DG-TTA: Out-of-domain Medical Image Segmentation through Augmentation and Descriptor-driven Domain Generalization and Test-Time Adaptation

TL;DR

This paper introduces a novel domain generalization and test-time adaptation method for medical image segmentation that leverages augmentation and descriptor-driven techniques to improve out-of-domain performance across various imaging modalities.

Contribution

It proposes a new combination of a generalizing SSC descriptor and GIN augmentation for pre-training, along with a test-time adaptation scheme, to enhance segmentation quality in unseen domains.

Findings

Significant performance improvements in cross-domain segmentation tasks.

Achieved up to 72.9% Dice score increase in spine imaging.

Method is effective across CT and MRI modalities.

Abstract

Purpose: Applying pre-trained medical deep learning segmentation models on out-of-domain images often yields predictions of insufficient quality. In this study, we propose to use a powerful generalizing descriptor along with augmentation to enable domain-generalized pre-training and test-time adaptation, achieving high-quality segmentation in unseen domains. Materials and Methods: In this retrospective study five different publicly available datasets (2012 to 2022) including 3D CT and MRI images are used to evaluate segmentation performance in out-of-domain scenarios. The settings include abdominal, spine, and cardiac imaging. The data is randomly split into training and test samples. Domain-generalized pre-training on source data is used to obtain the best initial performance in the target domain. We introduce the combination of the generalizing SSC descriptor and GIN intensity augmentation for optimal generalization. Segmentation results are subsequently optimized at test time, where we propose to adapt the pre-trained models for every unseen scan with a consistency scheme using the same augmentation-descriptor combination. The segmentation is evaluated using Dice similarity and Hausdorff distance and the significance of improvements is tested with the Wilcoxon signed-rank test. Results: The proposed generalized pre-training and subsequent test-time adaptation improves model performance significantly in CT to MRI cross-domain prediction for abdominal (+46.2% and +28.2% Dice), spine (+72.9%), and cardiac (+14.2% and +55.7% Dice) scenarios (p<0.001). Conclusion: Our method enables optimal, independent usage of medical image source and target data and bridges domain gaps successfully with a compact and efficient methodology. Open-source code available at: https://github.com/multimodallearning/DG-TTA