled data. Accordingly, our framework pushes the boundaries of deep-learning based segmentation in multi-domain applications, regardless of data availability - labeled, unlabeled, or nonexistent.