Unsupervised domain adaptation based COVID-19 CT infection segmentation network

TL;DR

This paper introduces an unsupervised domain adaptation segmentation network that leverages synthetic and unlabeled real COVID-19 CT images to improve infection area segmentation, achieving state-of-the-art results.

Contribution

It proposes a novel domain adaptation module and adversarial training scheme to enhance segmentation accuracy across different domains.

Findings

Achieves state-of-the-art segmentation performance on COVID-19 CT images.

Effectively aligns synthetic and real domain features.

Improves generalization of segmentation network to real data.

Abstract

Automatic segmentation of infection areas in computed tomography (CT) images has proven to be an effective diagnosis approach for COVID-19. However, due to the limited number of pixel-level annotated medical images, accurate segmentation remains a major challenge. In this paper, we propose an unsupervised domain adaptation based segmentation network to improve the segmentation performance of the infection areas in COVID-19 CT images. In particular, we propose to utilize the synthetic data and limited unlabeled real COVID-19 CT images to jointly train the segmentation network. Furthermore, we develop a novel domain adaptation module, which is used to align the two domains and effectively improve segmentation network's generalization capability to the real domain. Besides, we propose an unsupervised adversarial training scheme, which encourages the segmentation network to learn the domain-invariant feature, so that the robust feature can be used for segmentation. Experimental results demonstrate that our method can achieve state-of-the-art segmentation performance on COVID-19 CT images.