TCSA-UDA: Text-Driven Cross-Semantic Alignment for Unsupervised Domain Adaptation in Medical Image Segmentation

TL;DR

TCSA-UDA introduces a text-driven framework for unsupervised domain adaptation in medical image segmentation, aligning visual features with semantic class descriptions to improve cross-modality performance.

Contribution

It proposes a novel vision-language covariance cosine loss and prototype alignment module to enhance semantic and modality-invariant feature learning.

Findings

Significantly reduces domain shift in medical image segmentation.

Outperforms state-of-the-art UDA methods on multiple benchmarks.

Effectively leverages textual descriptions for cross-modal alignment.

Abstract

Unsupervised domain adaptation for medical image segmentation remains a significant challenge due to substantial domain shifts across imaging modalities, such as CT and MRI. While recent vision-language representation learning methods have shown promise, their potential in UDA segmentation tasks remains underexplored. To address this gap, we propose TCSA-UDA, a Text-driven Cross-Semantic Alignment framework that leverages domain-invariant textual class descriptions to guide visual representation learning. Our approach introduces a vision-language covariance cosine loss to directly align image encoder features with inter-class textual semantic relations, encouraging semantically meaningful and modality-invariant feature representations. Additionally, we incorporate a prototype alignment module that aligns class-wise pixel-level feature distributions across domains using high-level semantic prototypes. This mitigates residual category-level discrepancies and enhances cross-modal consistency. Extensive experiments on challenging cross-modality cardiac, abdominal, and brain tumor segmentation benchmarks demonstrate that our TCSA-UDA framework significantly reduces domain shift and consistently outperforms state-of-the-art UDA methods, establishing a new paradigm for integrating language-driven semantics into domain-adaptive medical image analysis.